Qr Code
-
WeChat
- Current Discuss
- First Published
- Album Paper
-
Research progress on defect detection of lithium battery based on machine vision
Abstract:
Lithium battery is one of the core components of new energy vehicles. But, the complex manufacturing process of lithium battery inevitably introduces various defects, which seriously affects the quality of products. Therefore, defect detection has become an important part of lithium battery manufacturing process. The machine vision method takes into account the advantages of accuracy and speed, which has been paid much attention to. In this article, the research progress of defect detection methods for lithium battery based on machine vision in recent 15 years is reviewed. Firstly, the common surface defect types of lithium battery are introduced, and the main flow of visual defect detection is clarified. Next, the defect detection method of lithium battery based on traditional image processing is emphasized. The four steps, including image preprocessing, image segmentation, feature extraction and classification recognition, are explained in detail. The advantages and disadvantages of each step are compared. Then, the defect detection methods based on deep learning are summarized according to the classification network, detection network and segmentation network. Afterwards, 10 self-built datasets of lithium battery and performance evaluation index of defect detection are sorted out. Finally, it is pointed out that the defect detection of lithium battery is faced with many technical challenges, and the future work is prospected.
-
Multispectral image feature fusion method for detecting surface defects in IC devices
Huang Zhihai, Deng Yaohua, Wu Guangdong
Abstract:
To address the issue where minor surface defects of IC devices are often obscured by redundant noise in traditional pixel-level fusion detection—hindering defect feature extraction—and the challenge of adaptively adjusting the contribution of visible and infrared images in complex detection scenarios with unstable lighting, this paper proposes a surface defect detection method for IC devices based on multispectral image feature fusion. The method employs a mid-fusion strategy to design a Multispectral Image Feature Fusion (MIFF) module and establishes a dual-path feature extraction channel within the YOLO framework. This leads to the development of an end-to-end YOLO-MIFF defect detection model specifically for multispectral image feature fusion. Experimental results demonstrate that the YOLO-MIFF fusion detection model achieves a mean Average Precision (mAP) that is 24.69% and 35.65% higher than that of single visible and single infrared image detection, respectively. Additionally, compared to the YOLO-Multiply, YOLO-Concat, and YOLO-Add models, YOLO-MIFF improves detection accuracy by 9.85%, 6.67%, and 3.44%, respectively.
-
Visual SLAM algorithm based on graph neural network feature point matching
Ji Zeyuan, Yu Xiaoying, Fu Wenxing
Abstract:
The vision-based simultaneous localization and mapping (SLAM) technology has significant applications in industries, such as augmented reality and autonomous driving. However, traditional visual SLAM faces challenges such as low positioning accuracy or failure in low-light conditions. This article proposes a visual SLAM algorithm based on graph neural network (GNN) for matching feature points between consecutive frames, e.g., VINS-GNN. In the front end of the visual SLAM, a feature point matching and tracking strategy is designed, integrating GNN with visual SLAM, which could effectively enhance the performance of feature point tracking. In the back end, a loop closure algorithm based on multi-frame fusion is designed to further improve global positioning accuracy. Comparative experiments on public datasets with low light and low texture show that VINS-GNN improves positioning accuracy by 17.33% compared to VINS-Fusion. In real indoor low-light experiments, VINS-GNN significantly improves the accuracy at the end of the trajectory compared to VINS-Fusion. Additionally, the article introduces neural network inference acceleration techniques to reduce resource consumption and enhance real-time performance. Experimental results show that the strategies proposed by VINS-GNN significantly enhance positioning accuracy under indoor low-light conditions, which is of great significance for the development of indoor pedestrian and mobile robot positioning technology.
-
Coupled artifacts removal in cone-beam computed tomography images based on multi-scale generative adversarial network
Chai Shijie, Huang Kuidong, Yang Fuqiang, Zhao Julong
Abstract:
To address the issue of incomplete correction of coupled artifacts in cone-beam computed tomography (CBCT) images, a coupled artifact correction method for CBCT images based on a multi-scale generative adversarial network (GAN) is proposed. Firstly, a CBCT coupling artifact dataset comprising both simulated and real images was constructed based on the artifact characteristics of CT images to enhance the model′s generalization capability. Additionally, the generator structure of the network was improved by integrating the feature pyramid network (FPN) and convolutional block attention module (CBAM) to capture more comprehensive feature information. We also employed a multi-scale discriminator (MSD) alongside these components to from a generative adversarial network framework, producing clearer and more realistic artifact-free images. Experimental analysis showed that the PSNR and SSIM of the corrected images increased by this method increased by 21.595 dB and 0.541 in the simulated dataset, and by 14.072 dB and 0.274 in the real dataset. The experimental results indicate that the proposed method can effectively correct coupled artifacts.
-
An intelligent detection method for cold-rolled strip deviation based on machine vision
Duan Bowei, Wang Dongcheng, Xu Yanghuan, Xing Jiawen, Liu Hongmin
Abstract:
The lack of measured deviation of the strip during cold-rolled process may cause problems of reduction in flatness the control accuracy and strip breakage. An intelligent detection method for cold rolled strip deviation based on machine vision is proposed in this article. A lightweight network structure is constructed, which is based on classic network UNet for intelligent segmentation of cold-rolled strips. The MobileNetV2 is used to replace the original contraction path of UNet and the channel attention ECA_Module is embedded in the connection structure, which effectively reduces the amount of network parameters while enhancing the perception ability of target features. The strip region segmentation model of the cold-rolled strip (SRS_M) can be obtained by training this network. The accuracy indicators mIoU and mPA of SRS_M could reach 98.83% and 99.36%, respectively. The running time of a single image is 40.57 ms. An intelligent detection model of strip deviation is formulated by combining SRS_M and edge position extraction algorithm. The 1 503 deviation sample data are collected through an edge detection device installed on site, which is used to evaluate the proposed method. The results show that the absolute error of 92.82% of the samples is within the range of ±2 mm, and the absolute error of all samples is within the range of ±3.5 mm, which shows the effectiveness of the proposed method in this article.
-
Research on visual detection method of levitation gap-based on column-oriented semantic segmentation
Jing Yongzhi, Ni Sheng, Jia Xingke, Liu Zhixin, Liu Guoqing
Abstract:
To address the problems of a large number of parameters, low real-time performance, and poor anti-interference ability of traditional semantic segmentation networks, a visual detection method of suspension gap based on column-oriented semantic segmentation is proposed. The method defines gap detection as finding the set of disaggregated locations of gaps in the middle of the image and simplifies the classification problem to reduce the computational complexity. Firstly, the structure of the visual gap detection-based suspension system is designed. A gap detection semantic segmentation network (GMSSNet) is designed based on position selection and classification in the column direction. The number of model parameters is further reduced by using a 1×1 convolution plus column-wise sub-pixel convolution layer module instead of a fully connected layer. Then, the suspension gap sample set is constructed and the training environment is configured. The designed GMSSNet model is tested for anti-jamming ability, ablation experiments, and closed-loop suspension experiments, respectively. The experimental results show that the GMSSNet model has high detection accuracy, the maximum detection error is ±0.1 mm and the linearity is 0.5% F.S for normal levitation gap detection samples. The maximum detection error of the network is ±0.15 mm and the linearity is 0.75% F.S in the presence of offset or specific occlusion. The closed-loop levitation experiments show that the levitation gap detection accuracy and speed of the GMSSNet model meet the requirements of the suspension system.
-
Research on image-based segmentation and quantification of road cracks
Yu Tianhe, Xu Bochao, Hou Shanchong, Zhao Sicheng, Liu Kexin
Abstract:
Aiming at the contradiction between high cost and high precision in the field of road crack detection and quantification, this paper proposes a low cost, high precision automatic segmentation and quantification system for road cracks. Firstly, the convolutional neural network with jump-stage round-trip multi-scale fusion module and attention gate mechanism is used for segmentation prediction, which is named SW-Net. Then, the cracks are classified by combining MCO, DFS and the trend of pixel statistical curves in different directions. Finally, in order to overcome the discontinuity of crack quantization and the limitation of traditional morphological skeleton quantization algorithm, this paper combined the A* algorithm and extended it to calculate the shortest length and maximum width of cracks. Experimental comparison results show that the system achieves the best accuracy (93.68%) and F1 score (0.896 5) among all comparison models on the Crack500 dataset. The average classification accuracy of the improved classification algorithm is 99.29%, and the classification speed is 109 pieces/s. The relative errors of the shortest length and maximum width are 12.34% and 15.85% respectively, which is 5.16% lower than the average error of the traditional skeleton method. These results show that the system has made remarkable progress in the segmentation, classification and quantification of cracks.
-
Improved ToMP network appearance changes under target tracking method
Chen Renxiang, He Jiale, Yang Lixia, Yu Tengwei, Liang Dong
Abstract:
To address the issue of poor tracking performance caused by target appearance changes during tracking, a target tracking method under appearance change based on improved ToMP network is proposed. Firstly, a target appearance state discrimination module was added to the ToMP network, using a cascaded Long Short-Term Memory (LSTM) network to output the target appearance state discrimination information. Secondly, the online sample storage criterion of the network was improved by adding the normal appearance discrimination information to the confidence score-based evaluation. This optimizes the model weights using reliable samples, enhancing the network′s classification ability for the target. Then, the mechanism for utilizing online samples during appearance changes was improved, updating the model weights with the latest samples to enhance classification performance for newly appeared targets. Finally, a center point trajectory prediction was used to weight the target response score generated by the network, improving the target feature mapping while reducing interference from similar objects, thereby stabilizing target tracking. The accuracy on public datasets reached 93.9% and 68.9%, respectively, outperforming other methods and feasibility is further confirmed through robotics experiments.
-
Multi-type insulator detection algorithm based on improved YOLOv7-tiny
Liu Xi, Chen Chen, Shuang Feng
Abstract:
Aiming at the problems of limited insulator type recognition, poor positioning accuracy and lack of robustness in existing insulator detection algorithms, a multi-type insulator detection algorithm based on improved YOLOv7-tiny is proposed. Firstly, the K-means++ algorithm is used to recluster the anchor box to obtain the anchor box size which is more suitable for multi-type insulator datasets. Secondly, the WIoUv3 loss function based on the dynamic non-monotone focusing mechanism is designed to address the imbalance between positive and negative samples in the training process. On the network structure, firstly, the Cross-stage Feature Fusion-ConvNeXt Block (CFFCB) is used to capture more context information at the Backbone, and some occluded insulators are accurately detected. At the same time, at the Neck, the SPPCSPF (Spatial Pyramid Pooling Cross Stage Partial-Fast) is proposed to replace the original SPPCSP, (Spatial Pyramid Pooling Cross Stage Partial), which effectively improves the detection success rate when the insulator is close to the background, and effectively improves the missed detection situation. After experimental testing, compared with YOLOv7-tiny, the mAP of the improved network model is increased by 2.1%, reaching 97.6%, which effectively improves the detection accuracy of various insulator types. Finally, the grabbing experiment is carried out on the UR5 manipulator by using the detection results of the improved algorithm. The actual grabbing success rate is about 90%, which verifies the feasibility of the algorithm.
-
Optimization of ECT image reconstruction algorithm for flow field measurement of supersonic separator
Wang Shiwei, Wang Chao, Guo Qi, Ding Hongbing
Abstract:
To address the issue of poor image reconstruction quality caused by the ill conditioned nature of the electrical capacitance tomography (ECT) system for measuring field parameters of gas-liquid two-phase flow in a supersonic separator, a new ECT image reconstruction strategy is proposed. This strategy combines regularization methods with guided image filtering. First, the media distribution image is reconstructed using L2 regularization and Lp (0
-
Attention mechanism and multi-source information fusion-based method for bearing fault diagnosis under variable operating conditions
Qiao Huihui, Zhao Erxian, Hao Rujiang, Li Dongsheng, Wang Yongchao
Abstract:
When bearings operate under variable conditions, they are affected by environmental noise and fluctuations in operating parameters. Existing bearing fault diagnosis methods relying on single-source signals struggle in these situations because such signals often fail to provide comprehensive and stable fault information. To address this issue, this paper proposes a multi-source information fusion network model (MSIFNM) based on an attention mechanism. The model′s multi-scale feature extraction module captures more detailed fault features, while the two-stage attention module enhances features that are less sensitive to changes in operating conditions. The multi-source information feature weighting module adaptively assigns weights to the features based on their sensitivity to faults. The feature fusion and classification output module further integrates these weighted features and delivers classification results through fully connected and softmax layers. To validate the effectiveness of the proposed MSIFNM model, bearing datasets under variable speed and load conditions were used. Experimental results demonstrate that the MSIFNM model significantly improves the accuracy, stability, and adaptability of bearing fault diagnosis under variable operating conditions.
-
Fault detection for ex-core neutron detectors in nuclear power plants using the spatial-temporal dynamic detection model
Jiang Hao, Ye Mingxin, Lin Weiqing, Chen Jing, Miao Xiren
Abstract:
In the safety monitoring systems of nuclear power plants, ex-core neutron detectors are essential. Existing fault detection methods for these detectors primarily focus on extracting temporal features and using fixed thresholds to identify faults. These methods do not fully leverage the spatial coupling relationships between detectors and lack flexibility. To address these limitations, this paper introduces a spatial-temporal dynamic detection model (STDDM) for fault detection in ex-core neutron detectors. The model comprises three components: a temporal convolutional network (TCN), a graph convolutional network (GCN), and dynamic thresholds. By combining the TCN and GCN, the model captures implicit spatial-temporal relationships between detectors to reconstruct detector signals. The residuals between the reconstructed and actual signals are then computed, with dynamic thresholds set based on the mean residual and the overall standard deviation of residuals across the reactor. This approach allows the model to adapt to varying reactor operating conditions. Tested with real data from a nuclear power plant, the STDDM not only provides accurate real-time signal reconstruction but also exhibits strong fault tolerance under various fault conditions, proving its effectiveness and practicality for fault detection in ex core neutron detectors.
-
Research on fault diagnosis based on improved federated learning long-tail data
Liu Weimin, Zhan Yihe, Zheng Aiyun, Huang Jide, Zheng Zhi
Abstract:
Due to the inability to collect sufficient fault samples of a certain fault type of gears and bearings failures, the data exhibits a long tail distribution, making it impossible to effectively construct a neural network diagnosis model. When the federal learning method is introduced to solve the above long tail problem, the feature information of the tail fault type sample cannot be effectively extracted. In view of the problems, this paper proposes an improved federated learning method. Firstly, the diagnosis model is retrained by using federal features to improve the fault feature extraction ability of tail samples. Secondly, the CBAM (convolutional block attention module) attention mechanism is introduced to improve the ResNet (residual network) network model in federated learning, boosting its ability and efficiency of extracting key local feature information of channel and space. Thirdly, the traditional convolution is replaced by asymmetric convolution to enhance the ability and efficiency of extracting asymmetric feature information of samples. Finally, the interval calibration algorithm is used to optimize the classification margin of the network model to obtain higher diagnostic accuracy and efficiency. The experimental analysis based on the measured fault samples of gears and bearings shows that the proposed improved federated learning method can effectively improve the average and highest accuracy, by 8.78% and 3.40%, respectively.
-
A transmission line icing level warning method based on LCPSO and heterogeneous ensemble learning model
Shang Qiufeng, Guyuanyu, Fan Xiaokai, Wang Jianjian, Yao Guozhen
Abstract:
Based on Φ-OTDR, distributed fiber optic sensing technology can be utilized to achieve online, real-time health monitoring of power transmission lines by analyzing the vibration characteristics of optical fibers within OPGW. By utilizing Φ-OTDR, vibration signals under three operating conditions—no icing, level 1 icing, and level 2 icing—were collected, and the temporal, frequency, and time-frequency domain features, along with their corresponding statistical properties, of both phase and amplitude signals were thoroughly explored. To enhance the accuracy of icing condition identification, this paper proposes an optimal feature subset selection algorithm based on LCPSO-AdaBoost-MCG. This algorithm employs the classification error rate of the AdaBoost-MCG as the fitness function and iterates with the LCPSO to calculate the optimal feature subset. The AdaBoost ensemble incorporates four weak classifiers: Simple Cartesian Network (SCN), K-Nearest Neighbors (KNN), Probabilistic Neural Network (PNN), and Support Vector Machine (SVM), forming a heterogeneous strong classifier. By leveraging the strengths of each weak classifier, the model′s generalization performance and recognition accuracy are improved. Field data validation demonstrates that the proposed method achieves a 98.7% accuracy in identifying icing levels. Using the optimal feature subset identified in this study, an icing level warning feature library can be established, offering a valuable reference for intelligent transmission line inspection.
-
Wavelet knowledge-driven mechanical equipment fault detection with zero-fault samples
Li Guoqiang, Wei Meirong, Wu Defeng, Wu Jun, Duan Chaoqun
Abstract:
For the zero-fault sample problem of mechanical equipment, transfer learning-based and data generation-based methods have attracted much attention. However, the methods usually require the support of similar fault samples, which makes it difficult to ensure training data is aligned with the real-world fault sample of mechanical equipment in data distribution. The generalization is insufficient in applications. To address the aforementioned problems, a novel fault detection (FD) method is proposed based on a designed new loss function and vision transformer (ViT). First, the continuous wavelet transform knowledgebase is established by combining three different mother wavelet functions, which are used to analyze the monitored signals of mechanical equipment from different time-frequency perspectives. Secondly, a new contrastive loss function is designed based on different time-frequency features and the cosine similarity analysis to effectively optimize the parameters of a constructed ViT. Finally, a fault detection algorithm is proposed to parse the real-time monitored signals of mechanical equipment to achieve the FD. The proposed FD method is evaluated by the industrial robot test rig. The results show that a deep network-based feature encoder with high-performance feature extraction can be established with zero-fault samples, and accurate fault detection for different fault conditions can also be realized.
-
Study on dynamic wireless time comparison technology for visual range based on carrier phase
Zhang Yingbo, Liu Yinhua, Liu Ya, Li Xiaohui, Zhang Jifeng
Abstract:
Carrier phase observations can significantly reduce noise and address substantial clock bias noise that pseudo-range observations alone cannot resolve. However, carrier phase measurements face challenges with whole-cycle ambiguities and cycle slips. This paper explores the principles of bidirectional pseudo-range and carrier phase measurements and the methods for calculating clock bias. It proposes a new approach for sampling and extracting bidirectional carrier phase observations to mitigate the effects of link asymmetry due to inconsistent sampling times on clock bias measurement results. To address the technical challenges of whole-cycle ambiguity resolution and cycle slip detection and correction in bidirectional carrier phase time comparison, an improved pseudo-range method is used to enhance ambiguity estimation accuracy, while a Kalman filter is employed for real-time cycle slip detection. A vehicle-mounted test platform was developed to validate the dynamic time comparison performance. Compared to pseudo-range bidirectional time comparison, the carrier phase bidirectional time comparison improves accuracy by 71.3%, with the root-mean-square error of the measured clock bias reaching the order of hundreds of picoseconds.
-
Study on real-time subdivision and equal division averaging technology for multi-channel grating moire signals
Li Qiang, Liu Hongwei, He Tao, Du Kun, Xia Yangqiu
Abstract:
Circular gratings are susceptible to factors such as engraving errors and installation eccentricity, preventing them from directly meeting the positioning accuracy requirements of high-precision angle measurement systems. Averaging data from multiple read heads can effectively self-calibrate the accuracy of circular gratings, thereby enhancing the precision of angle measurement systems. However, commercial servo control systems typically do not support real-time access to multi-read head signals on the same servo axis, making the online application of the equal division averaging method challenging. This paper explores the sampling, error compensation, real-time subdivision, and equal division averaging of multi-channel grating moire signals. A real-time processing system for multiple read heads was developed using an FPGA platform and applied to an angle measurement turntable. Experiments were designed to evaluate the stability of subdivision values, control resolution, angular positioning accuracy, and subdivision error, verifying the performance of the real-time multi-read head processing system. The system successfully enabled the online application of the equal division averaging method for multiple read heads, meeting the control resolution requirement of 0.005″ on a rotating test bench for angle measurement, achieving a positioning accuracy of 0.58″, and obtaining a repeatability of 0.06″ in positioning accuracy.
-
A joint rotation modulation of two rotating inertial navigation systems based on sensor-level fusion
Zha Feng, Xiao Liangfen, Tong Yude, Lin Hongyi, Bu Haoyu
Abstract:
Large surface and underwater vehicles are typically equipped with two sets of Rotational Inertial Navigation Systems (RINS), which operate independently following the same rotation modulation strategy. The systems only serve as mutual backups, lacking effective information fusion. To improve RINS accuracy, a joint rotation modulation scheme based on sensors information fusion is proposed under the configuration of two RINSs. The classic single axis rotation scheme is optimized at first without changing the structure and arrangement of any single RINS. The rotational strategies and start-stop sequences of the Inertial Measurement Units (IMUs) of both systems are jointly designed to ensure that at any given moment, one RINS is in a stationary state. The output information (from the gyroscopes and accelerometers) of the IMUs during the stationary phases is fused in sequence to reduce the coupling effect between IMU rotation, scale factor errors, and installation errors. Theoretical analysis of error characteristics verifies the advantages of the proposed modulation scheme. Simulation results show that the position error of the system using the joint rotational modulation scheme has decreased from 2.3 n mile/72 h of a single normal RINS to 0.7 n mile/72 h under the typical sensor level.
-
Inversion of geometric parameters and ultrasonic wave velocity in an irregularly liquid-filled pipe model
Shi Shaopeng, Wang Hua, Li Shengya, Fang Zhilong, Wu Wenhe
Abstract:
Fluid-filled pipes are susceptible to deformation and corrosion due to the influence of the media inside and outside the pipe as well as environmental conditions. To ensure system safety and reliable operation, it is necessary to conduct regular inspections using specialized instruments. The practical measurement environment is complex, affected by factors such as pipe deformation, instrument gravity, improper use of centralizers, and variations in the properties of the internal fluid, all of which can hinder the measurement instruments from functioning ideally and reduce the precision of inspections. This paper focuses on ultrasonic logging within fluid-filled casings as a case study to address these challenges. Leveraging the characteristics of ultrasonic logging tools, we have developed an iterative optimization algorithm. This innovative algorithm allows for the ultrasonic measurements to accurately locate the eccentric trajectory of the inspection tool within an irregular fluid-filled pipe, calculate the ultrasonic wave velocity in the pipe′s internal fluid, and infer the internal boundaries of an irregular pipe. Both synthetic data and field data indicate that the proposed method can effectively evaluate the instrument′s measurement state, precisely determine the ultrasonic wave velocity in the internal fluid, and ascertain the geometric parameters (such as shape and position) of irregular pipes. This approach significantly improves the usability of ultrasonic logging tools in fluid-filled casings, providing reliable data for further assessments, such as external cement evaluation in the context of well casings.
-
Design and study of coupling cavity in trans-ice acoustic signal pick-up device
Zhang Zimu, Zhu Guangping, Chen Chaori, Li Lei
Abstract:
In view of the limitations of the existing polar ice acoustic detection technology in terms of detection bandwidth and deployment convenience, this article focuses on the design and testing of the coupling cavity part of a cross-ice underwater acoustic signal pickup device. The device converts elastic waves in the ice into pressure waves in the coupling fluid through a coupling chamber, a flexible shell with an acoustic-solid coupling fluid, and receives the sound pressure signal through a hydrophone immersed in the coupling fluid. Firstly, the material of the coupling cavity is selected through theoretical analysis. Then, the geometric parameters of the coupling cavity are explored by simulation software simulation. Finally, the pool experiment and low-temperature condition test are carried out, and the influence of the geometric parameters of the coupling cavity on the trans-ice detection effect and the reliability of the device is further determined. The test results show that the thickness of the sidewall of the coupling cavity is an important parameter affecting the cross-ice underwater acoustic signal pickup effect and low-temperature adaptability of the device.
-
Autonomous berthing strategy based on Linear Quadratic Regulator(LQR)
Abstract:
Autonomous berthing is a prominent topic in ship automatic navigation research. To effectively address the challenges of path planning and control during ship berthing, this paper proposes an improved LQR control method based on the Frenet framework along with an adaptive berthing strategy decision model. The approach integrates ship motion control, path planning, and berthing strategy selection to achieve adaptive autonomous berthing. First, a ship dynamics model accounting for wind flow interference is established, and the berthing mode is automatically selected based on the spatial relationship between the current wind flow environment and the berth. Next, the berthing path is planned, and the LQR controller is utilized to enable the ship′s autonomous berthing. To verify the controller′s effectiveness, the simulation experiment fully considers the ship′s large drift angle characteristics and the shore effect during berthing. The simulation results demonstrate that the proposed method exhibits strong robustness against environmental disturbances, can select appropriate berthing strategies under varying conditions, and successfully achieves autonomous ship berthing
-
Multi-robot collaborative path planning algorithm for many-to-one task handover
Mao Jianlin, He Zhigang, Zhang Shufan, Li Ruiqi, Zhang Kaixiang
Abstract:
To solve the problems of traditional multi-robot path planning algorithms dealing with a single form of task and large non-essential loss, this article proposes a multi-group many-to-one task processing mode of cooperative dynamic priority safe interval path planning algorithm (Co-DPSIPP). Firstly, the algorithm utilizes the simulated annealing and diffusion search to determine the task handover point of each group of robots with the objective of minimizing the total path length. Then, the improved safe interval path planning algorithm is used to carry out the segmented path planning for all the robots. Furthermore, to deal with the problem that some irrational task handover points may cause regional congestion and lead to solution failure, a cluster prioritization and intermediate point dynamic adjustment planning strategy is designed. Finally, the test results on four benchmark maps show that, compared with the cooperative conflict-based search algorithm (Co-CBS), the proposed algorithm can improve the solution success rate by 73% on average, and reduce the running time and total path length by 56% and 5% on average, respectively. The experimental results show that the proposed algorithm provides a more flexible and scalable solution for the collaborative path planning problem of multi-robot in multi-group many-to-one task scenarios.
-
A hysteresis loop modeling method considering the transition characteristics of hardened ferromagnetic components
Wu Qi, Wang Yujue, Wu Bin, Liu Xiucheng, Han Zhenhua
Abstract:
The depth of the effective hardened layer and the extent of the transition layer are critical factors that influence the performance of surface-hardened mechanical components. They constitute significant parameters in the process of quality control for product evaluation. By exploiting the variances in magnetic properties among different structures, a magnetic detection technique can be developed for assessing hardened layers, offering non-destructive and rapid advantages over conventional metallographic observation methods. This method holds promising potential for direct testing and analysis of hardened layers in components. The Boltzmann function is proposed to describe the gradient law of hysteresis characteristic parameters of materials along the depth direction. By discretely layering multi-layered materials and considering the magnetic field coupling between layers, a T(x) hysteresis loop model for multi-layer structural materials is formulated. The hysteresis loops of the hardened layer sample, obtained by cutting them layer by layer, are calculated by using the particle swarm optimization algorithm and the proposed multilayer hysteresis loop model. The accuracy of the model is validated to evaluate its capability in accurately describing the hysteresis loop of multilayer materials, as well as determining and characterizing both the depth of the hardened layer and transition layer.
-
Improved three-dimensional object detection algorithm based on PointPillars
Tang Xinhua, Dai Daowen, Chen Xiyuan, Pan Shuguo
Abstract:
LiDAR-based object detection technology is widely used in fields such as autonomous driving, robotic navigation, and drones. However, due to the sparsity and uneven distribution of LiDAR point cloud data, object detection and classification face significant challenges. Aiming at this problem, this paper proposes an improved 3D object detection algorithm based on the PointPillars algorithm. Firstly, a more efficient point cloud pillar feature encoding network is designed, incorporating a dual attention encoding network with point-wise and channel-wise attention, enhancing the feature representation capability of each pillar. Secondly, in the backbone network part, the global context information network (GCNet) and CSPDarknet network are integrated to improve the feature map representation ability, allowing the network to extract rich contextual semantic information more comprehensively during the feature extraction phase. Experiments conducted on the KITTI dataset demonstrate that the proposed method achieves higher detection accuracy compared to the baseline model, with mean Average Precision improvements of 2.12%, 2.51%, and 1.84% in easy, moderate, and hard scenarios, respectively. Additionally, the improved algorithm achieves a detection speed of 35.6 FPS, demonstrating that this method effectively enhances detection accuracy while maintaining real-time performance.
-
A blind deconvolution method for terahertz thermal barrier coating adapted by novel window function
Gong Yunli, Cao Binghua, Sun Fengshan, Fan Mengbao, Ye Bo
Abstract:
The seriously overlapped terahertz (THz) signals of thermal barrier coatings (TBCs) result in unrecognizable echoes and reduce the accuracy of thickness measurement. Therefore, the blind deconvolution method for THz thermal barrier coating adapted by the novel window function is proposed in this article. The similarity between the window function and the echo is enhanced to improve the reconstruction precision of the deconvolution signal. Firstly, the features of THz echo of TBCs are explored based on the analytical model. A novel window function is presented to improve the similarity between the window function and the THz echo by cross correlation theory and swarm intelligence algorithm. The FIR filter with the novel window function is used to separate the overlapped echoes. Secondly, the time of flight and refractive index are obtained by the first three echoes to calculate the thickness of TBCs, and the Kirchhoff approximation is employed to characterize the influence of the rough surface of TBCs, followed by correcting the refractive index to reduce the thickness measurement errors. Finally, experiments are implemented to evaluate the effectiveness of the proposed method. Compared with frequency wavelet domain deconvolution and the improved maximum correlated kurtosis deconvolution, the results show that the refractive index measurement accuracy of the proposed method is improved by 76.32% and 83.51%. The thickness measurement accuracy is improved by 76.20% and 89.67%, respectively.
-
Research on blind spot elimination of spread spectrum time domain reflectometry based on signal separation
Chen Hongzhen, Wang Li, Xu Hao, Yang Shanshui
Abstract:
When using the spread spectrum time domain reflectometry for cable fault detection, it is difficult to effectively identify near-distance faults due to the presence of detection blind spot. This not only reduces the detection range but also limits its wide applicability in practical applications. In response to this issue, this paper thoroughly analyzes the mechanism behind the formation of blind spot, and proposes a signal separation-based method to eliminate them. Leveraging the linear superposition of incident and reflected signals, this method applies subtraction in digital signal processing to successfully separate the two signals, enabling accurate identification of reflected signals within the blind zone and extraction of fault information. This method eliminates blind spot through digital signal processing technology, avoiding changes to the hardware platform. The algorithm is simple, easy to implement, and adaptable to a wide range of scenarios. Experimental results show that using this method for online detection of open and short circuit faults in a detection system with a central frequency of 62.5 MHz, the maximum absolute error is only 0.16 m, fully demonstrating its effectiveness and practicality.
-
A review on the research of magnetohydrodynamic angular velocity sensors
Tuo Weixiao, Jiang Haojiao, Li Xingfei, Xing Weida, Wang Tianyu
Abstract:
As detection distances in the aerospace field increase, the impact of micro-angle vibrations—characterized by a broad frequency range (0.1 Hz to 1 kHz) and small amplitude (in the microradian range)—on payload pointing accuracy is becoming increasingly significant. Magnetohydrodynamic (MHD) angular velocity sensors offer advantages such as low noise, wide frequency bandwidth, and compact size, making them the most suitable method for on-orbit micro-angle vibration measurement. These sensors can provide comprehensive data across a wide spectrum, aiding in the analysis of on-orbit micro-angle vibration dynamics and enabling active compensation. This paper reviews the working principles, technical characteristics, and applications of MHD angular velocity sensors while discussing current research challenges. It begins by comparing the two structural forms of MHD angular velocity sensors and analyzing the differences in their output characteristics based on the sensor′s mathematical model. Next, the technical specifications and applications of typical MHD angular velocity sensors, both domestic and international, are introduced.
-
A novel single-row absolute displacement sensor featuring dual-frequency magnetic field time-sharing excitation and linear time-grating displacement measurement
Yang Jisen, Xiu Fu, Zhang Jing, Wen Jie, Liu Jiacheng
Abstract:
To address the need for compact installation and high-precision linear displacement measurement, a novel single-row absolute linear time-grating displacement sensor structure is designed using a dual-frequency magnetic field time-sharing excitation scheme and outlier frequency reduction principle. This design resolves the conflict between increasing the frequency of sensor excitation signals for a higher signal to noise ratio and achieving high resolution. First, a transient magnetic field coupling model for a planar coil is established to create a single absolute linear time-grating measurement model and its sensing mechanism. The proposed solution for absolute position determination minimizes measurement error influence. Electromagnetic field simulations are used to analyze the coupling characteristics of various induction coil shapes and air gap magnetic fields, leading to an optimized sensor installation gap of 0.6 mm. The sensor employs 500 kHz and 1 MHz time-sharing excitation drive schemes and introduces a novel decoupling method for outlier frequency reduction, which ensures high resolution while enhancing the signal-to-noise ratio. The sensor prototype was fabricated and tested, with experimental results demonstrating a 36.4% increase in measurement accuracy using the outlier frequency reduction method compared to the original direct decoupling method. The sensor provides an effective measurement range of 187.68 mm, with an accuracy of ±4.9 μm and a resolution of 0.14 μm after error compensation. Compared to mainstream international products, this time-grating displacement sensor offers high accuracy, high resolution, compact size, and low cost while reducing reliance on ultra-precise grating etching and electronic subdivision technologies.
-
The bistable piezoelectric energy harvester with magnetic attraction and repulsion between upper and lower beams
Jin Hong, Tang Baoping, Zhao Chunhua, Du Siyu
Abstract:
To address the problems of narrow frequency bands and significant barriers posed by bistable potential wells in piezoelectric energy harvesting with single cantilever beams, a novel solution is proposed by utilizing the bistable piezoelectric energy harvester employing magnetic attraction and repulsion between upper and lower beams. Firstly, a bistable structure with mutually-excited upper and lower beams and externally fixed magnets is constructed to broaden the vibration response and operational frequency range of the structure. By modeling the magnets as magnetic dipoles and establishing a planar magnetic force model, the impact of different parameters on the structural potential energy and magnetic force are determined. Subsequently, based on Euler-Bernoulli beam theory and Lagrangian equations, a comprehensive dynamic model of the system is formulated to analyze the influence of dynamic characteristics on the system. Experimental results show that, compared with the single upper or lower beams without fixed magnets, the proposed bistable piezoelectric energy harvester with magnetic attraction and repulsion expands the operational frequency range of the upper cantilever beam from 5~6 Hz to 3~11 Hz, resulting in a 26.35% increase in average power output within the 3~11 Hz band. Similarly, the operational frequency range of the lower cantilever beam is extended from 13~14 Hz to 4~13 Hz, with a 23.28% enhancement in average power output within the 4~13 Hz range.
-
Orthodontic force modeling of vertical closing loop combined with micro-implant for the treatment of anterior teeth adduction
Jiang Jingang, Tan Yujian, Li Changpeng, Zhai Shuojian, Zhang Yongde
Abstract:
Dental crowding is the most prevalent type of malocclusion. To alleviate the condition of tooth crowding in orthodontic treatment, the typical approach involves the extraction of the first premolars and subsequent comprehensive retraction of the anterior teeth. During the retraction process, a combination of vertical closing loops and micro-implant anchorage is often used to close the gaps between the teeth. However, in the course of treatment, physicians often rely on qualitative non-quantitative methods to describe forces and movements, which makes it challenging to accurately predict the treatment outcomes. To address this problem, the superposition theorem was used to obtain a model for predicting the orthodontic force and torque of vertical closing loop combined with micro-implant anchorage, which was parameterized by the cross-section size and shape of the archwire and the height of micro-implant traction. The theoretical and simulated values for each operational condition were compared, and the error was determined through finite element analysis. The error between the simulated and theoretical values for corrective force was within 0.09 N, while for corrective torque, it was within 0.75 N·mm. Further measurements of orthodontic force and moment on mandibular wax models revealed that, with low traction, the error between theoretical and experimental values for orthodontic force ranged from 0.03~0.18 N, and for orthodontic moment from 0.51~1.1 N·mm. With high traction, the force error ranged from 0.03~0.17 N, and the moment error from 0.23~1.30 N·mm. This serves to validate the accuracy of the theoretical model and the reliability of the simulation conditions. The model can parametrically represent the force in the correction process, thus providing a foundation for personalized treatment planning and enhancing treatment efficacy and safety.
Visual inspection and Image Measurement
Industrial Big Data and Intelligent Health Assessment
Precision Measurement Technology and Instrument
Information Processing Technology
传感器技术
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