• Issue 3,2021 Table of Contents
    Select All
    Display Type: |
    • >Visual inspection and Image Measurement
    • System error separation and compensation of the continuous full circle angle standard device

      2021(3):1-9.

      Abstract (503) HTML (0) PDF 9.61 M (1260) Comment (0) Favorites

      Abstract:A continuous angle standard device for the full circle (0° ~ 360°) is developed at the National Institute of Metrology, China. The self-calibration method of equal division average (EDA) with four reading heads is utilized for hardware compensation. The angle measurement accuracy of the standard device after compensation is smaller than 0. 1″. To further improve the angle measurement accuracy of the standard device, the mathematical model of the self-calibration error separation algorithm is derived from the harmonic angle based on the error characteristics, and the error separation method based on cross-calibration and the EDA method is proposed. Secondly, the Fourier series distribution of the original error of full combination separation and the full circle discrete angular position deviation after EDA compensation is analyzed quantitatively. And the error attenuation rate after EDA compensation is analyzed to achieve the optimal EDA compensation function. The inverse Fourier transform method is used to generate the continuous full circle (0° ~ 360°) angle compensation function. Finally, the laser interferometer is utilized to verify the accuracy of the full circle continuous angle standard in the small angle range, which is compared with the laser small angle benchmark of National Institute of Metrology. Experimental results show that the proposed method takes advantage of the high-precision measurement capability of the cross-calibration off-line. The compensated EDA angle measurement accuracy is close to the cross-calibration method within the continuous angle range of the full circle. This method can improve the calibration capability and limitations of self-calibration. The continuous full circle angle standard device is further improved.

    • A location method of building structure information / inertial navigation combination based on the cascade filtering

      2021(3):10-16.

      Abstract (471) HTML (0) PDF 5.30 M (1207) Comment (0) Favorites

      Abstract:The position and heading error cannot be effectively corrected in the shoe mounted indoor pedestrian navigation and positioning system with the inertial sensors as the core. To address this issue, an indoor fusion positioning method based on Kalman particle filter with cascade structure is proposed, which integrates the micro electro mechanical system (MEMS) inertial sensor and indoor building structure information. Firstly, the zero-velocity update is used in the lower Kalman filter to initially correct the inertial navigation solution error. Then, the upper particle filter utilizes indoor building structure information to further calibrate pedestrian position and heading through wall detection. Experimental results show that the building structure information / inertial navigation pedestrian navigation algorithm based on the cascade filter can effectively reduce the accumulation of inertial conductivity error. The “wall-crossing” behavior of pedestrian trajectory caused by the unobservability of course can be corrected when the traditional algorithm adopts zero velocity update and correction system error. This method can reduce the error in the process of location update. Compared with the single inertial pedestrian positioning, the root mean square error of location is reduced from 0. 69 m to 0. 39 m, and the heading of the root mean square error is reduced from 0. 81° to 0. 72°.

    • Dynamic suppression of harmonic distortion for optical encoders via Vold-Kalman filtering

      2021(3):17-24.

      Abstract (411) HTML (0) PDF 7.00 M (1175) Comment (0) Favorites

      Abstract:Harmonic distortion is one of the key factors that affects the measurement accuracy and resolution of optical encoders. However, it is a challenge to dynamically suppress the harmonic distortion when optical encoders operate at variable speed. We explore the application of the Vold-Kalman filter in the signal processing of optical encoders. The basic principle of the Vold-Kalman filter is introduced. The dynamic harmonic suppression model based on the Vold-Kalman filter is formulated. The weight factor of the VoldKalman filter is optimized. The Vold-Kalman filter is developed, which is based on Labview and FPGA. It is applied to signal processing of an optical encoder. Simulation results show that the amplitudes of the dominant third and fifth harmonics are suppressed by about 95% and 98% , respectively. Experimental results indicate that the amplitude of the third harmonic is suppressed by about 71. 3% , and the amplitude of the fifth harmonic is suppressed about 83. 2% . Both simulation and experimental results demonstrate the effectiveness of the Vold-Kalman filter in dynamic harmonic suppression of optical encoders.

    • A digital phase-locking subdivision method for grating Moiré signal

      2021(3):25-34.

      Abstract (547) HTML (0) PDF 4.87 M (1350) Comment (0) Favorites

      Abstract:In order to ensure and improve the accuracy and real-time performance of the signal subdivision results of grating sensor in dynamic measurement process, in this paper, a new digital phase-locking subdivision method for Moiré signal is proposed. This method adopts an open-loop structure, and according to the real-time frequency of the grating Moiré signal, subdivision function of the grating Moiré signal is completed adopting fractional frequency division method. An FPGA-based modified digital phase-locking subdivision circuit was developed, and the key links of the circuit in the process of subdivision were analyzed. The effectiveness of the subdivision algorithm was verified using the homemade subdivision circuit board with the signal generator and actual signal as input, and the modified digital phase-locking subdivision method was compared with the traditional phase-locking subdivision method in subdivision effect. The experiment results show that the digital phase-locking subdivision algorithm designed in this paper can complete the subdivision function in the case of higher Moiré signal frequency and frequency change rate, so it can be better applied to dynamic measurement occasions.

    • Research on fuel dynamic flow measurement technology for aero-engine based on laminar flow meter

      2021(3):35-41.

      Abstract (929) HTML (0) PDF 6.28 M (1196) Comment (0) Favorites

      Abstract:In order to solve the problem of aero-engine fuel measurement in transient state, the dynamic flow measurement technology of aero-fuel flow based on laminar flow meter (LFM) was studied. A liquid LFM with a range of 0~ 3 000 mL/ min was designed. The LFM was tested on a small turbojet-engine test bench and compared with a coriolis mass flowmeter (CMF). The test results show that the LFM has excellent dynamic characteristics. When the engine is in a steady state, the LFM measurement results are in good agreement with those of the CMF, and the relative error is -1. 25% ~ 1. 16% , which is consistent with the relative error of -1. 24% ~ 1. 26% when the LFM is calibrated. When the engine is in a transition state, the liquid LFM can follow the oil pressure after pump well, while the CMF cannot follow the oil pressure after the pump and there is a lag of about 2 s. The measurement result of the LFM can be consistent with the control logic during starting and acceleration / deceleration. The study in this paper provides valuable references for the research of aero-engine test technology and the expansion of LFM applications.

    • >传感器技术
    • AT-cut quartz resonant magnetic field sensor with surface sputtered FeGa film

      2021(3):42-49.

      Abstract (506) HTML (0) PDF 5.90 M (1234) Comment (0) Favorites

      Abstract:Aiming at the problem of low quality factor of the resonant sensor, which will cause high output signal noise and low sensing resolution, a high-quality-factor resonant magnetic field sensor with FeGa film and AT-cut quartz crystal composites is proposed. When excited with ac voltage, the AT-cut quartz wafer will vibrate in thickness-shear mode. The FeGa film is magnetized in the magnetic field, then the film is subjected to the force of the magnetic field, which is transferred to the quartz wafer through interlayer coupling. Due to the force-frequency characteristics of the quartz wafer, its resonant frequency will change. In addition, the proposed sensor has a particularly high quality factor because of the low internal loss and high Q value of the quartz wafer, the stable and low noise sensing output is achieved. Experiment results show that the sensitivity of the resonant magnetic field sensor reaches 0. 2 Hz/ Oe, its quality factor reaches above 46 000 and the power consumption is lower than 8 μW. In addition, the sensor has a wide magnetic field measurement range, and the maximum measured magnetic field reaches 1 200 Oe. The proposed resonant magnetic field sensor has obvious advantages of high signal-to-noise output signal and low power consumption, also has simple structure, and can be fabricated easily.

    • Design and research of a micro-sensor for multi-pollutant detection in hydraulic oil

      2021(3):50-58.

      Abstract (630) HTML (0) PDF 7.12 M (1037) Comment (0) Favorites

      Abstract:A hydraulic oil micro sensor is designed, which is based on the microfluidic chip technology. The microsensor consists of an inductive sensor embedded in silicon planar coils opposing silicon. Metal particles are detected in the inductance detection mode. Water droplets and bubbles are detected in the capacitance mode. Four kinds of pollutants in the oil can be distinguished, which are ferromagnetic particles, non-ferromagnetic particles, water droplets and bubbles. Experiments verify that the new structure of the micro sensor improves the detection accuracy of metal particles in the inductive mode. It can detect 28 μm iron particles and 85 μm copper particles in the oil. In the capacitance mode, the micro sensor has the best detection effectiveness when the excitation voltage is 2 V and the excitation frequency is 0. 9 MHz. In addition, it can detect 95 μm water droplets and 160 μm bubbles in the oil. This research is great significant for failure prediction and diagnosis of hydraulic systems.

    • Distributed acoustic sensing type special optical fiber structure design

      2021(3):59-69.

      Abstract (689) HTML (0) PDF 9.94 M (1518) Comment (0) Favorites

      Abstract:The distributed optical fiber acoustic wave sensing (DAS) is a new kind of optical fiber acoustic sensing technology. Its mechanism is that the acoustic signal generates a weak deformation on the distributed optical fiber, which causes the change of the refractive index of the optical fiber. Light produces coherent effects during transmission. DAS is not sensitive to quadrature component signals. It is difficult to detect such signals with the straight horizontal optical fiber. In this article, the sensitivity of DAS detection acoustic signal detection is mainly studied. The ANSYS simulation software is utilized to analyze the sound field′s ability to perceive acoustic signals of different physical models. The best solution of the physical model structure is analyzed, a physical model with high sound perception ability is formulated. Experimental results show that the signal amplitude is increased by 3. 94% on average. After signal demodulation and denoising processing, it is concluded that the designed special optical fiber high-sensing structure can effectively achieve better acoustic signal acquisition quality.

    • TDLAS online measurement system for plume temperature of ramjet engine

      2021(3):70-77.

      Abstract (522) HTML (0) PDF 6.97 M (1097) Comment (0) Favorites

      Abstract:In order to realize the online measurement of hydrocarbon fuel ramjet engine plume temperature, the spectral line pair of (7 444. 352+7 444. 371) cm -1 ~ (7 185. 586 5+7 185. 597 3) cm -1 was optimally selected through molecular spectral simulation taking H2O molecular as the target composition, and tunable diode laser absorption spectroscopy (TDLAS) system was designed to measure the parameters of hydrocarbon fuel ramjet engine plume using scanned-wavelength direct absorption spectroscopy-time division multiplexing (SDAS-TDM) strategy. Moreover, the temperature measurement accuracy of this system was verified by premixed flat-flamed burner. The results show that the relative deviations between the measured flame temperature at the height of 1cm of the premixed flat-flamed burner using this system and the standard reference value are all within 15% . On this basis, the system was applied to measure the parameters of the hydrocarbon fuel ramjet engine plume. Through alternately modulating of 1 392 nm and 1 343 nm diode lasers, the intensity attenuation signal of the modulated laser passing through the measuring zone was measured, the H2O molecule absorption spectrums at (7 444. 352 + 7 444. 371) cm -1 and ( 7 185. 586 5 + 7 185. 597 3) cm -1 lines were obtained. The online temperature measurement of the hydrocarbon fuel ramjet engine plume was realized with two-line technique, which provides important references for the combustion organization and performance evaluation of hydrocarbon fuel ramjet engine.

    • Development of a new type of high temperature thin film heat flux sensor

      2021(3):78-87.

      Abstract (296) HTML (0) PDF 9.02 M (1682) Comment (0) Favorites

      Abstract:When the spacecraft reenters the earth atmosphere, the outer surface of spacecraft will generate megawatt of heat flux. As a result, the instantaneous temperature rise can reach 1800℃ . In order to ensure the stability and safety of the spacecraft, it is significantly important to accurately measure the heat flux on the thermal protection system surface of the spacecraft in real time. Aiming at the technical difficulty of heat flux measurement in high-temperature condition, a novel heat flux sensor structure integrating the lead and substrate is proposed. Combining the ceramic sintering static pressing molding and magnetron sputtering technologies, a new high temperature thin film heat flux sensor was developed by sequentially depositing PtRh30-PtRh6 thermopile thin film, Al 2O3 film and ZrO2 film on the substrate surface of 99 alumina ceramics where the PtRh6 leads are embedded. Then, the static and dynamic performance, high-temperature resistant and repeatability tests were conducted. The result shows that the sensitivity of the sensor reaches better than 0. 01 μV/ (W/ m 2 ), the dynamic response time is 3. 97 s. High temperature test at 1200℃ was conducted on the sensor, the test result shows that the output signal of the sensor has no obvious change before and after the high temperature test, and the maximum repeatability error is 2. 38% . The developed high temperature thin film heat flux sensor can provide scientific basis for the high temperature heat flux measurement and thermal protection system optimization.

    • Planar two-dimensional electric field time-grating displacement sensor with discrete array

      2021(3):88-96.

      Abstract (157) HTML (0) PDF 11.50 M (1005) Comment (0) Favorites

      Abstract:The planar two-dimensional grating displacement measurement technology is difficult to achieve balance between high precision and large range. To address this issue, a two-dimensional electric field time grating displacement sensor with a discrete array structure is designed, which is based on the research foundation of the one-dimensional electric field time grating in the early stage. A coding method of planar orthogonal discrete grating surface spatial arrangement is utilized in the sensor to realize the coding of planar two-dimensional electric field time grid excitation electrodes. A planar two-dimensional electric field time grating displacement measurement model is formulated, and the coupling signal expression modulated by the displacement information in the X and Y directions is theoretically deduced. A method for direct decoupling of two-dimensional displacement measurement signals is proposed, which uses the spatial position relationship of differential sensing electrodes to realize the decoupling of measurement signals through simple addition and subtraction operations. The sensor prototype is established by using PCB technology and performance tests are implemented to evaluate the feasibility of the proposed coding and decoupling method. The final result shows that the sensor proposed in this paper has the peak - to - peak measurement errors of 13. 1 μm and 11. 8 μm in X and Y directions respectively in the measurement range of 160 mm×160 mm.

    • >智能系统与人工智能
    • Research on variable stiffness drive method based on magnetorheological principle

      2021(3):97-104.

      Abstract (742) HTML (0) PDF 6.75 M (1375) Comment (0) Favorites

      Abstract:Aiming at the variable stiffness drive function requirement of robot joint, a variable stiffness drive method based on magnetorheological principle is proposed. The method utilizes the characteristics that the rotary magnetorheological damper torque is controllable and the response speed is fast, and the rotary magnetorheological damper and servo motor are combined to form a variable stiffness driver. Based on the configuration of the variable stiffness driver, the principle of variable stiffness drive was analysed. The design of the rotary magnetorheological damper used for variable stiffness driver was completed, the torque value and the magnetomotive force value of the damper were obtained. Based on the configuration of the variable stiffness driver and the mechanical characteristics of the damper, the variable stiffness control process was designed and the control program was developed. On this basis, the test system of rotary magnetorheological damper and its mechanical performance was developed, the actual mechanical performance parameters of the damper were obtained, the maximum torque of damper was 18. 1 N·m. The variable stiffness driver test system was established and the experiment research was completed, when the stiffness coefficient was set as 5. 0 N·m/ rad,6. 0 N·m/ rad,7. 0 N·m/ rad respectively, the actual stiffness coefficient obtained was 5. 1 N·m/ rad,6. 2 N·m/ rad,7. 1 N·m/ rad corresponding. The experiment results show that the proposed method can realize the stiffness adjustment of the driver nicely.

    • Navigation system of security mobile robot based on FM millimeter wave

      2021(3):105-113.

      Abstract (521) HTML (0) PDF 7.10 M (1146) Comment (0) Favorites

      Abstract:As the application field of security robots are expanded, the complexity of their working environment increases. In special environments such as smoke, dust and dark, the navigation system with visual and laser navigation manners are no longer applicable. Aiming at this problem, based on the research on the principle of the distance measurement with millimeter-wave radar, firstly, the research of two pulse canceller is carried out to filter out the static clutter. Then, a dynamic threshold detector is designed to accurately obtain the distance between the millimeter-wave radar and mobile robot. In order to improve the navigation accuracy, a segmentation clustering method is proposed to process the distance data set. Finally, a security robot navigation system based on the principle of triangulation was designed. Experiment results show that using the segmentation clustering method, the robot navigation accuracy is higher compared with that using the mean method. The robot can operate along the set straight and curvilinear lines in smoke and dark environment. The navigation error is about 0. 11 m.

    • Research on magnetic field drive modeling of micro robot based on gradient intensification

      2021(3):114-122.

      Abstract (224) HTML (0) PDF 13.24 M (1139) Comment (0) Favorites

      Abstract:Magnetic field drive technology has become current research hotspot in the field of micro-robot operation. A gradient-enhanced magnetic field drive system is designed. Firstly, the iron core end model was established, the end shape was designed, and finite element method was used to optimize the iron core coil parameters to meet the system index requirements, such as magnetic induction intensity, magnetic field gradient, magnetic field uniformity and working space. Secondly, the designed magnetic drive system was simulated with ANSYS and tested in experiment. It is concluded that the maximum magnetic induction intensity at the working space center of the magnetic drive system is 73. 93 mT, the maximum magnetic field gradient at the working space center is 8. 68 T/ m. Compared with the gradient magnetic field drive system studied in other literatures, the performance has improved significantly. At the same time, the motion control experiment analysis on the designed new magnetic drive system was conducted. The results show that the system can perform closed-loop position control of the magnetic beads under different environments. The average error for driving the magnetic beads moving along the predetermined trajectories under the silicone oil environment with different viscosity is 0. 066 mm at most, and the maximum root mean square error is 0. 078 mm.

    • Operator recognition and adaptive speed control method of teleoperation robot based on CNN-GRU

      2021(3):123-131.

      Abstract (448) HTML (0) PDF 6.88 M (1279) Comment (0) Favorites

      Abstract:The movement speed of the slave manipulator arm in traditional space teleoperation system completely depends on the operating speed of the operator. In order to improve the safety of the space teleoperation system, an adaptive speed control method based on the recognition of the operating speed of the operator is proposed. Combining with the theory of deep learning, a fusion model based on convolutional neural network (CNN) and gate recurrent unit (GRU) neural network is proposed to identify and classify the speed of operator. Nine subjects were selected to construct an operator speed sample library. The operating speed of the operators is divided into three categories, and the final recognition accuracy rate reaches 92. 71% . And, on this basis, the cascade PID is used to realize the adaptive speed control of the slave manipulator arm. Experiments confirm that the model can also accurately identify new operators. At the same time, the accuracy of the model is better than that of the fusion model of convolutional neural network and recurrent neural network (RNN), and the real-time performance of the model is better than that of the fusion model of convolutional neural network and long short-term memory (LSTM) neural network. Besides, the adaptive speed control based on this model can reduce the end linear speed of the manipulator arm while ensuring that the movement trajectory of the slave manipulator arm remains unchanged, which helps to improve the safety of the space teleoperation system.

    • Research on robot random obstacle avoidance method based on fusion of improved A ∗ algorithm and dynamic window method

      2021(3):132-140.

      Abstract (665) HTML (0) PDF 11.80 M (1808) Comment (0) Favorites

      Abstract:Aiming at the problems of collision or failure of path planning when the robot uses the A ∗ algorithm to plan a path in the environment with random obstacles, a random obstacle avoidance method for robots that combines the improved A ∗ algorithm with the dynamic window method is proposed. In the improved A ∗ algorithm, firstly, the search point selection strategy and the evaluation function are optimized to improve the search efficiency of the A ∗ algorithm, then the redundant point deletion strategy is proposed to eliminate the redundant nodes in the path, and the dynamic window method is used for the local planning between every two adjacent nodes to ensure that on the basis of the global optimal path, random obstacle avoidance is achieved in real time, so that the robot can reach the target point successfully. The experiment results show that the improved A ∗ algorithm proposed in this paper can reduce the path length by 4. 39% and the calculation time by 65. 56% on average compared with the traditional A ∗ algorithm. After fusing the dynamic window method, on the global path basis the local path can be modified to achieve random obstacle avoidance, which verifies the effectiveness of the proposed algorithm.

    • >Detection Technology
    • An arc fault detection method based on the self-normalized convolutional neural network

      2021(3):141-149.

      Abstract (526) HTML (0) PDF 9.41 M (1554) Comment (0) Favorites

      Abstract:The electric arc fault is an important cause of electrical fire. When the series arc fault occurs in the low-voltage circuit, the traditional feature extraction method cannot fully express all the data features of the time domain signal. The feature expression ability of arc fault is limited, which may bring high false alarm rate and miss alarm rate of detection results. To solve this problem, an arc fault detection method based on the self-normalized convolutional neural network is proposed. This method intercepts the current time series of different kinds of loads according to half period. Then, they are normalized. The two-dimensional images of arc faults and normal operation are generated by the grayscale data transformation. The gray transformation features of arc faults are extracted by using the convolutional neural network. The arc fault convolution features are identified by multi-layer full connection layer fitting calculation of the following sampling information. The evaluation shows that the accuracy of the proposed method is 99. 67% , which is better than the traditional convolutional neural network and has good generalization performance.

    • Study on detection method of weak series DC fault arc in PV power generation systems

      2021(3):150-160.

      Abstract (1006) HTML (0) PDF 7.89 M (1420) Comment (0) Favorites

      Abstract:Fault arc in photovoltaic power generation system is difficult to accurately detect due to strong randomness, weak signal, and easy to be affected by load sudden change. According to the U-I output characteristics of photovoltaic cells, this paper analyzes the generation mechanism of DC series weak fault arc in photovoltaic power generation system, and analyzes the characteristics of weak DC series fault arc signal by building a photovoltaic power generation system fault arc simulation experimental platform, and then a method to detect weak DC series fault arc based on the wavelet energy entropy features of current signal is proposed. The proposed method firstly calculates the current pulse factor, which are used to detect the fault arc with threshold comparison method. On this basis, the current wavelet energy entropy features are calculated to identify weak fault arc based on extreme learning machine (ELM). The experimental results show that the proposed method can not only detect strong DC fault arc, but also detect weak DC fault arc with high average identification rate 98% .

    • Anomaly detection of cigarette-smuggling vehicles based on label propagation

      2021(3):161-167.

      Abstract (326) HTML (0) PDF 1.68 M (1086) Comment (0) Favorites

      Abstract:The tobacco industry is closely related to government revenue. Smuggling of counterfeit cigarettes will not only cause the loss of national tax, but also disrupt the market and endanger consumers′ health. How to effectively regulate cigarette-smuggling vehicles is of great significance to the development of the tobacco industry. Aiming at the issue of cigarette-smuggling vehicles, this paper combines the actual collected vehicle data and proposes an anomaly detection algorithm based on label propagation. Firstly, the features of the vehicle data set are extracted. Second, random forest algorithm is adopted to conduct the feature selection. On this basis, label propagation algorithm is utilized to classify the anomaly vehicles. The results show that in the case of less historical data and abnormal vehicle tags, the anomaly detection algorithm of cigarette-smuggling vehicles based on label propagation can effectively detect most cigarette-smuggling vehicles. In the given dataset, the recall rate of the proposed algorithm in detecting anomaly is 57. 7% , which outperforms those of other traditional machine learning models. The algorithm can provide auxiliary support for the detection of the vehicles transporting prohibited items.

    • Multi-object tracking algorithm for UAV based on the thin plate spline function

      2021(3):168-176.

      Abstract (411) HTML (0) PDF 12.96 M (993) Comment (0) Favorites

      Abstract:The multi-object using Unmanned aerial vehicle, UAV monocular camera under object may have problems of position drift and failure of state prediction. To address these issues, a multi-object tracking method for UAV based on the thin plate spline function is proposed, the UAV motion is formulated by the space transformation function. The state space model for UAV motion is established, and the tracklets and detection correspondence are initialized by appearance characteristics. The unknown parameters of the model are obtained by calculating the least square solution of the thin plate spline function based on the initial correspondence. Then, the tracklets motion state is predicted according to the model. And the appearance to data association is combined. In addition, the space transformation parameters are introduced into the Kalman filter equation to realize the optimal estimation of tracklets state under camera motion. The process of tracklets initialization and termination, missed detection and false detection is realized by the effective tracklets management method. Experimental results on UAV data set show that the proposed algorithm has better performance than the existing state-of-the-art algorithms. Compared with the existing mainstream algorithm MDP, the multi-object tracking accuracy of the proposed algorithm is increased by 2. 75% .

    • Research on an electronic differential control strategy based on active disturbance rejection control

      2021(3):177-191.

      Abstract (108) HTML (0) PDF 16.43 M (852) Comment (0) Favorites

      Abstract:When the electric vehicle is steering, the driving wheel will bear more disturbing load under the combined actions of complex road conditions and vehicle conditions, and the proportion uncertainty of the sliding motion of the driving wheel increases, which will affect the driving stability and safety. Therefore, the active disturbance rejection control (ADRC) based electronic differential control (EDC) strategy is designed, and the Chaos Particle Swarm Optimization (CPSO) algorithm is used to design the controller parameters. A seven degree of freedom complete automobile model is constructed, and the electronic differential controller based on CPSO-ADRC is designed taking slip rate as the control quantity and driving wheel motor torque as the output, so that the slip rate is always kept at the target value in the steering process. The proposed EDC system is compared with the EDC systems equipped with fuzzy PID controller and sliding mode controller and analyzed, and the EDC experiments under different road conditions were carried out on Simulink / CarSim platform and real vehicles. The results show that the electronic differential control strategy based on CPSO-ADRC has strong antiinterference ability, its speediness is increased by 20% and 14. 4% , respectively compared with the other two strategies, the amplitude of yaw rate is reduced by about 50% , the speediness and robustness of EDC are enhanced, and the driving safety in electric vehicle steering process is more effectively guaranteed.

    • >Information Processing Technology
    • Fast recognition and elimination of the interference signal caused by bending deformation of the whole roller seamless flatness meter

      2021(3):192-200.

      Abstract (177) HTML (0) PDF 4.40 M (999) Comment (0) Favorites

      Abstract:The whole roller flatness meter is the development trend of the contact type flatness meter for cold rolling. Under the action of dead-weight and external load, the bending deformation of the detection roller will change the pre-tightening force of the sensors that are installed in the meter. The additional signals are generated, which are similar with sinusoidal waveforms. They will affect the accuracy of flatness detection by superposition with the effective signals. This paper studies the generation mechanism. The recognition and elimination method of this interference signal is also studied. Firstly, according to the experimental calibration and industrial testing results, the wave characteristics of the flatness signal and the interference signal are found and analyzed. The effective flatness waveform distributed along the wrapping angle is located on the peak and trough of the interference waveform distributed along the circumference. Secondly, the elastic theory is applied to calculate and analyze the stress and deformation caused by the beding of the flatness roller. It reveals that the periodic change of the pre-tightening force of the sensor is the cause of the interference sinusoidal waveform. Finally, based on the detection signal data outside the wrapping angle range, a mathematical model that has accurate and quick feature to identify and eliminate the interference signal waveform is formulated, which utilizes the curve fitting optimization method of minimum error. The industrial application show that the proposed method can effectively eliminate the beding signal and improve the effect of flatness detection and control by 1~ 2 I.

    • A fault diagnosis method of rolling bearing based on the improved DQN network

      2021(3):201-212.

      Abstract (408) HTML (0) PDF 13.54 M (1196) Comment (0) Favorites

      Abstract:Under normal and fault states in practice, rolling bearing vibration data are imbalanced and the fault diagnosis accuracy is low. Based on the deep reinforcement learning, an improved deep Q network (DQN) fault diagnosis method for rolling bearing is proposed. The short time Fourier transform is performed on the vibration data to establish sample sets of time-frequency graph. The distance between the sample and the center point in the K-means algorithm is used as the bias of the return value. The imbalance ratio is utilized as the benchmark to formulate a personalized reward function for the training set. Meanwhile, the residual network (Resnet-18) is used to realize the deep extraction of features. In which, the agent takes the new reward function and time-frequency graph as input. The diagnosis action is executed at each time step. And the reward is judged and returned. Finally, the agent learns the fault diagnosis strategy under imbalanced data. Compared with other methods, experimental results show that the improved diagnostic model is improved by 5% to 8% under imbalanced conditions. At the same time, it also performs outstandingly under imbalanced and variable load conditions. The imbalanced index score can reach about 0. 982, which shows better generalization.

    • Energy maximization analog-to-information converter system for ECG signal

      2021(3):213-220.

      Abstract (618) HTML (0) PDF 6.60 M (999) Comment (0) Favorites

      Abstract:The current portable ECG acquisition system needs analog to digital conversion module, which has low power consumption and high resolution. Although the analog to information converter (AIC) based on pulse width modulation can effectively reduce the sampling rate of the system, the quantization part of the conversion clock is proportional to the quantization accuracy. Therefore, there is a problem of high power consumption. In this study, a precision adjustable time to digital conversion (TDC) module based on the power entropy is proposed, which is based on the energy imbalance of ECG signal. The energy maximization is taken as the basic principle of design. The minimum quantization accuracy of system observation vector is determined by analyzing the power spectrum entropy of ECG signal. In this way, the optimal design of AIC time coding system is achieved. Experimental results show that this method can reduce 80% TDC clock dynamic reversal while sampling ECG signals under the conditions of compression ratio of 4, SNR of 38. 91 dB and reconstruction accuracy of 0. 36% . Therefore, the power consumption can be reduced effectively.

    • Emotion EEG recognition based on the adaptive optimized spatial-frequency differential entropy

      2021(3):221-230.

      Abstract (332) HTML (0) PDF 14.53 M (936) Comment (0) Favorites

      Abstract:The affective brain-computer interface (ABCI) aims to provide a channel for emotional communication between people and external devices. Emotion electroencephalography (EEG) recognition is the most basic and key part of the ABCI system. To adaptively select the optimal combination of spatial electrodes and frequency bands to optimize the emotion EEG feature and improve the classification effectiveness, an adaptive optimized spatial-frequency differential entropy (AOSFDE) feature is proposed. We design an importance measurement method of spatial electrodes based on the relative entropy. According to the importance of various spatial electrodes, the most important spatial electrodes are selected automatically. The sparse regression algorithm is used to optimize the differential entropy features in multiple local spatial-frequency domains. The emotion EEG database (SEED) provided by Shanghai Jiao Tong University is utilized for experimental analysis. Results show that the proposed AOSFDE method can effectively improve the recognition accuracy. For 15 subjects in this dataset, the average recognition accuracy values of positive / negative, positive / neutral and neutral/ negative binary emotional classifications are 91. 8% , 93. 3% and 85. 1% , respectively. The proposed algorithm provides a new idea and method for emotion EEG recognition.

    • Imbalanced classification for epileptic EEG signals based on deep learning

      2021(3):231-240.

      Abstract (523) HTML (0) PDF 13.90 M (1460) Comment (0) Favorites

      Abstract:Automatic seizure detection is of great significance to the diagnosis and treatment of patients with epilepsy. Due to the short duration of epileptic seizure period, the EEG signal distribution between the seizure period and the non-seizure period is imbalanced. To solve this problem, an automatic detection method of epilepsy based on the fusion of imbalanced classification and deep learning is proposed. Firstly, the Borderline-SMOTE algorithm is applied to one-third training set to prevent the boundaries between different classes from blurring. Then, a pyramidal one-dimensional convolutional neural network is designed, which is trained with the balanced processing data. Different from the common 2D convolutional neural network, the 1D convolutional neural network reduces the number of training parameters. The training rate is improved, and the overfitting is avoided effectively which is caused by the small number of training samples. By utilizing the 991 hours long scalp EEG database, the effectiveness of the seizure detection after balanced treatment is significantly improved. The sensitivity, specificity, positive predictive value, and negative predictive value reach 92. 35% , 99. 88% , 90. 68% , and 99. 91% , respectively. Meanwhile, the comparison with other seizure detection methods shows that the proposed method has better performance. It is suitable for satisfying requirements of clinical application.

    • >Visual inspection and Image Measurement
    • Research on multi-task convolutional neural network facing to AR-HUD

      2021(3):241-250.

      Abstract (585) HTML (0) PDF 16.26 M (841) Comment (0) Favorites

      Abstract:AR-HUD has been widely used in automobile. Its environment perception module needs to complete target detection, lane segmentation and other tasks, but multiple deep neural networks running at the same time will consume too much computing resources. In order to solve this problem, a lightweight multi-task convolutional neural network ( DYPNet) applied in AR-HUD environment perception is proposed in this paper. DYPNet is based on YOLOv3-tiny framework, and fuses the pyramid pooling model, DenseNet dense connection structure and CSPNet network model, which greatly reduces the computing resources consumption without reducing the accuracy. Aiming at the problem that the neural network is difficult to train, a linear weighted sum loss function based on dynamic loss weight is proposed, which makes the loss of the sub-networks tend to decline and converge synchronously. After training and testing on the open data set BDD100K, the results show that the detection mAP and segmentation mIOU of the neural network are 30% and 77. 14% , respectively, and after accelerating with TensorRt, it can reach about 15 FPS on Jetson TX2, which has met the application requirements of AR-HUD. It has been successfully applied to the vehicle AR-HUD.

    • Manufacturing error detection method based on ICT image and design model

      2021(3):251-261.

      Abstract (388) HTML (0) PDF 7.22 M (1020) Comment (0) Favorites

      Abstract:When the two-dimensional CT image and the design model are used to analyze the manufacturing error, it requires the position of CT image in the design model. To determine the position, it needs to find a reference position matching the design model on the workpiece. Then, the machining the position needs to be manufactured, which increases the processing cycle and manufacturing cost of the workpiece. To address this issue, this study proposes an automatic search CT image design model to complete the workpiece manufacturing error detection method. Firstly, the edge of CT image is extracted. Secondly, the point cloud design model is sliced layer by layer. Then, through Hu moment matching, ICP registration automatically searches the point cloud slices that match the CT image. Finally, according to the registration results, the error distribution of the workpiece is calculated, and the manufacturing error detection of the workpiece is completed. To evaluate the feasibility of this method, the standard workpiece with known size is used for detection. Experimental results show that the detection error is concentrated between 0~ 0. 25 pixel, which has high detection accuracy. Two actual workpieces are tested to evaluate the applicability of the method. In this study, without the need to determine the benchmark position of the workpiece, the two-dimensional CT image and design model of the workpiece are used to realize the calculation, analysis and display of the manufacturing error of the workpiece, which has high practical significance.

    • Collision analysis of non-convex complex shape objects based on bincular vision

      2021(3):262-269.

      Abstract (155) HTML (0) PDF 9.89 M (1002) Comment (0) Favorites

      Abstract:The traditional collision detection algorithm based on the point cloud typically uses bounding volume hierarchies or space decomposition to determine whether there is collision. This method cannot achieve the accurate safety distance value between the object and the scene. In this study, a collision analysis method based on the binocular stereo point cloud is proposed, which is mainly for nonconvex complex surface objects. Firstly, the binocular stereo algorithm is used to reconstruct the point cloud of the scene captured by the calibrated camera. Then, the point clouds of the object and the scene are both utilized to solve the collision problem. The process distance values are rapidly obtained by the K-D tree search algorithm. The symbol is defined by the coordinate relationship of the point cloud along the optical axis of the camera. The accuracy of this method in the field experiment of a detector is 100% . Compared with the existing algorithms, this method can obtain the distance of each point on the surface of the object. The complexity of calculation is reduced efficiently under the reference of camera′s optical axis. The detection time at the drop sampling single point is not larger than 0. 15 s, which can satisfy the need of rapid collision analysis between non-convex objects and complex topography. This method can successfully complete the sampler-terrain collision analysis task of sampling point selection in the lunar surface sampling package of Chang′e-5.

Current Issue


Volume , No.

Table of Contents

Archive

Volume

Issue

Most Read

Most Cited

Most Downloaded