Pub Date : 2024-05-06DOI: 10.1177/01423312241247873
Zhi-ying He, Hongji Pu
Original iterative learning control (OILC) has been proved a powerful tool in dealing with the model-free control problems by repetitively corrections based on the control error. However, the steady-state error under widely-used proportional-type original iterative learning control (P-type OILC) is highly corresponded to the proportional learning gain, making the algorithm parameter-determined. Therefore, a new gradient-descent iterative learning control (GDILC) algorithm is proposed to achieve a parameter-free approach by simulating the gradient-descent process. First, GDILC problem is formulated mathematically. Next, the idea of the algorithm is proposed, the analyses of the convergence and the steady-state error are conducted and the algorithm is implemented. GDILC will generate a random correction with a gradient-descent upper bound, rather than a correction proportional to the error in P-type OILC. Finally, illustrative and application simulations are conducted to validate the algorithm. Results show that the algorithm will be convergent after adequate iterations under proper corrections. The steady-state error will be less affected by the algorithm parameters under GDILC than that under OILC.
{"title":"A gradient-descent iterative learning control algorithm for a non-linear system","authors":"Zhi-ying He, Hongji Pu","doi":"10.1177/01423312241247873","DOIUrl":"https://doi.org/10.1177/01423312241247873","url":null,"abstract":"Original iterative learning control (OILC) has been proved a powerful tool in dealing with the model-free control problems by repetitively corrections based on the control error. However, the steady-state error under widely-used proportional-type original iterative learning control (P-type OILC) is highly corresponded to the proportional learning gain, making the algorithm parameter-determined. Therefore, a new gradient-descent iterative learning control (GDILC) algorithm is proposed to achieve a parameter-free approach by simulating the gradient-descent process. First, GDILC problem is formulated mathematically. Next, the idea of the algorithm is proposed, the analyses of the convergence and the steady-state error are conducted and the algorithm is implemented. GDILC will generate a random correction with a gradient-descent upper bound, rather than a correction proportional to the error in P-type OILC. Finally, illustrative and application simulations are conducted to validate the algorithm. Results show that the algorithm will be convergent after adequate iterations under proper corrections. The steady-state error will be less affected by the algorithm parameters under GDILC than that under OILC.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141010691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-25DOI: 10.1177/01423312241233622
Mohammad Fiuzy, S. Shamaghdari
This paper investigates a type of robust observer-based tracking control for the well-known inverted pendulum as an uncertain fractional-order system with two-norm bounded uncertainties subject to input saturation along with external disturbance that focuses on the case of a fractional-order α such that [Formula: see text]. The main goal is to design a robust model following tracking control servomechanism by the side of stability analysis. This paper addresses the well-known quadratic Lyapunov function in addition to the Gronwall–Bellman lemma and the sector condition of the saturation function stability synthesis. This paper proposes an outstanding strategy to achieve the best model following robust observer in the cast of output-feedback control subject to input saturation in the framework of an uncertain fractional-order system based on the linear matrix inequality (LMI) solution. The LMI procedure can be used to achieve static output-feedback tracking control and observer gain. Several valuable theorems support this approach, and various simulation scenarios are available to showcase its effectiveness.
{"title":"Robust H∞ adaptive model following control for fractional-order systems with polytopic and bounded uncertainties subject to input saturation","authors":"Mohammad Fiuzy, S. Shamaghdari","doi":"10.1177/01423312241233622","DOIUrl":"https://doi.org/10.1177/01423312241233622","url":null,"abstract":"This paper investigates a type of robust observer-based tracking control for the well-known inverted pendulum as an uncertain fractional-order system with two-norm bounded uncertainties subject to input saturation along with external disturbance that focuses on the case of a fractional-order α such that [Formula: see text]. The main goal is to design a robust model following tracking control servomechanism by the side of stability analysis. This paper addresses the well-known quadratic Lyapunov function in addition to the Gronwall–Bellman lemma and the sector condition of the saturation function stability synthesis. This paper proposes an outstanding strategy to achieve the best model following robust observer in the cast of output-feedback control subject to input saturation in the framework of an uncertain fractional-order system based on the linear matrix inequality (LMI) solution. The LMI procedure can be used to achieve static output-feedback tracking control and observer gain. Several valuable theorems support this approach, and various simulation scenarios are available to showcase its effectiveness.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140653833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-24DOI: 10.1177/01423312241239233
Yang Lu, Guochen Pang, Yang Liu, A. Zhang, Jianlong Qiu, Jinde Cao
This paper deals with the problem of event-triggered control (ETC) for vertical take-off and landing (VTOL) aircraft systems under actuator saturation. By introducing the VTOL system with saturated actuators, a composite quadratic Lyapunov function (CQLF) approach is first proposed to describe the problem of the event triggering mechanism. The proposed transfer mechanism not only further saves communication resources, but also further suppresses the conservatism of the attraction domain estimation. Then, sufficient conditions are established via linear matrix inequality (LMI) by incorporating new event triggering mechanism condition and prescribed passive performance metrics under the closed-loop system that is asymptotically stable and strictly passive, respectively. Finally, the method proposed in this paper can effectively reduce the number of triggers and expand the domain of attraction. The validity and feasibility of the obtained results are demonstrated by two example simulations.
{"title":"Composite quadratic Lyapunov function event-triggered control of vertical take-off and landing aircraft systems with actuator saturation","authors":"Yang Lu, Guochen Pang, Yang Liu, A. Zhang, Jianlong Qiu, Jinde Cao","doi":"10.1177/01423312241239233","DOIUrl":"https://doi.org/10.1177/01423312241239233","url":null,"abstract":"This paper deals with the problem of event-triggered control (ETC) for vertical take-off and landing (VTOL) aircraft systems under actuator saturation. By introducing the VTOL system with saturated actuators, a composite quadratic Lyapunov function (CQLF) approach is first proposed to describe the problem of the event triggering mechanism. The proposed transfer mechanism not only further saves communication resources, but also further suppresses the conservatism of the attraction domain estimation. Then, sufficient conditions are established via linear matrix inequality (LMI) by incorporating new event triggering mechanism condition and prescribed passive performance metrics under the closed-loop system that is asymptotically stable and strictly passive, respectively. Finally, the method proposed in this paper can effectively reduce the number of triggers and expand the domain of attraction. The validity and feasibility of the obtained results are demonstrated by two example simulations.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140659509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-24DOI: 10.1177/01423312241242845
Hui Bi, Tao Zou, Lihua Wu
This paper introduces an adaptive neural network compensatory control approach designed for a 2-degree-of-freedom (2-DOF) helicopter system facing challenges such as input backlash and state constraints. The proposed methodology leverages a radial basis function (RBF) neural network to effectively approximate system uncertainties, mitigating the impact of nonlinear dynamics on control performance. To address the presence of nonlinear input backlash, a compensation technique is introduced to enhance the smoothness of input signals. In addition, for enhanced system safety, a barrier Lyapunov function is integrated to impose restrictions on position and velocity states, resulting in constrained control. Through a rigorous analysis using the Lyapunov direct method, this paper demonstrates the effectiveness of the proposed approach in achieving bounded stability of the system. The validation of the approach is further established through the presentation of simulation and experimental results, showcasing its effectiveness and feasibility in real-world applications.
{"title":"Full-state constraints and input backlash–based neural network control of a 2-DOF helicopter system","authors":"Hui Bi, Tao Zou, Lihua Wu","doi":"10.1177/01423312241242845","DOIUrl":"https://doi.org/10.1177/01423312241242845","url":null,"abstract":"This paper introduces an adaptive neural network compensatory control approach designed for a 2-degree-of-freedom (2-DOF) helicopter system facing challenges such as input backlash and state constraints. The proposed methodology leverages a radial basis function (RBF) neural network to effectively approximate system uncertainties, mitigating the impact of nonlinear dynamics on control performance. To address the presence of nonlinear input backlash, a compensation technique is introduced to enhance the smoothness of input signals. In addition, for enhanced system safety, a barrier Lyapunov function is integrated to impose restrictions on position and velocity states, resulting in constrained control. Through a rigorous analysis using the Lyapunov direct method, this paper demonstrates the effectiveness of the proposed approach in achieving bounded stability of the system. The validation of the approach is further established through the presentation of simulation and experimental results, showcasing its effectiveness and feasibility in real-world applications.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140665974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Constant tension control is essential for excellent winding quality. However, the system’s nonlinearity and external disturbance make it challenging to guarantee tension control accuracy with conventional control methods. Thus, a self-coupling fractional-order proportional–integral–derivative (SC-FOPID) control scheme combined with a disturbance observer is proposed to enhance the system’s anti-vibration performance. The fractional-order dynamic model of the unwinding roller and swing rod is established by analyzing the tension mechanism. Based on deliberate analysis and calculation, the vibration shock signal can be decomposed into periodic sinusoidal disturbance and bounded noise approximately. As such, an output-based anti-vibration method using a fractional-order model can be realized, where a back recursive disturbance observer is designed to estimate the periodic component. Simultaneously, the bounded noise exhibited in vibration can be attenuated by the SC-FOPID controller. The stability is guaranteed using the Lyapunov theorem, and the simulation results show the proposed method’s effectiveness in improving the tension control performance.
{"title":"Research on anti-vibration tension control based on self-coupling fractional-order PID","authors":"Jianguo Liang, Yujie Duan, Xinyu Wen, Yinan Zhao, Haifeng Gao, Xiaodong Zhao, Uwayezu Emmanuel","doi":"10.1177/01423312241239113","DOIUrl":"https://doi.org/10.1177/01423312241239113","url":null,"abstract":"Constant tension control is essential for excellent winding quality. However, the system’s nonlinearity and external disturbance make it challenging to guarantee tension control accuracy with conventional control methods. Thus, a self-coupling fractional-order proportional–integral–derivative (SC-FOPID) control scheme combined with a disturbance observer is proposed to enhance the system’s anti-vibration performance. The fractional-order dynamic model of the unwinding roller and swing rod is established by analyzing the tension mechanism. Based on deliberate analysis and calculation, the vibration shock signal can be decomposed into periodic sinusoidal disturbance and bounded noise approximately. As such, an output-based anti-vibration method using a fractional-order model can be realized, where a back recursive disturbance observer is designed to estimate the periodic component. Simultaneously, the bounded noise exhibited in vibration can be attenuated by the SC-FOPID controller. The stability is guaranteed using the Lyapunov theorem, and the simulation results show the proposed method’s effectiveness in improving the tension control performance.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140669639","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-22DOI: 10.1177/01423312241245474
Qixun Zhou, Yihang Wang, Yingxing Zhang, Long Zhang, Keke Shi
Aiming at the problems of chattering and low accuracy of rotor position observation in traditional sliding mode observer sensorless control of permanent magnet synchronous motor, an improved sliding mode observer sensorless control method is proposed. First, the sigmoid function is used to replace the traditional switching function, which can effectively improve the chattering problem of the sliding mode observer. Then, an adaptive back electromotive force observer is designed to avoid the use of low-pass filter and phase compensation and reduce the high-frequency harmonics in the back electromotive force estimation. Finally, the speed and rotor position information are obtained by improving the phase-locked loop, and the fundamental harmonics caused by the inverter nonlinearity are suppressed. The simulation and experimental results show that the improved sliding mode observer can effectively weaken the chattering problem of the system, and improve the position observation accuracy and the robustness of the system.
{"title":"Sensorless control of permanent magnet synchronous motor based on improved sliding mode observer","authors":"Qixun Zhou, Yihang Wang, Yingxing Zhang, Long Zhang, Keke Shi","doi":"10.1177/01423312241245474","DOIUrl":"https://doi.org/10.1177/01423312241245474","url":null,"abstract":"Aiming at the problems of chattering and low accuracy of rotor position observation in traditional sliding mode observer sensorless control of permanent magnet synchronous motor, an improved sliding mode observer sensorless control method is proposed. First, the sigmoid function is used to replace the traditional switching function, which can effectively improve the chattering problem of the sliding mode observer. Then, an adaptive back electromotive force observer is designed to avoid the use of low-pass filter and phase compensation and reduce the high-frequency harmonics in the back electromotive force estimation. Finally, the speed and rotor position information are obtained by improving the phase-locked loop, and the fundamental harmonics caused by the inverter nonlinearity are suppressed. The simulation and experimental results show that the improved sliding mode observer can effectively weaken the chattering problem of the system, and improve the position observation accuracy and the robustness of the system.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140672165","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-22DOI: 10.1177/01423312241239216
Xiaohe Li, Jing Bai, Guoguang Wen, Xue Xia
This paper investigates the observer-based bipartite consensus tracking of nonlinear fractional-order multi-agent systems (FOMASs) by employing a pull-based dynamic event-triggered mechanism (DETM). First, considering that the relevant state information of each agent is not always measurable, a class of distributed observers is considered for each agent to estimate its state information. Then, a pull-based DETM for FOMASs is proposed to avoid continuous controller updates, in which the dynamic threshold is modulated according to the preset conditions. The pull-based DETM constructed in this paper enables each agent to update the controller only based on its own trigger instants. Furthermore, an observer-based dynamic event-triggered control protocol is designed to guarantee the bipartite consensus tracking of FOMASs. Correspondingly, sufficient conditions are obtained by using graph theory and choosing suitable Lyapunov candidate functions. Moreover, the Zeno behavior is precluded. Finally, two simulation examples are presented to illustrate the theoretical results efficiently.
{"title":"Observer-based bipartite consensus tracking of nonlinear fractional-order multi-agent systems with pull-based dynamic event-triggered mechanism","authors":"Xiaohe Li, Jing Bai, Guoguang Wen, Xue Xia","doi":"10.1177/01423312241239216","DOIUrl":"https://doi.org/10.1177/01423312241239216","url":null,"abstract":"This paper investigates the observer-based bipartite consensus tracking of nonlinear fractional-order multi-agent systems (FOMASs) by employing a pull-based dynamic event-triggered mechanism (DETM). First, considering that the relevant state information of each agent is not always measurable, a class of distributed observers is considered for each agent to estimate its state information. Then, a pull-based DETM for FOMASs is proposed to avoid continuous controller updates, in which the dynamic threshold is modulated according to the preset conditions. The pull-based DETM constructed in this paper enables each agent to update the controller only based on its own trigger instants. Furthermore, an observer-based dynamic event-triggered control protocol is designed to guarantee the bipartite consensus tracking of FOMASs. Correspondingly, sufficient conditions are obtained by using graph theory and choosing suitable Lyapunov candidate functions. Moreover, the Zeno behavior is precluded. Finally, two simulation examples are presented to illustrate the theoretical results efficiently.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140673068","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-22DOI: 10.1177/01423312241247346
Yuequan Yang, Wei Li, Zhiqiang Cao, Jiatong Bao, Fudong Li
Grasping detection is one of the crucial capabilities for robot systems. Deep learning has achieved remarkable outcomes in robot grasping tasks; however, many deep neural networks were at the expense of high computation cost with memory requirements, which hindered their deployment on computing-constrained devices. To solve this problem, this paper proposes an end-to-end lightweight network with dual attention and inverted residual strategies (LiDAIR), which adopts a generative pixel-level prediction to achieve grasp detection. The LiDAIR is composed of the convolution modules (Conv), the inverted residual convolution module (IRCM), the convolutional block attention connection module (CBACM), and the transposed convolution modules (TConv). The Convs are utilized in downsampling processes to extract the input image features. Then, the IRCM is proposed as a bridge between the downsampling and upsampling phases. In the upsampling phase, the CBACM is designed to focus on the valuable regions from spatial and channel dimensions, where the skip connection is employed to attain multi-level feature fusion. Afterwards, the TConvs are used to restore image resolution. The LiDAIR is lightweight with 704K parameters and enjoys a good tradeoff among lightweight structure, accuracy, and speed. It was evaluated on both the Cornell data set and the Jacquard data set within 10 ms inference time, and the detection accuracy on both the data sets were 97.7% and 92.7%, respectively.
{"title":"Lightweight robotic grasping detection network based on dual attention and inverted residual","authors":"Yuequan Yang, Wei Li, Zhiqiang Cao, Jiatong Bao, Fudong Li","doi":"10.1177/01423312241247346","DOIUrl":"https://doi.org/10.1177/01423312241247346","url":null,"abstract":"Grasping detection is one of the crucial capabilities for robot systems. Deep learning has achieved remarkable outcomes in robot grasping tasks; however, many deep neural networks were at the expense of high computation cost with memory requirements, which hindered their deployment on computing-constrained devices. To solve this problem, this paper proposes an end-to-end lightweight network with dual attention and inverted residual strategies (LiDAIR), which adopts a generative pixel-level prediction to achieve grasp detection. The LiDAIR is composed of the convolution modules (Conv), the inverted residual convolution module (IRCM), the convolutional block attention connection module (CBACM), and the transposed convolution modules (TConv). The Convs are utilized in downsampling processes to extract the input image features. Then, the IRCM is proposed as a bridge between the downsampling and upsampling phases. In the upsampling phase, the CBACM is designed to focus on the valuable regions from spatial and channel dimensions, where the skip connection is employed to attain multi-level feature fusion. Afterwards, the TConvs are used to restore image resolution. The LiDAIR is lightweight with 704K parameters and enjoys a good tradeoff among lightweight structure, accuracy, and speed. It was evaluated on both the Cornell data set and the Jacquard data set within 10 ms inference time, and the detection accuracy on both the data sets were 97.7% and 92.7%, respectively.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140672473","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-22DOI: 10.1177/01423312241239020
Qing-Huang Song, Boyuan Wang, Yuandong Ma, Mengjie Hu, Chun Liu
In the domain of autonomous driving, object detection presents several complex challenges, particularly concerning the accurate identification of small and salient objects. This paper introduces DL-YOLOX (Dilated Enhancement YOLOX), which flexibly uses dilated convolution to enhance features to achieve the purpose of improving small objects and silent objects. As we all know, a large receptive field covers a larger area and has greater contextual information, which is more advantageous for detecting large targets. A small receptive field helps capture local details and has better detection capabilities for detecting small targets. To bolster the representation of objects across various scales, we propose the integration of Dilated Adaptive Feature Fusion (DAFF) which has the ability to adaptively fuse features with different receptive fields. This innovative fusion mechanism allows for a more comprehensive understanding of objects, enabling improved detection accuracy even for objects of varying sizes. In addition, we tackle the issue of small object loss during feature propagation by introducing Stack Dilated Module (SDM), a powerful module that mitigates this phenomenon and contributes to better detection performance. Moreover, we endeavor to enhance small object detection further by replacing the conventional Intersection over Union (IoU) metric with Normalized Gaussian Wasserstein Distance (NWD), a novel distance metric that proves to be more effective in accurately gauging small object detection, thus elevating the precision of our algorithm. To thoroughly evaluate the robustness and generalization capabilities of our proposed method, we conduct extensive experiments on two benchmark datasets, namely MS COCO 2017 and BDD100K. The results from our evaluation not only affirm the significant improvements achieved in multi-scale object detection but also highlight the real-time capability of our approach. The impressive performance across these datasets demonstrates the promising potential of DL-YOLOX in revolutionizing object detection techniques in the context of autonomous driving.
在自动驾驶领域,物体检测面临着一些复杂的挑战,尤其是如何准确识别小物体和突出物体。本文介绍了 DL-YOLOX(扩张增强 YOLOX),它灵活地利用扩张卷积来增强特征,从而达到改善小物体和无声物体的目的。众所周知,大的感受野覆盖面积更大,具有更多的上下文信息,这对检测大型目标更有利。小的感受野有助于捕捉局部细节,在检测小目标时具有更好的检测能力。为了加强对不同尺度物体的表征,我们建议整合稀释自适应特征融合(DAFF),它能够自适应地融合不同感受野的特征。这种创新的融合机制可以更全面地了解物体,即使是不同大小的物体,也能提高检测精度。此外,我们还通过引入堆栈稀释模块(SDM)来解决小物体在特征传播过程中的损失问题,该模块功能强大,可减轻这一现象,并有助于提高检测性能。此外,我们还用归一化高斯瓦瑟斯坦距离(NWD)取代了传统的 "交集大于联合"(IoU)度量,从而进一步增强了小目标检测能力。为了全面评估我们提出的方法的鲁棒性和泛化能力,我们在两个基准数据集(即 MS COCO 2017 和 BDD100K)上进行了广泛的实验。评估结果不仅肯定了我们在多尺度物体检测方面取得的显著改进,还突出了我们方法的实时性。DL-YOLOX 在这些数据集上的出色表现证明了它在自动驾驶背景下革新物体检测技术的巨大潜力。
{"title":"DL-YOLOX: Real-time object detection via adjustable dilated enhancement for autonomous driving scene","authors":"Qing-Huang Song, Boyuan Wang, Yuandong Ma, Mengjie Hu, Chun Liu","doi":"10.1177/01423312241239020","DOIUrl":"https://doi.org/10.1177/01423312241239020","url":null,"abstract":"In the domain of autonomous driving, object detection presents several complex challenges, particularly concerning the accurate identification of small and salient objects. This paper introduces DL-YOLOX (Dilated Enhancement YOLOX), which flexibly uses dilated convolution to enhance features to achieve the purpose of improving small objects and silent objects. As we all know, a large receptive field covers a larger area and has greater contextual information, which is more advantageous for detecting large targets. A small receptive field helps capture local details and has better detection capabilities for detecting small targets. To bolster the representation of objects across various scales, we propose the integration of Dilated Adaptive Feature Fusion (DAFF) which has the ability to adaptively fuse features with different receptive fields. This innovative fusion mechanism allows for a more comprehensive understanding of objects, enabling improved detection accuracy even for objects of varying sizes. In addition, we tackle the issue of small object loss during feature propagation by introducing Stack Dilated Module (SDM), a powerful module that mitigates this phenomenon and contributes to better detection performance. Moreover, we endeavor to enhance small object detection further by replacing the conventional Intersection over Union (IoU) metric with Normalized Gaussian Wasserstein Distance (NWD), a novel distance metric that proves to be more effective in accurately gauging small object detection, thus elevating the precision of our algorithm. To thoroughly evaluate the robustness and generalization capabilities of our proposed method, we conduct extensive experiments on two benchmark datasets, namely MS COCO 2017 and BDD100K. The results from our evaluation not only affirm the significant improvements achieved in multi-scale object detection but also highlight the real-time capability of our approach. The impressive performance across these datasets demonstrates the promising potential of DL-YOLOX in revolutionizing object detection techniques in the context of autonomous driving.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140675124","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-20DOI: 10.1177/01423312241239714
Linyan Han, Jianliang Mao, Haibo Du, Yahui Gan, Shihua Li
Facing the system uncertainties caused by unmodeled dynamics and unpredictable external disturbances, the robot position control for meeting the high-performance control requirements on higher accuracy and faster beat is vital for many industrial applications, such as welding and laser cutting tasks. This work aims to cope with the problem of precise and fast position tracking for robot manipulators with an effective and safe control scheme. Specifically, a discrete-time super-twisting observer (STO) is integrated into the scheme to estimate the uncertain dynamics (e.g. unmodeled dynamics and external disturbances) in the feedforward compensation part of the dynamics. Subsequently, a discrete-time fast terminal sliding mode controller (FTSMC) dominates the robot control to guarantee fast convergence of the position tracking error. The significant improvement of the proposed method with respect to other discrete-time sliding mode control approaches lies in that it is capable of alleviating the chattering-like problem, achieving a fast convergence and improving the robustness of sliding mode control against uncertain dynamics. To illustrate the effectiveness of the presented control scheme, several experiments on a six degrees of freedom (6DoF) robot manipulator are provided.
{"title":"Disturbance-rejection position tracking control of industrial robots via a discrete-time super-twisting observer–based fast terminal sliding mode approach","authors":"Linyan Han, Jianliang Mao, Haibo Du, Yahui Gan, Shihua Li","doi":"10.1177/01423312241239714","DOIUrl":"https://doi.org/10.1177/01423312241239714","url":null,"abstract":"Facing the system uncertainties caused by unmodeled dynamics and unpredictable external disturbances, the robot position control for meeting the high-performance control requirements on higher accuracy and faster beat is vital for many industrial applications, such as welding and laser cutting tasks. This work aims to cope with the problem of precise and fast position tracking for robot manipulators with an effective and safe control scheme. Specifically, a discrete-time super-twisting observer (STO) is integrated into the scheme to estimate the uncertain dynamics (e.g. unmodeled dynamics and external disturbances) in the feedforward compensation part of the dynamics. Subsequently, a discrete-time fast terminal sliding mode controller (FTSMC) dominates the robot control to guarantee fast convergence of the position tracking error. The significant improvement of the proposed method with respect to other discrete-time sliding mode control approaches lies in that it is capable of alleviating the chattering-like problem, achieving a fast convergence and improving the robustness of sliding mode control against uncertain dynamics. To illustrate the effectiveness of the presented control scheme, several experiments on a six degrees of freedom (6DoF) robot manipulator are provided.","PeriodicalId":49426,"journal":{"name":"Transactions of the Institute of Measurement and Control","volume":null,"pages":null},"PeriodicalIF":1.8,"publicationDate":"2024-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140681422","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}