Qr Code
-
WeChat
- Current Discuss
- First Published
- Album Paper
-
A temperature and humidity compensation method for microwave ammonia gas sensors experiencing frequency drift
Shi Guolong, Hu Guoping, He Yigang, Meng Fanli
Abstract:
High-precision detection of harmful gases is urgently required in key sectors, such as livestock farming, agricultural product quality monitoring, and industrial environmental management. However, fluctuations in indoor ambient temperature and humidity can lead to frequency drift in gas sensors, thereby affecting detection accuracy. To address this issue, electromagnetic simulations are conducted to analyze the electromagnetic loss characteristics of the microstrip resonator, thereby identifying the optimal coating position for the gas-sensitive material and enhancing the microwave sensor′s sensitivity to ammonia. Furthermore, the correlation between the sensor′s radiation gain and ammonia concentration. A wireless ammonia detection system based on a wireless power transmission model is constructed. By utilizing the detection principles of radio frequency identification, an experimental platform is developed to test sensor performance under various temperature and humidity conditions. The back propagation (BP) neural network temperature-humidity compensation algorithm is introduced to the model, analyze, and correct the frequency drift caused by environmental variations, combined with Pearson correlation analysis. Experimental results indicate that temperature and humidity significantly affect the microwave ammonia sensor′s frequency stability. After compensation, the frequency drift amplitude is reduced by 14 MHz, the concentration error is decreased to 0.06×10-6, and the relative error is limited to 2%, resulting in a 31.11% improvement in gas detection accuracy. Compared with the temperature compensation model of the BP network or the temperature-humidity compensation model of the support vector machine, the proposed method demonstrates superior performance. In conclusion, this research effectively enhances the detection accuracy of microwave ammonia gas sensors under complex temperature and humidity environmental conditions. It provides a more robust technical foundation for high-precision harmful gas monitoring.
-
Research on the magnetic suspension control method of pendulous integrating gyro accelerometer under transverse acceleration
Wang Long, Li Liang, Sheng Andong, Zhu Zhigang, Ren Moxuan
Abstract:
A pendulous integrating gyro accelerometer is installed in the inertial navigation system as a single-axis sensitive accelerometer. It is generally used to sensitive apparent acceleration in the direction of the input axis. Meanwhile, it will be subjected to transverse acceleration, which is perpendicular to the direction of the input axis. The PIGA is subjected to transverse acceleration, which generates a large amplitude of the float cyclic swing around the input axis, thus affecting the accuracy of the instrument. To address this problem, the force analysis of the float is carried out, and the dynamic equations of radial motion of the float have been established under the simultaneous action of transverse acceleration and sensitive acceleration. A proportional-integral control method of magnetic suspension is proposed on the basis of the purely proportional control method. The new method can provide a greater stiffness for the magnetic suspension system to overcome pendulous moments and eliminate the static position error. It would suppress the amplitude of the float swing. A Simulink simulation and comparison are conducted on the radial motion of the float under two control methods. When the transverse acceleration 0.866g and the sensitive acceleration 0.5g act simultaneously, the amplitude of the float swing decreases from ±0.62 μm to ±0.24 μm. When the transverse acceleration 3g and the sensitive acceleration 1g act simultaneously, the amplitude of the float swing decreases from ±2.2 μm to ±0.74 μm. By carrying out centrifuge tests on PIGA, it is found that the amplitude of the float swing decreased from ±2 μm to ±0.5 μm at a transverse acceleration 3.15g and a sensitive acceleration 0.313g. When the transverse acceleration and the sensitive acceleration change at the same time, the amplitude of the float swing is decreased by about 50%. The results show that the proportional-integral control method of magnetic suspension can significantly suppress the amplitude of the float swing and really reduce the swing amplitude by 50% or more compared with the pure proportional control method of magnetic suspension. The new method has practical engineering significance and can improve the instrumentation measurement accuracy when PIGA is subjected to lateral acceleration in use.
-
Design and experimental study of a dual-stripe temperature-resistant optical current transformer based on Koopman filtering
Zhang Jing, Li Yansong, Zhang Ziao, Hou Kaiyun, Liu Jun
Abstract:
Optical current transformers align with the intelligent and digital development needs of new power systems. However, the existing optical current transformers lack temperature-resistant solutions for wide temperature range measurements in power systems that simultaneously meet real-time processing requirements and ensure long-term operational stability. To address these issues, a dual-strip temperature-immune magneto-optical current transformer (DS-MOCT) based on the Koopman adaptive filtering is proposed. First, the principles of magneto-optical sensing and the temperature perturbation mechanism in magneto-optical sensing are introduced. By using the Jones matrix, a polarization analytical model for optical current sensing at arbitrary angles of the polarizer-analyzer in a straight-through optical path configuration is formulated. Subsequently, the influence of varying polarizer-analyzer angles on the transformer′s output is analyzed. A temperature compensation method for specific angles of the polarizer-analyzer is proposed. This enabled the construction of a temperature-resistant DS-MOCT structure comprising a measurement arm and a temperature compensation arm. The measurement arm outputs the measured current information, while the temperature compensation arm generates temperature compensation signals. These compensation signals are applied in real-time to counteract temperature disturbances in the measured current signal. Consequently, the DS-MOCT produces temperature-immune measurement current values resilient to thermal perturbations. The error sources of DS-MOCT are analyzed, and a Koopman theory-based denoising method is proposed. Subsequently, finite element simulations are performed to model the DS-MOCT′s multi-physics coupling environment. Results demonstrate the optical wave′s thermal stability by visualizing its temperature resistance behavior. Finally, the DS-MOCT experimental system with hardware-software co-design is constructed. The experimental results show that, within a broad temperature range of -40℃ to 40℃, the DS-MOCT exhibits a measurement error of less than 0.2%, complying with the Class 0.2 measurement standard for electronic transformers specified in GB/T 20840.8—2007. The dynamic response time remains under 14 ms, satisfying the real-time monitoring requirements of power systems. The proposed Koopman adaptive filtering-based DS-MOCT resolves the trilemma of temperature resistance, real-time performance, and long-term operational stability in optical current transformers within new power systems.
-
Research on the performance of piezoelectric multi-dimensional force sensor based on d33 mode
Li Fen, Hou Xiaojuan, Zhang Xin, Fu Xiaoyan, He Jian
Abstract:
With the rapid development of modern manufacturing, robotics, precision measurement and other fields, the demand for multi-dimensional force sensors to accurately measure multi-directional forces has become increasingly prominent. Research on piezoelectric multi-dimensional force sensors has been initiated to address the problem of improving the accuracy of force control and feedback in robot arms and complex mechanical systems. Based on the d33 mode of piezoelectric ceramics, this paper designs a piezoelectric multi-dimensional force sensor that can accurately measure forces in the X, Y, and Z directions. The sensor adopts a cubic structure, and the piezoelectric sheets are placed vertically to the X, Y, and Z axes, respectively to achieve independent measurement and self-decoupling of multi-dimensional forces. The parameters and meshes are set in COMSOL Multiphysics to simulate and analyze the stress distribution and deformation of the overall structure when loads are applied in different directions. A three-dimensional sensor measurement system is established by combining a microcomputer-controlled electronic universal testing machine, sensors, and signal acquisition equipment to achieve three-dimensional force measurement and analysis. The study shows that the force in each loading direction can be accurately applied to the target piezoelectric sheet through a reasonably designed transfer path, and it is able to test the normal force of 0~500 N (Z direction) and the shear force of 0~200 N (X and Y directions). The designed sensor has good dynamic response capability, high sensitivity, precise directional resolution high-precision directionality and relatively stable repeatability, with a sensitivity of 0.080 V/N in the Z direction, 0.110 V/N in the X direction, and 0.113 V/N in the Y direction, and crosstalk in each direction is less than 0.2%. It has a broad application prospect in the field of multi-dimensional force measurement, providing important experimental data and theoretical support for the design and optimization of multi-dimensional force sensors.
-
Design of FBG tactile sensing unit based on graphene-silicone composite packaging
Sun Shizheng, Chen Shengkang, He Jiang, Dong Shaojiang
Abstract:
The level of refined operation of a robotic hand depends on its fingertip tactile perception performance. To enhance the tactile perception performance of FBG-based robotic fingertips, a diagonal cross-shaped FBG tactile perception unit was developed, featuring a flexible packaging structure composed of a graphene-silicone composite material. This design addressed key challenges such as poor thermal conductivity and the cross-sensitivity between contact force and contact temperature commonly observed in conventional FBG soft perception units. To resolve the coupling issue between force and temperature, a decoupling method based on an Osprey optimization algorithm optimized convolutional neural network (OOA-CNN) was proposed. First, simulation analyses were conducted to compare the temperature response and strain response of FBG sensors embedded in graphene-silica gel composite packaging versus pure silica gel packaging. Then, experimental analyses were performed to investigate the impact of different graphene mass fractions (1%, 1.5%, 2%, 2.5%, and 3%) on the thermal conductivity of composite materials. Using a three-fingered robotic hand, fingertip perception experiments were carried out to evaluate the sensitivity of the FBG tactile perception unit to both contact force and contact temperature. Finally, coupled analysis was performed on the composite perception data of contact force and temperature. The decoupling performance of the proposed OOA-CNN model was validated through comparative experiments against a standard CNN model and a least squares method. Simulation and experimental results show that incorporating 1.5% mass fraction graphene as a thermally conductive filler in the silicone matrix significantly enhances its thermal conductivity while preserving the FBG′s tactile sensing performance. The FBG tactile unit exhibited a sensitivity of 31.281 pm/N to contact force and 10.787 pm/℃ to contact temperature. Furthermore, the OOA-CNN decoupling model has a better decoupling effect compared to the least squares method and the CNN decoupling model, with an average absolute error reduction of 40.3% for contact temperature and 41.33% for contact force.
-
Data fusion method for multi-line laser sensor based on KD-Tree acceleration
Li Xinfei, Yan Ran, Xia Lei, Zhao Qing, Zhang Kaifei
Abstract:
To address the challenges of low fusion efficiency, significant stitching errors, and high processing complexity in large-scale point cloud data fusion during collaborative scanning with multiple line laser sensors, this paper proposes a KD-Tree-accelerated data fusion method for multi-line laser sensor systems. The method leverages a dynamic neighborhood search strategy and an adaptive radius adjustment mechanism to enable efficient ordering and parallel smoothing optimization of point cloud data. First, a KD-Tree spatial indexing structure is constructed, and an innovative dynamic neighborhood search strategy is designed to rapidly reorganize disordered 2D contour data into ordered sequences, reducing the algorithm′s time complexity from the traditional O(n2logn) to O(nlogn). Second, by integrating OpenMP multi-threaded parallel computing with an improved Moving Least Squares algorithm, a K-MLS parallel smoothing method is proposed, optimizing the time complexity from O(n2) to O(nlogn), significantly enhancing the processing efficiency for large-scale point clouds. The proposed method was validated in a train wheel measurement system. When processing 2.09 million points, the sorting algorithm achieved a 35.7-fold speedup compared to traditional approaches, while the smoothing algorithm exhibited an 84.5-fold performance improvement. Comparative analysis further demonstrates the method′s effectiveness in improving point cloud quality: it successfully fills data gaps, reduces the maximum deviation in wheel tread measurement from ±0.279 mm to ±0.085 mm, and lowers the mean square error of 3D point cloud registration from 0.323 mm to 0.106 mm. Experimental results confirm that the proposed method maintains sub-millimeter accuracy while significantly boosting processing efficiency for million-scale point clouds. It effectively addresses issues such as stitching errors and uneven point density in overlapping regions during multi-sensor data fusion, proving its robustness and applicability in industrial online measurement scenarios.
-
Dual-source perception indoor localization algorithm under dynamic radiation conditions
Wu Yan, Qin Ningning, Song Shulin, Wang Yan
Abstract:
In underground garage environments, traditional RSSI-based fingerprint positioning is compromised by fluctuating radiation, multipath effects, and interference, leading to feature distortion and positioning errors. This paper proposes a dual-source sensing indoor positioning method that integrates environmental radiation perception with signal analysis to enhance system robustness. In the offline phase, a bidirectional fusion model combining BiLSTM and BiGRU is employed to capture both long-term and short-term radiation effects on RSSI. By leveraging multi-head self-attention, the model constructs an adaptive fingerprint database that accommodates varying radiation conditions. During the online matching phase, an exponential power normalization technique is used to map RSSI signals to a unified scale, mitigating hardware-related interference. An AP-aware clustering algorithm is introduced to select RPs based on AP signal quality and suppress matching deviations through density estimation. Experimental results in underground garages demonstrate the method′s strong performance. Under known radiation conditions, it achieves an average positioning accuracy improvement of 11.05%~25.38% over baseline methods. Under unknown radiation conditions, the BGLA-constructed fingerprint database enables it to outperform comparative approaches by 27.55%~35.71% in average positioning accuracy.
-
Structural design and verification of the hyperspectral imager for the Gaofen-5 Satellite
Liu Shufeng, Liu Yinnian, Cao Kaiqin, Ke Youlong, Jia Xiaowei
Abstract:
The visible and shortwave infrared hyperspectral imager (AHSI) onboard the Gaofen-5 satellite represents a core payload within China′s high-resolution Earth observation program. Unlike conventional off-axis three-mirror cameras primarily used for visible or multispectral imaging, hyperspectral imagers require a specialized architecture comprising a telescope, spectrometer, and detector array. The AHSI features an extended 3-meter optical path integrating three off-axis systems and 22 off-axis optical components, presenting significant challenges in terms of structural complexity, alignment precision, and environmental robustness within constrained space. To address these challenges, this study presents the design and implementation of a novel integrated structural configuration using advanced composite materials. A systematic methodology encompassing material selection, structural design, simulation analysis, and experimental validation is established. The opto-mechanical framework utilizes a high-volume fraction (55% ) SiCp/Al composite as the primary structural material, offering superior stiffness-to-weight performance. The design unifies the off-axis three-mirror telescope, Offner spectrometer, and detector assemblies into a compact and rigid composite structure, optimizing mechanical stability and mass efficiency. The framework is fabricated using an ultrasonic-assisted gradient brazing process to ensure precise bonding and structural integrity. Structural strength and stiffness are confirmed through finite element modeling and mechanical testing. On-orbit imaging results show strong agreement with ground-based calibration data, verifying the design′s environmental adaptability and long-term stability. As the first hyperspectral imaging instrument worldwide to achieve simultaneous wide swath coverage and broad spectral range, AHSI demonstrates the viability of advanced composite-based structural solutions for complex spaceborne optical systems. It delivers critical technical support for strategic national applications, including land resource management and environmental monitoring.
-
A differential confocal comprehensive measurement method of lens geometry multi-parameters based on reference ring positioning
Liu Jinchen, Liu Yuhan, Ying Ronghui, Qiu Lirong, Zhao Weiqian
Abstract:
The cumulative effect of geometric parameter errors of a single lens has a particularly prominent impact on the overall imaging quality of the optical system. Among various geometric parameters of the lens, the center thickness, wedge error and center deviation have a more significant influence on the imaging quality. Among various geometric parameters of the lens, the center thickness, wedge error, and center deviation have a significant influence on the imaging quality. To address the difficult problems in measuring geometric parameters such as thickness, center deviation, and wedge error of optical lenses and characterizing the correlation of multi-surface shapes, this article proposes a differential confocal optical lens geometric multi-parameter comprehensive measurement method based on reference ring-assisted positioning. This method uses the characteristics of the precise correspondence between the zero point of the laser differential confocal measurement curve and the focus of the sensor to achieve high-precision fixed-focus measurement of a single surface of the optical lens. Through the common reference ring-assisted positioning method, the precise positioning of the optical lens during the flip measurement of the double-sided surface is achieved. Through the reference ring attitude alignment method, the position correlation alignment and surface reconstruction of the double-sided surface of the optical lens are achieved. Finally, the comprehensive measurement and evaluation of the optical lens geometric multi-parameters are achieved. By conducting comprehensive measurements of multiple geometric parameters of the lens, the overall system can be optimized in a targeted manner, reducing repetitive positioning errors and cumulative errors. This helps improve the system′s accuracy, enhances imaging quality, and ultimately boosts the overall comprehensive performance of the optical system. Experimental verification shows that the measurement error of the optical lens thickness is less than 1.200 μm, the measurement error of the center deviation is better than 1.000 μm, and the wedge error is less than 0.002°. This method provides a new technical approach for the measurement of optical lens geometric multi-parameters.
-
Self-calibration identification method for rotary axis geometric errors based on tower-shaped workpiece
Ni Juntao, Xiang Sitong, Chen Kejian, Zhang Hainan, Yang Jianguo
Abstract:
With the growing demand for manufacturing precision, efficiently identifying geometric errors in five-axis machine tools has become essential for achieving high-accuracy machining. This paper presents a self-calibration method utilizing a tower-shaped workpiece, enabling the simultaneous identification of geometric errors in the rotary axis (C-axis), linear axes, and the workpiece itself. The approach involves designing a five-tier stepped tower-shaped workpiece and incorporating a three-dimensional volumetric error model for the linear axes, which discretizes linear axis errors at grid nodes within a 3D space. An overdetermined system of equations is constructed from multi-angle probing data, and the error parameters are estimated using the least squares method. Experimental results successfully identify four position-independent and six position-dependent geometric errors of the C-axis, alongside the geometric errors of the linear axes and the workpiece. The findings demonstrate the method′s high stability in identifying C-axis errors under repeated clamping, confirming its potential for long-term monitoring. To further assess accuracy, a comparative experiment using a disk-shaped reference workpiece was conducted, revealing an average consistency of 89.8% in error identification results. A Monte Carlo-based uncertainty analysis confirms the method′s robustness and reliability in the presence of measurement system errors and environmental disturbances. Importantly, the proposed method does not rely on high-end measurement equipment or complex path planning. It supports repeatable clamping and automated measurement, offering advantages such as ease of operation, low cost, and strong adaptability. This makes it a practical and efficient solution for high-precision geometric error identification in five-axis machine tools.
-
Research on full-wavelength ultrasonic attenuation model of fluid-solid two-phase flow particle parameters
Wang Mi, Liu Jiegui, Tong Kaijie, Cui Xueli, Fang Lide
Abstract:
With the development of industrial process precision, fluid-solid particle two-phase flow systems have been widely used in the chemical industry, energy, and other fields. According to the different mediums in the continuous phase, the fluid-solid two-phase flow can be divided into gas-solid and liquid-solid two-phase flow, in which the accurate measurement of particle size and concentration characteristic parameters are very important for process control and efficiency optimization. To address the problem that the classical ultrasonic attenuation model is difficult to adapt to different medium types and cannot cover the full wavelength range, this article analyzes the advantages and disadvantages of the classical theoretical model based on the numerical simulation method of thermal viscous acoustic physical field. The coupled superposition method of the McClements model and the BLBL model are adopted to achieve the comprehensive characterization of different attenuation mechanisms. An ultrasonic attenuation McBL model suitable for low-concentration fluid-solid two-phase flow in the full wavelength region is formulated. Combined with the particle swarm optimization algorithm, the simultaneous inversion of particle size and concentration is achieved. The simulated attenuation coefficient is taken as the inversion input to preliminarily verify the applicability of the model. Compared with the particle parameters of the simulation model, the simulation results show that the errors of the average particle size and concentration obtained by inversion are both within ±15%. In addition, an ultrasonic attenuation experimental device is set up to measure the ultrasonic attenuation coefficient of the fluid-solid particle two-phase flow employing a dual-frequency transmission method, and the McBL model combined with particle swarm optimization algorithm is used to invert the experimental attenuation coefficient. Compared with the measurement results of the laser particle size analyzer, the experimental results show that the average particle size error of the particles is within ±10%. This further verifies that the McBL model combined with the particle swarm optimization algorithm has high accuracy in the inversion of characteristic parameters in the full wavelength region of fluid-solid two-phase flow. This research is of great significance for the online precise measurement of characteristic parameters of flu-solid two-phase flow.
-
Research on double-sided anti-metal UHF RFID tag with spiral loop structure
Lai Wangfu, Shi Zibin, Lin Changgui, Yuan Jiade
Abstract:
A miniaturized double-sided anti-metal UHF RFID tag with a spiral loop structure is proposed. The tag antenna comprises three metallic layers: a middle radiation patch, a lower ground plane, and an upper ground plane, separated by two layers of 1 mm-thick foam substrates. The middle radiation patch consists of an external ring patch and a central spiral loop patch, with the tag chip positioned between them. The middle radiation patch and the upper ground plane together generate effective radiation when the lower ground plane is mounted on the metallic plate. Conversely, the middle radiation patch and the lower ground plane act as the radiating pair when the upper ground plane is mounted on the metallic plate. This design enables the proposed antenna to achieve both double-sided anti-metal performance and miniaturization. Theoretical analysis and simulation results confirm that adjusting the length of the spiral loop patch in the middle layer effectively tunes the antenna′s inductive reactance, which enables conjugate impedance matching between the antenna and the chip. The tag antenna has dimensions of 35 mm×20 mm×2.15 mm. Simulation and experimental results demonstrate that the maximum power transfer coefficient between the tag antenna and the chip exceeds 98% for both sides. When the two ground planes of the tag antenna are individually mounted on a 200 mm×200 mm metallic plate under an effective isotropic radiated power (EIRP) of 3.28 W, within the 902~928 MHz frequency range, the read distances reach 7.5 and 7.6 m, respectively. The minimum sensitivity is -14 dBm for both sides. The proposed tag offers advantages such as compact size, consistent double-sided anti-metal performance, and long read distance. It effectively enhance the identification capability and stability of suspended tags in metallic scenarios, providing an efficient solution for RFID tag applications in complex metallic scenarios.
-
Research on losses measurement of magnetic component with wide-range frequency excitation based on calorimetric method
Wang Jinghui, Lin Shuang, Fan Peng, Huang Fubin, Chen Wei
Abstract:
When measuring the losses of magnetic component excited at frequencies up to 100 MHz using electrical measurement methods, the high-frequency parasitic parameters can lead to significant measurement errors. The calorimetric method is utilized to measure the losses of magnetic component under ultra-high-frequency excitation in this paper. The errors of the closed calorimetric method stem from the determination of specific heat capacity, heat dissipation, and heat from accessories. Typically, the calibration calorimetric method is employed to eliminate the measurement errors of the closed calorimetric method. Traditional calorimetric method uses direct current (DC) power as a standard to verify the power-temperature rise (P-ΔT) relationship. The errors of the DC power calibration calorimetric method arise from inconsistent environmental conditions between the calibration and measurement processes, particularly the difference between the DC and AC equivalent resistance of connecting wires, which causes measurement errors that increase with the excitation frequency. This paper proposes an alternating current (AC) power calibration calorimetry method for measuring the losses of magnetic components under ultra-high-frequency excitation. By using AC power as a standard to verify the P-ΔT relationship, and ensuring the same excitation frequency during both calibration and measurement, this method eliminates measurement errors caused by inconsistent of connecting wire losses. The remaining measurement errors in this approach mainly result from the influence of high-frequency parasitic parameters. Following a detailed analysis of the principles and error sources associated with the calibration-based calorimetric method, corresponding solutions are proposed. A measurement platform for magnetic component loss evaluation is constructed, and an upper computer measurement interface is developed to enable automated measurement. Finally, an air-core inductor with accurately measurable losses is used as the inductive device under test. Experimental results verify that the calibration calorimetric measurement platform can accurately measure the losses of magnetic components excited by sinusoidal waves within the frequency range of 100 MHz.
-
Research on blind source separation of rain sound signal based on the improved ICA
Abstract:
An improved independent component analysis (ICA) method is proposed to address the issues of amplitude amplification and poor separation performance in blind source separation of rain sound signals using the traditional fast independent component analysis (FASTACA) method based on objective functions such as negative entropy. Traditional complex objective functions (negative entropy, kurtosis, etc.) are no longer used. Instead, we choose non-Gaussianity based on maximizing the signal. A combination of the hyperbolic cosine function and the logarithmic function for nonlinear transformation is utilized. Meanwhile, we reconstruct the objective function based on the mean difference square between the source signal and the separated signal. To improve the algorithm′s running, convergence speed, and optimization ability, the particle swarm optimization (PSO) algorithm is utilized instead of the traditional gradient descent method. Its fast global search ability is adopted to optimize the objective function, effectively avoiding the problem of ICA getting trapped in local optima during the iteration process. After obtaining the optimal solution mixing matrix, the rain sound mixed signal and extracting a purer rain sound signal are separated. The experimental results show that the improved ICA can meet the requirements of blind source separation, and the separation index (PI) reaches a level of 10-2. To further evaluate the effectiveness and stability of the proposed algorithm, blind source separation experiments are conducted in mixed scenarios of different types of rain sounds and environmental noise. The results show that the improved ICA algorithm can effectively separate and recover the source rain sound signal in mixed signals under different environmental noise backgrounds. In addition, comparing the ICA algorithm with the improved objective function and the FASTACA algorithm based on negative entropy, the proposed algorithm not only effectively solves the amplitude expansion problem caused by the FASTACA algorithm, but also converges faster, reducing the mean square error (MSE) by two orders of magnitude. The signal distortion ratio (SDR) under the rain sound type is increased by nearly 20 dB.
-
A three-winding coupled-inductor bipolar-output high-gain DC-DC converter
Chen Xianjun, Zhou Mingzhu, Shang Yunhui, Mao Xingkui, Zhang Yiming
Abstract:
The bipolar DC microgrid system represents an innovative power supply architecture offering high reliability, flexibility, and efficiency. Building upon advanced research findings, this study proposes a three-winding coupled-inductor bipolar-output high-gain DC-DC converter. By incorporating coupled inductor coils and a switched-capacitor boosting mechanism into a bipolar-output three-level Boost converter topology, the proposed converter achieves low input current ripple and reduced stress on switching devices. The output voltages can be regulated via the switch duty cycle and the turns ratio of the coupled inductor. This paper presents the topology derivation, operating principles, and mode analysis of the proposed converter. The voltage gain, along with the voltage and current stresses on key components, is theoretically derived and numerically evaluated. A comparative analysis is conducted against existing high-gain DC-DC converters to highlight the advantages in efficiency and performance. Furthermore, the efficiency of the converter is thoroughly analyzed. An experimental prototype was developed to validate the converter′s performance. The prototype operates at a switching frequency of 50 kHz, with an input voltage of 28 V and an input power of 200 W, delivering bipolar output voltages of +190 V and -190 V. A suitable input inductor is selected based on the desired input current ripple ratio. Full-load tests (200 W) were conducted at input voltages of 24, 28, and 32 V, with the system successfully boosting to a total output voltage of 380 V. Experimental waveforms of input current and output voltage are presented under both half-load and full-load conditions, demonstrating efficiencies of 95.65% and 93.63%, respectively. These results confirm that the proposed converter performs efficiently under high-frequency operation, making it highly suitable for applications requiring high efficiency and compact design. The converter shows strong potential for widespread application and holds significant research value in the field of electric power systems.
-
Research on ocean observation system based on the acoustic scattering of shallow sea bubbles
Cheng Yuezhu, Shi Jie, Cao Yuan, Fu Xiaoyue
Abstract:
Research on marine observation equipment is crucial for marine infrastructure, particularly the environmental parameter measurement systems that effectively capture physical properties of shallow seas. This paper presents a method to characterize the relationship between dynamic medium parameters, acoustic excitation conditions, and acoustic field evolution based on the theory of acoustic scattering by shallow sea bubble groups. It details how the distribution of bubble groups relates to the medium′s acoustic parameters and leverages their unique scattering features to measure environmental parameters. The system employs narrow frequency-swept pulses to measure the incident sound pressure as well as the forward and backscattered sound pressure amplitudes of the target medium. From these measurements, the scattering and attenuation coefficients are obtained in real time and interpreted using bubble scattering and nonlinear parameter theories to derive shallow sea environmental parameters. Field tests conducted in a lake environment validate the measured scattering coefficients. Notably, while the nonlinear parameter of pure water is known to be about 3.5, measurements indicate values up to 90 in shallow seas, confirming that bubbles significantly increase medium nonlinearity and that pure water parameters cannot approximate shallow sea conditions. These results highlight the critical role of nonlinear parameter measurement. Compared with conventional systems, the proposed measurement system is simpler, highly adaptable to complex environments, and capable of rapidly measuring shallow sea environmental parameters. Importantly, it addresses the challenge of swiftly measuring nonlinear parameters in shallow sea settings, underscoring its practical significance.
-
An improved camera-LiDAR fusion perception algorithm in the bird′s-eye view perspective
Abstract:
In autonomous driving perception tasks, a multi-modal fusion of camera and LiDAR features based on a bird′s-eye view has become a mainstream research paradigm to combine information from different modalities into a unified spatial representation. Although representative frameworks such as BEVFusion achieve high 3D object detection accuracy, they rely heavily on depth prediction during the perspective transformation from 2D image features to the BEV space. This depth module is often complex, parameter-intensive, and results in low inference efficiency and high memory consumption, posing challenges for deployment on edge devices or resource-constrained platforms. To address these issues, we build upon the BEVFusion framework and focus on improving the accuracy and efficiency of the perspective transformation process. A BEV visual feature optimization algorithm is proposed, which integrates camera and LiDAR information by embedding LiDAR-provided depth data into the image feature representation, replacing the original depth prediction module. Additionally, the BEV space construction and pooling modules are restructured for computational efficiency. Experimental results show that, without compromising 3D detection accuracy, the proposed method reduces the inference time of key modules to 16% of the original, improves end-to-end inference speed by 83%, and lowers peak memory usage by 27%. It also significantly reduces sensitivity to input image resolution, enhancing adaptability to varying compute resources and improving deployment feasibility in real-world applications.
-
Low-slow small infrared target detection method based on spatio-temporal correlation under complex background interference
Bu Desen, Su Shaojing, Wang Yinglong, Sun Bei, Sun Xiaoyong
Abstract:
To enhance the detection performance of infrared targets for low-altitude, slow-moving, and small (LSS) UAVs under complex background interference, a spatio-temporal correlation-based detection method is proposed. This approach addresses both single-frame static object detection and dynamic trajectory prediction. First, for static detection in single frames, improvements are made to the YOLOv8 algorithm to mitigate the loss of fine-grained information typically caused by downsampling. This is achieved by introducing a no-stride convolutional layer and a P2 detection head, thereby enhancing the capability to detect small targets. Second, for dynamic trajectory prediction, a Kalman filter is employed to estimate and track the UAV's motion trajectory. By integrating this prediction module with the single-frame detector, the system can maintain target localization even when detection confidence drops. Based on confidence evaluation, the system adaptively switches to the trajectory prediction mode to ensure continuous tracking. Temporal correlation is further reinforced by aligning target information across consecutive frames and enabling inter-frame information interaction, effectively establishing spatio-temporal associations. Experimental results show that the improved YOLOv8-P2-SPD model achieves an average precision (mAP@0.5) of 86.8% for single-frame detection. Under challenging backgrounds such as clouds, mountains, and urban structures, the proposed spatio-temporal correlation method improves detection accuracy by 12.1% and recall by 12.2% compared to single-frame detection alone. This approach effectively addresses the limitations of conventional deep learning models in detecting LSS targets under complex background interference and is well-suited for real-world deployment in such scenarios.
-
Retinal vessel segmentation method based on fusion of frequency domain and spatial domain
Song Weiwei, Xu Ming, Yu Xiaosheng, Li Haixing, Wang Hongyu
Abstract:
Accurate segmentation of retinal blood vessels plays a vital role in diagnosing various eye diseases. It not only aids in identifying conditions such as diabetes, arteriosclerosis, and cardiovascular diseases but also enhances clinicians′ ability to diagnose and treat patients effectively. While existing convolutional neural network (CNN) approaches excel at capturing local spatial features through convolutional operations, they face challenges in extracting global spatial information. Conversely, frequency domain methods can capture the overall spectral distribution and global structural features of images but struggle to precisely locate local details and preserve high-frequency information due to spatial information blurring during frequency transformation. To address these limitations, this study proposes a retinal blood vessel segmentation method based on spatial-frequency domain fusion, leveraging the strengths of both domains for local and global feature extraction. The approach features a dual-branch spatial-frequency feature extraction and fusion module in the encoding stage, designed to integrate frequency and spatial features and mitigate detail loss during downsampling. Additionally, a multi-scale Gaussian filter is incorporated to enhance the model′s capability in accurately locating vessel boundaries and preserving continuity of small vessels. Finally, an adaptive spatial-frequency fusion module dynamically calculates fusion weights across feature map regions, improving the precision of small vessel segmentation. Experiments conducted on two widely-used open-source datasets, DRIVE and CHASE_DB1, demonstrate accuracy rates of 96.9% and 97.81%, respectively. Results indicate that the proposed method achieves competitive performance in segmentation accuracy, consistency of small vessel detection, and robustness in handling lesions.
-
Research on the localization and mapping based on closed-loop detection and ground optimization
Wang Yu, Wang Binbin, Wang Fei, Zou Qiang
Abstract:
Accurate estimation of a robot′s posture is fundamental for navigation and path planning during the mapping process. However, the states of yaw, roll and pitch of robots are unobservable and difficult to evaluate and eliminate errors in the outdoor environments. This often leads to serious drift of the Z-axis in the generated map, thereby preventing the construction of a globally consistent and accurate map. To address this issue, we propose a complete robot mapping system architecture, which improves the performance of robot mapping combined with two methods: ground segmentation and closed-loop detection. To efficiently integrate data from multiple sensors and estimate the dynamic biases of the Inertial Measurement Unit in real time, the system employs a LiDAR-inertial odometry simultaneous localization and mapping method. By constructing a factor graph framework, the system incorporates Lidar odometry factors, IMU pre-integration factors, and loop closure detection factors. Through factor graph optimization, the robot′s global pose is estimated to reduce accumulated errors and ultimately generate a globally consistent map. In addition, the algorithm has been deployed on the quadruped robot dog platform, conducting outdoor experiments and using public datasets for extensive evaluation. The experimental results showed that our system has better mapping effects and accuracy, significantly reducing the average positional error compared to baseline methods.
-
End-to-end camera-LiDAR extrinsic calibration method based on stereo camera depth estimation
Abstract:
Accurate and reliable extrinsic calibration of sensors is essential for achieving high-precision localization and navigation in camera-LiDAR fusion systems. However, existing end-to-end camera-LiDAR calibration methods suffer from various limitations, such as large model parameter sizes and mismatched cross-modal feature correlation computation. To address these issues, this article proposes a novel joint calibration method based on stereo camera-estimated depth maps and initial LiDAR-projected depth maps. Specifically, the SGBM algorithm is used to perform stereo matching and generate high-accuracy depth estimation maps. These maps, along with the initial LiDAR depth projections, are fed into a lightweight deep neural network designed for multi-modal feature fusion, effectively mitigating modality inconsistency. A correlation matching layer is then utilized to compute feature-level correspondences, and two separate self-attention mechanisms are introduced to independently model rotational and translational extrinsic. Finally, an iterative refinement training strategy is adopted to enhance calibration accuracy. Compared with the state-of-the-art method LCCNet, experimental results on the KITTI Odometry dataset show that the proposed method achieves an average translation error of 0.67 cm and an average rotation error of 0.09°, representing reductions of 59.64% and 72.73%, respectively. And it requires fewer model parameters. In addition, real-world vehicle tests further demonstrate the effectiveness of the proposed method. When used as the initial extrinsic calibration in the LVI-SAM system, the absolute trajectory root mean square error is reduced by 5.18% compared with LCCNet, validating the accuracy and practical applicability of the method.
-
A dual-level saliency-driven vehicle component detection method
Zhai Yongjie, Wu Zifeng, Zhou Xunqi, Wei Letao, Wang Qianming
Abstract:
High-precision vehicle component detection and segmentation play a vital role in intelligent damage assessment systems by assisting in the accurate localization of damaged parts. However, challenges remain due to complex backgrounds and the performance bottlenecks of traditional detection methods constrained by single-level feature representations. To address these issues, this article proposes a dual-level saliency-driven vehicle component detection method. At the image level, DeepLabV3 is employed with a combination of three loss functions to extract salient foreground regions and suppress background interference. At the feature level, a detection and segmentation framework is formulated based on YOLOv11, where a spatial attention pyramid pooling structure is integrated during feature extraction to enhance multi-scale feature aggregation. Additionally, an attention-guided saliency map module is designed to achieve global modeling and spatial enhancement. To evaluate the effectiveness of the proposed method, a customized vehicle component dataset for multi-part detection is constructed, and extensive experiments are conducted. Ablation studies confirm the contribution of each module. In comparative experiments, the method achieves improvements of 3.5% in detection accuracy and 3.7% in segmentation accuracy over the baseline model. Visualization results further show that the proposed approach focuses more accurately on salient component regions and effectively reduces false detections and missed detections caused by complex backgrounds. Moreover, the method shows strong generalization capability on the public Car Seg dataset, achieving superior performance across multiple evaluation metrics. Overall, the dual-level saliency-driven architecture significantly enhances vehicle component detection performance through salient foreground extraction and attention-guided multi-scale feature aggregation, providing new practical insights for intelligent damage assessment in the vehicle insurance industry.
-
VoxelFSD: voxel-based fully sparse detector with sparse convolution for 3D object detection
Zhou Weidian, Hong Ru, Gai Shaoyan, Da Feipeng
Abstract:
Voxel-based 3D object detection methods often suffer from poor real-time performance when processing large-scale LiDAR point clouds due to their heavy dependence on dense 2D backbone networks. In this paper, we propose VoxelFSD, a voxel-based fully sparse 3D object detector that significantly enhances the real-time capability of long-range detection. The model features three core components: Firstly, parallel convolutional branches (PCB), which expand the receptive field and comprehensively extract object features while mitigating the impact of missing object center features; Then, a sparse region proposal network (SRPN) head that predicts objects sparsely, reducing redundant computations compared to dense prediction and thus improving efficiency for large-scale point clouds; Finally, an ROI head with an attention fusion module (AFM-ROI) that employs cross-attention to effectively fuse 3D backbone features with compressed bird′s eye view (BEV) features in the second stage, refining object representation for improved detection accuracy. By removing the dense 2D backbone from traditional voxel-based detectors and integrating PCB and SRPN, we first present VoxelFSD-S, a fully sparse, single-stage, lightweight detector that achieves a superior balance between speed and accuracy relative to existing lightweight voxel-based models. Building upon VoxelFSD-S, we introduce VoxelFSD-T, a two-stage detector enhanced with AFM-ROI, which boosts accuracy with minimal additional computational cost. On the KITTI test set, VoxelFSD-S and VoxelFSD-T achieve accuracies of 77.67% and 81.50% , respectively.
-
Identification of residential air conditioning loads via channel-attention-optimized variational autoencoder
Wang Lingyun, Tang Tao, Bao Gang, Ruan Shengdong, Zhang Tao
Abstract:
Accurate identification of residential air conditioning load is essential for leveraging their regulation potential and enabling effective demand response. To overcome the limitations of existing residential air conditioning power estimation methods, which often suffer from insufficient accuracy and high computational complexity, this paper proposes a novel non-intrusive neural network model that combines a variational autoencoder (VAE) with an enhanced efficient channel attention (ECA) mechanism. The improved ECA incorporates a dual-pooling strategy-combining global average pooling and global maximum pooling-to capture rich statistical information while highlighting prominent local responses. Additionally, a compression-reconstruction mechanism is introduced: after dimensionality reduction, fast dynamic convolutional kernels adaptively model local channel interactions, focusing on key features and assigning appropriate channel weights. This enhanced ECA module is embedded within the VAE decoder to improve feature reconstruction for air conditioning load estimation. Furthermore, a multi-task learning framework jointly optimizes power disaggregation and state recognition tasks, promoting effective information sharing and complementarity to boost overall identification accuracy. An output module with post-processing state threshold constraints is employed to suppress interference from non-air conditioning loads. The proposed model is validated on real-world residential electricity datasets, showing a mean absolute error (MAE) reduction of 59.71% and 9.22%, respectively, compared to two baseline models across three regions, while achieving an air conditioner state recognition F1 score of 84.58%. Ablation studies reveal that the improved ECA contributes to MAE reductions of 56.23% and 12.47% in two regions, and the multi-task learning framework further improves identification accuracy by 3.17% and 5.90%. Moreover, the minimally intrusive measurement approach-training with intrusive data from only 30% of users-significantly reduces reliance on extensive user data while maintaining high accuracy, demonstrating strong potential for practical deployment.
-
Joint state estimation and trajectory prediction-based GNSS-RTK reliable localization
Zhao Chenhao, Zhang Zhiyong, Fang Xin, Du Xuefei, Tan Rui
Abstract:
Accurate and reliable location information is essential for the safe operation of unmanned transport trains. However, in steel production and transportation environments, multipath interference and signal blockage often produce numerous pseudo-fixed solutions in GNSS-RTK positioning data, leading to unreliable train position estimates and posing serious safety risks. To tackle these challenges, this paper presents a robust GNSS-RTK localization method that integrates state estimation with trajectory prediction. Initially, the double-difference positioning model combined with the least squares algorithm is employed to obtain the GNSS-RTK float solution. This float solution is then fixed to achieve centimeter-level accuracy using an ambiguity decorrelation algorithm and ratio test. To overcome the limitations of the fixed threshold in the ratio test—which can result in pseudo-fixed solutions under multipath interference and signal obstruction—this approach diverges from conventional multi-sensor cross-validation methods. Instead, it leverages the spatiotemporal characteristics of the train′s motion by comparing the current position against the predicted trajectory derived from previous states. This enables rapid identification and correction of pseudo-fixed solutions without requiring additional hardware, thereby enhancing positioning reliability. The method was validated across semi-occluded environments, urban canyons, and steel transportation sites. Experimental results demonstrate its superior ability to detect pseudo-fixed solutions accurately while maintaining high-precision localization. Compared to threshold- and trajectory prediction-based anomaly detection methods, it achieves higher recognition rates, ensuring dependable train positioning in real-world industrial scenarios.
-
Shearer double-IMU position and attitude calculation method
Dai Weiwei, Wang Shibo, Wang Shijia
Abstract:
The inertial measurement unit (IMU) can achieve the advantages of autonomous full-parameter navigation and has the technical advantage of being applied in underground GPS-denied environments. The shearer positioning based on redundant IMUs is a feasible and low-cost positioning method for longwall mining equipment. However, it also faces the problem of a large drift of IMUs over time. When two IMUs are installed on the carrier of a coal mining machine, the differences in their respective output positions and attitudes should all be constants. They should meet the dual-IMU pose constraint conditions. Based on the constraints of position and attitude, a method for calculating the shearer double-IMU position and attitude is proposed based on the information filter. Taking the attitude quaternions of IMU-1 and IMU-2 as state quantities, the information filtering state equation is established based on the quaternion update equation. Using the raw output of the accelerometer, the raw output of the magnetometer, the position difference, and the attitude difference as the measurement values, the Jacobian matrices for converting the measurement values into attitude quaternions are derived, and the measurement equations are constructed. In the experiment of four cutting cycles, for IMU-1, the spherical probability errors (SPEs) of the third and fourth cutting cycles after processing by the algorithm are reduced from 3.618 0 m and 8.220 2 m to 3.618 0 m and 8.220 2 m. The positioning accuracy is improved by 64.9%. For IMU-2, the SPEs of the third and fourth cutting cycles after processing are reduced from 4.342 0 m and 5.736 8 m to 1.617 8 m and 2.352 3 m. The positioning accuracy is improved by 59.0%. The average values of the attitude angles processed by the IMU-1 and IMU-2 algorithms are obtained. The position coordinates of the mobile robot are calculated using the position estimation algorithm, and the SPEs of the third and fourth cuts are 0.790 7 m and 1.431 7 m, respectively. This method provides a low-level solution algorithm for redundant IMU positioning and improves positioning accuracy.
-
Research on measurement method of bridge cable tension with guided wave based on multi-resolution singular value decomposition
Wang Ziye, Duan Shuyu, Sun Lingsi, Wu Xinjun
Abstract:
Cable tension is a critical indicator for evaluating the load-bearing capacity and service life of cable structures. Among various non-destructive testing methods, ultrasonic guided wave-based tension measurement-rooted in the acoustoelastic effect-has shown significant promise. This technique not only enables defect detection but also offers accurate assessment of the cable′s overall stress state. However, in practical applications, the complex waveform characteristics of guided waves-caused by the cable′s structural intricacies pose challenges in extracting precise acoustic time-of-flight feature points. To address this problem, this study proposes a novel cable tension measurement method based on multi-resolution singular value decomposition (MRSVD). The approach constructs a new binary recursive matrix using a dichotomous recursion strategy, integrated with MRSVD. A sliding window technique is employed to segment the signal, enabling extraction of a newly defined echo localization feature-the singular correlation value (SCV). This metric effectively quantifies the correlation between the segmented signal and its corresponding echo. By constructing the SCV spectrum, the method achieves high-precision localization of the echo arrival time at the cable end. To validate the proposed approach, guided wave experiments were conducted on 5~55 parallel steel wire cables with anchor heads under varying tension conditions. The results demonstrate that the method accurately identifies echo signals at the cable anchorage zone, enabling calculation of guided wave velocity and corresponding cable tension. The relative error between the measured and actual cable tension values remains within 10%. A comparative analysis with the conventional cross-correlation method highlights the superior performance of the proposed technique in both measurement accuracy and noise resistance, offering a novel and effective technical solution for cable tension evaluation.
-
Random finite set theory-based method for radar anti-interference target tracking
Jiang Kang, Zhang Zhiyong, Zhang Zhenyuan, Du Xuefei, Tan Rui
Abstract:
In dense vehicle-perception environments, the lack of coordination among vehicular radar transmissions leads to severe mutual interference, resulting in false target generation and failure in tracking actual objects. To overcome these challenges, this paper proposes a radar anti-interference target-tracking method based on the Gaussian Mixture Probability Hypothesis Density (GM-PHD) filter. Initially, time-frequency transformation techniques are employed to extract salient signal features, revealing how interfering signals distort genuine radar echoes. Recognizing that interference intensity causes time-varying detectability of targets, the proposed approach departs from traditional explicit state-measurement association strategies. Instead, both state and measurement sets are modeled within the Random Finite Set (RFS) framework, and an adaptive association weight is introduced to represent their implicit correlation. To further leverage the spatiotemporal distribution characteristics of true and false targets, the GM-PHD filter is utilized to perform multi-target tracking and false target suppression under dynamically changing target counts. Real-world validation is conducted using a TI millimeter-wave radar platform, where an identically tuned radar placed in front of the test vehicle introduces deliberate RF interference. Under these conditions, the radar′s noise floor is significantly elevated and the signal-to-noise ratio drops below -10 dB. Experimental results demonstrate that the proposed method maintains robust and accurate tracking performance in challenging scenarios, including target crossing and occlusion. Compared with benchmark algorithms, it achieves approximately a 50% reduction in tracking error, thereby validating its effectiveness and strong anti-interference capability in practical vehicular environments.
-
Research on thermal management technology for ultra-low temperature inspection robot based on flexible phase change energy storage materials
Wang Guishan, Yang Can, Yu Chengguo, Wang Han, Wang Weidong
Abstract:
Cryogenic inspection robots can perform various tasks in environments that are inaccessible to humans. Extreme cryogenic environments, such as space and deep sea, can cause irreversible damage to most conventional electronic components and mechanical parts. Therefore, energy-efficient thermal management technology is of great significance for cryogenic inspection robots and remains one of the technical challenges in this field. Phase change materials exhibits unique advantages, including high energy storage density and near-isothermal thermal management capabilities, making it highly valuable for thermal regulation. However, the energy storage mechanisms of phase change materials at ultra-low temperatures (-163℃) remain unclear, and the influence of phase change materials and spatial distribution on energy storage performance is not well understood. This study starts from the energy storage mechanism of cryogenic phase change materials. The temperature variations of different phase change materials under -163℃ conditions were examined, and the phase transition processes and energy storage efficiency were quantitatively analyzed. Experimental validation confirmed the accuracy of the analytical results, leading to the optimal selection of phase change materials. Furthermore, simulation studies were conducted to evaluate the energy storage performance of phase change materials at different scales and spatial distributions, followed by an analysis and optimization of their spatial arrangement. Finally, based on the studied phase change energy storage thermal management technology, a cryogenic inspection robot was designed and low-temperature inspection experiments were conducted. The results show that the inspection robot can carry out long-term inspection work at an ultra-low temperature of -163℃, verifying the feasibility of phase change energy storage thermal management technology and providing an effective technical approach for designing cryogenic inspection robots with efficient thermal management capabilities.
-
Deeply optimized integrated learning model EKSSA-CatBoost: Towards highly accurate intelligent diagnosis of PV array faults
Peng Ziran, Xu Huaishun, Xiao Shenping, Pan Changning
Abstract:
Photovoltaic (PA) arrays may be affected by a variety of factors during operation, leading to different types of failures. Real-time monitoring, fault diagnosis, and predictive maintenance of PV array data can be realized through machine learning algorithms, an approach that is not limited by geography and can improve system reliability and efficiency. The current-voltage (I-V) curve of a PV array is an important metric that contains a great deal of information about the health of the PV module, which is crucial for timely fault detection and health assessment. However, existing methods only extract part of the information from the I-V curve for diagnostic analysis, without digging deeper into all the information. As a result, the range of detectable PV array faults remains limited. To address the problems, an I-V curve correction algorithm is proposed to correct the effects of irradiance and temperature on the characterization of the same fault type, effectively eliminating the coupling effect of environmental variables on the characterization of fault features. Then, the CatBoost model is used to realize real-time, high-accuracy fault intelligent diagnosis of PV arrays with small samples. The model′s key hyperparameters are optimized using the sparrow search algorithm. Finally, in order to further enhance the optimization ability of the sparrow search algorithm, the sparrow search algorithm is improved by introducing the fusion elite inverse learning strategy and the Cauchy Gaussian variation strategy, so that it achieves the best effect in optimizing the CatBoost model. The results show that when using simulated data for model training and field data for fault diagnosis, only one and two misdiagnosed samples appear in the test set, respectively. The classification accuracy of the deeply optimized integrated learning model CatBoost reaches 99.9% in both cases, demonstrating its exceptional diagnostic performance.
-
Research on data-driven texture friction modeling and tactile rendering method
Chen Dapeng, Ding Yi, Lou Juncheng, Liu Jia, Song Aiguo
Abstract:
As an important haptic perception dimension of texture, the friction feature has a significant impact on the haptic realism of virtual textures. Previous studies utilize traditional physical friction models to model surface friction. However, such methods are often accompanied by high computational complexity and cumbersome parameter setting. To avoid complex texture modeling processes and predict real-time sliding friction that needs to be fed back to the user when interacting with virtual textures, this study establishes an end-to-end texture friction prediction model (TFPM) based on an encoder decoder that integrates attention mechanisms. This model takes friction data from the previous period and the user′s action information as inputs, which can generate real-time friction signals with high accuracy. It shows a strong generalization effect when dealing with common textures. Subsequently, a haptic device with the function of real-time collection of operation information (pressing pressure and sliding speed) is developed. By combining with the Touch device, data was collected when interacting with 70 real textures, and it was used in conjunction with the SENS3 database to train the model. In order to further verify the generalization ability of the model, a performance evaluation experiment is carried out for the texture samples in the test set. The results show that the model can render the frictional properties of virtual textures with high quality (root mean square error is 0.025 7), and can effectively model the tactile textures outside the database. Finally, the optimal gain parameters of various virtual texture friction signals are determined through psychophysical experiments. Based on this, three user experience experiments are carried out. The experimental results show that the proposed method achieves the highest perceived average similarity score currently (6.25), which can bring users a more realistic virtual texture interaction experience.
-
MBIT* for asymptotically optimal path planning of mobile robot
Chen Zhengsheng, Tian Chukai, Liu Kaixuan, Wang Xuesong, Cheng Yuhu, Chen Yanjie
Abstract:
The multi-batch informed trees (MBIT*) algorithm is proposed for mobile robot path planning. The algorithm integrates multiple informed subset generation and path optimization to reduce path planning time and path length. First, the generalized voronoi diagram (GVD) is utilized to generate a heuristic initial collision-free reference path. Then, a multi-informed subset exploration method is constructed to reduce the sampling range and improve the convergence efficiency based on batch informed trees (BIT*) and the reference path. On this basis, to overcome the problem of uneven distribution of sampling points in existing BIT* algorithm, a biased Gaussian sampling strategy based on the distribution of obstacles and the informed subset is leveraged to obtain the optimal path in narrow environment. The theoretical analyses confirm that the proposed algorithm exhibits probabilistic completeness and asymptotic optimality, while also maintaining measurable computational complexity and storage requirements. Furthermore, MBIT* has been developed as a package to be integrated in the robot operating system (ROS). To further validate its effectiveness, simulation studies and performance comparisons with other prevalent sampling-based path planning algorithms were conducted in typical map scenarios. Furthermore, the real-world experiments under identical environmental conditions were carried out. Results indicate that the proposed algorithm offers obvious advantages in terms of path length and planning time, and is feasible for implementation.
-
A path virtualization and adaptive preview-based method for parking trajectory tracking
Liu Ping, Pan Yewei, Li Yang, Liu Mingjie, Piao Changhao
Abstract:
In traditional parking control, the pure pursuit algorithm has certain limitations due to discontinuities in the path and a fixed look-ahead distance, particularly in terms of tracking accuracy and smoothness. To address the issues of selecting the lookahead distance, poor endpoint performance, and front wheel angle jitter in the pure pursuit method for parking scenarios, a parking trajectory tracking method based on path virtualization and adaptive preview is proposed. Firstly, the geometric relationship model using the pure pursuit method is analyzed. On this basis, optimization strategies for endpoint handling and preprocessing of the parking trajectory path are introduced. These strategies address the issues of oscillation caused by discontinuous curvature in the parking trajectory and jitter resulting from front wheel angle changes near the endpoint by virtually extending and simulating the tracking of the parking trajectory. Furthermore, an adaptive curve preview distance strategy is proposed to reduce the variation amplitude of the front wheel angle during parking, thereby mitigating oscillations and enhancing the tracking accuracy of the parking trajectory. Finally, the implementation steps of the proposed method are presented. The testing and validation are conducted. Compared with the unmodified pure pursuit algorithm, simulation, and real-vehicle test results show that the proposed method exhibits superior tracking performance and endpoint accuracy, effectively reducing the jitter caused by the front wheel angle during parking tracks. For the proposed algorithm, a performance evaluation index matrix, including maximum lateral error, parking endpoint distance error, and cumulative front wheel angle oscillation value and mean difference, indicates average performance improvements of 54.08%, 83.61%, 71.34%, and 48.95%, respectively. These results highlight its effectiveness and practical application value.
传感器技术
Precision Measurement Technology and Instrument
电子测量技术与仪器
Visual inspection and Image Measurement
Information Processing Technology
人机融合与人工智能
About
Organizer：China Association for Science and Technology
Governing Body：China Instrument and Control Society
Chief editorial unitf：Zhang Zhonghua
Address：23rd Floor, Building A, Horizon International Tower,No.6 Zhichun Road, Haidian District,Beijing, China
Zip Code：100088
Phone：010-64004400
Email：cjsi@cis.org.cn
ISSN：11-2179/TH