Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216953
L. Moormann, J. V. D. Mortel-Fronczak, W. Fokkink, J. Rooda
Supervisor synthesis enables the design of supervisory controllers for large cyber-physical systems, with high guarantees for functionality and safety. The complexity of the synthesis problem, however, increases exponentially with the number of system components in the cyber-physical system and the number of models of this system, often resulting in lengthy or even unsolvable synthesis procedures. In this paper, a new method is proposed for reducing the model of the system before synthesis to decrease the required computational time and effort. The method consists of three steps for model reduction, that are mainly based on symmetry in dependency graphs of the system. Dependency graphs visualize the components in the system and the relations between these components. The proposed method is applied in a case study on the design of a supervisory controller for a road tunnel. In this case study, the model reduction steps are described, and results are shown on the effectiveness of model reduction in terms of model size and synthesis time.
{"title":"Exploiting Symmetry in Dependency Graphs for Model Reduction in Supervisor Synthesis","authors":"L. Moormann, J. V. D. Mortel-Fronczak, W. Fokkink, J. Rooda","doi":"10.1109/CASE48305.2020.9216953","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216953","url":null,"abstract":"Supervisor synthesis enables the design of supervisory controllers for large cyber-physical systems, with high guarantees for functionality and safety. The complexity of the synthesis problem, however, increases exponentially with the number of system components in the cyber-physical system and the number of models of this system, often resulting in lengthy or even unsolvable synthesis procedures. In this paper, a new method is proposed for reducing the model of the system before synthesis to decrease the required computational time and effort. The method consists of three steps for model reduction, that are mainly based on symmetry in dependency graphs of the system. Dependency graphs visualize the components in the system and the relations between these components. The proposed method is applied in a case study on the design of a supervisory controller for a road tunnel. In this case study, the model reduction steps are described, and results are shown on the effectiveness of model reduction in terms of model size and synthesis time.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130012400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216771
Seyed Mahdi Shamsi, N. Napp
Parameter calibration is a burdensome yet essential part of robotics development which traditionally was done through manual calibration routines. Over the recent years, many authors have proposed to automatically calibrate parameters directly from the input data, during the operation of the robot. While the majority of methods are discussed within the context of a specific application and benefit from the particularity of the problem, some generic approaches are proposed that are applicable to a wide spectrum of problems. However in practice, they require re-implementation and customization every time applied to a different domain, due to coupling of formulations with the model, e.g. linearization steps, matrix decomposition, closed form solving, etc. In this paper, we exploit the expressiveness of general purpose probabilistic programming languages (PPLs) to build a generic online calibration framework that can estimate the parameters of arbitrary robotic systems during operation. The proposed approach, based on Bayes filter and Monte Carlo methods, only requires model specification and works as a black-box otherwise. Hence, it spans the generality to the implementation aspect of the calibration problem which facilitates a range of new applications, e.g. fast prototyping of arbitrary robots. We show a short PPL program is capable of calibrating kinematic, extrinsic, and noise parameters of a classic SLAM dataset with minimum knowledge about the system and the parameters.
{"title":"A Generic Online Parameter (Re-)calibration Framework Using PPL","authors":"Seyed Mahdi Shamsi, N. Napp","doi":"10.1109/CASE48305.2020.9216771","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216771","url":null,"abstract":"Parameter calibration is a burdensome yet essential part of robotics development which traditionally was done through manual calibration routines. Over the recent years, many authors have proposed to automatically calibrate parameters directly from the input data, during the operation of the robot. While the majority of methods are discussed within the context of a specific application and benefit from the particularity of the problem, some generic approaches are proposed that are applicable to a wide spectrum of problems. However in practice, they require re-implementation and customization every time applied to a different domain, due to coupling of formulations with the model, e.g. linearization steps, matrix decomposition, closed form solving, etc. In this paper, we exploit the expressiveness of general purpose probabilistic programming languages (PPLs) to build a generic online calibration framework that can estimate the parameters of arbitrary robotic systems during operation. The proposed approach, based on Bayes filter and Monte Carlo methods, only requires model specification and works as a black-box otherwise. Hence, it spans the generality to the implementation aspect of the calibration problem which facilitates a range of new applications, e.g. fast prototyping of arbitrary robots. We show a short PPL program is capable of calibrating kinematic, extrinsic, and noise parameters of a classic SLAM dataset with minimum knowledge about the system and the parameters.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129551626","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216769
Pengcheng Xia, Dapeng Zhang
Crowd counting, which requires to estimate crowd density from an image, is still a challenging task in computer vision. Most of the current methods are focused on large scale variation of people and ignore the huge distribution difference of crowd. To tackle these two problems together, we propose a novel framework named Spatial Normalization Network (SNNet). We normalize multi-scale features from parallel subnetworks to a particular scale and then fuse them to acquire rich spatial information for final accurate density map predictions. Furthermore, we propose a novel normalization layer called Spatial Group Normalization (SGN), which firstly split feature maps along the spatial dimension and then perform group-wise normalization. It’s useful to solve statistic shift problems caused by the great difference of distribution in crowd counting. Moreover, SGN can be naturally plugged into existing solutions and brings significant improvement in crowd counting. Our proposed SNNet achieves state-of-the-art performance on four challenging crowd counting datasets (ShanghaiTech, UCFQNRF, GCC and TRANCOS datasets), which demonstrates the effectiveness and robust feature learning capability of our methods.
{"title":"Crowd Counting with Spatial Normalization Network","authors":"Pengcheng Xia, Dapeng Zhang","doi":"10.1109/CASE48305.2020.9216769","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216769","url":null,"abstract":"Crowd counting, which requires to estimate crowd density from an image, is still a challenging task in computer vision. Most of the current methods are focused on large scale variation of people and ignore the huge distribution difference of crowd. To tackle these two problems together, we propose a novel framework named Spatial Normalization Network (SNNet). We normalize multi-scale features from parallel subnetworks to a particular scale and then fuse them to acquire rich spatial information for final accurate density map predictions. Furthermore, we propose a novel normalization layer called Spatial Group Normalization (SGN), which firstly split feature maps along the spatial dimension and then perform group-wise normalization. It’s useful to solve statistic shift problems caused by the great difference of distribution in crowd counting. Moreover, SGN can be naturally plugged into existing solutions and brings significant improvement in crowd counting. Our proposed SNNet achieves state-of-the-art performance on four challenging crowd counting datasets (ShanghaiTech, UCFQNRF, GCC and TRANCOS datasets), which demonstrates the effectiveness and robust feature learning capability of our methods.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129657417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216994
E. Safronov, Michele Colledanchise, L. Natale
Task planning in a probabilistic belief space generates complex and robust execution policies in domains affected by state uncertainty. The performance of a task planner relies on the belief space representation of the world. However, such representation becomes easily intractable as the number of variables and execution time grow. To address this problem, we developed a novel belief space representation based on the Cartesian product and union operations over belief substates. These two operations and single variable assignment nodes form And-Or directed acyclic graph of Belief States (AOBSs). We show how to apply actions with probabilistic outcomes and how to measure the probability of conditions holding true over belief states. We evaluated AOBSs performance in simulated forward state space exploration. We compared the size of AOBSs with the size of Binary Decision Diagrams (BDDs) that were previously used to represent belief state. We show that AOBSs representation more compact than a full belief state and it scales better than BDDs for most of the cases.
{"title":"Compact Belief State Representation for Task Planning","authors":"E. Safronov, Michele Colledanchise, L. Natale","doi":"10.1109/CASE48305.2020.9216994","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216994","url":null,"abstract":"Task planning in a probabilistic belief space generates complex and robust execution policies in domains affected by state uncertainty. The performance of a task planner relies on the belief space representation of the world. However, such representation becomes easily intractable as the number of variables and execution time grow. To address this problem, we developed a novel belief space representation based on the Cartesian product and union operations over belief substates. These two operations and single variable assignment nodes form And-Or directed acyclic graph of Belief States (AOBSs). We show how to apply actions with probabilistic outcomes and how to measure the probability of conditions holding true over belief states. We evaluated AOBSs performance in simulated forward state space exploration. We compared the size of AOBSs with the size of Binary Decision Diagrams (BDDs) that were previously used to represent belief state. We show that AOBSs representation more compact than a full belief state and it scales better than BDDs for most of the cases.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128999907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9217035
B. Yan, Mikhail A. Bragin, P. Luh
Job shops are an important production environment for low-volume high-variety manufacturing. When there are urgent orders, the speeds of certain machines can be adjusted with a high energy and wear and tear cost. Scheduling in such an environment is to achieve on-time deliveries and low energy costs. The problem is, however, complicated because part processing time depends on machine speeds, and machines need to be modeled individually to capture energy costs. This paper is to obtain near-optimal solutions efficiently. The problem is formulated as a Mixed-Integer Linear Programming (MILP) form to make effective use of available MILP methods. This is done by modeling machines in groups for simplicity while approximating energy costs, and by linking part processing status and machine speed variables. Nevertheless, the resulting problem is still complicated. The formulation is therefore transformed by extending our previous tightening approach for machines with constant speeds. The idea is that if constraints can be transformed to directly delineate the convex hull, then the problem can be solved by linear programming methods. To solve the problem efficiently, our advanced decomposition and coordination method is used. Numerical results show that nearoptimal solutions are obtained, demonstrating significant benefits of our approach on on-time deliveries and energy costs.
{"title":"Tightened Formulation and Resolution of Energy-Efficient Job-Shop Scheduling","authors":"B. Yan, Mikhail A. Bragin, P. Luh","doi":"10.1109/CASE48305.2020.9217035","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9217035","url":null,"abstract":"Job shops are an important production environment for low-volume high-variety manufacturing. When there are urgent orders, the speeds of certain machines can be adjusted with a high energy and wear and tear cost. Scheduling in such an environment is to achieve on-time deliveries and low energy costs. The problem is, however, complicated because part processing time depends on machine speeds, and machines need to be modeled individually to capture energy costs. This paper is to obtain near-optimal solutions efficiently. The problem is formulated as a Mixed-Integer Linear Programming (MILP) form to make effective use of available MILP methods. This is done by modeling machines in groups for simplicity while approximating energy costs, and by linking part processing status and machine speed variables. Nevertheless, the resulting problem is still complicated. The formulation is therefore transformed by extending our previous tightening approach for machines with constant speeds. The idea is that if constraints can be transformed to directly delineate the convex hull, then the problem can be solved by linear programming methods. To solve the problem efficiently, our advanced decomposition and coordination method is used. Numerical results show that nearoptimal solutions are obtained, demonstrating significant benefits of our approach on on-time deliveries and energy costs.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129032960","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216979
Lening Wang, Yutong Zhang, Xiaoyu Chen, R. Jin
Smart manufacturing enables real-time data streaming from interconnected manufacturing processes to improve manufacturing quality, throughput, flexibility, and cost reduction via computation services. In these computation services, machine learning pipelines integrate various types of computation method options to match the contextualized, on-demand computation needs for the maximum prediction accuracy or the best model structure interpretation. On the other hand, there is a pressing need to integrate Fog computing in manufacturing, which will reduce communication time latency and dependency on connections, improve responsiveness and reliability of the computation services, and maintain data privacy. However, there is a knowledge gap in using machine learning pipelines in Fog manufacturing. Existing offloading strategies are not effective, due to the lack of accurate prediction model for the performance of computation services before the execution of those heterogeneous computation tasks. In this paper, machine learning pipelines are implemented in Fog manufacturing. The computation performance of each sub-step of pipelines is predicted and analyzed via linear regression models and random forest regression models. A Fog manufacturing testbed is adopted to validate the performance of the employed models. The results show that the models can adequately predict the performance of computation services, which can be further integrated into Fog manufacturing to better support offloading strategies for machine learning pipelines.
{"title":"Online Computation Performance Analysis for Distributed Machine Learning Pipelines in Fog Manufacturing","authors":"Lening Wang, Yutong Zhang, Xiaoyu Chen, R. Jin","doi":"10.1109/CASE48305.2020.9216979","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216979","url":null,"abstract":"Smart manufacturing enables real-time data streaming from interconnected manufacturing processes to improve manufacturing quality, throughput, flexibility, and cost reduction via computation services. In these computation services, machine learning pipelines integrate various types of computation method options to match the contextualized, on-demand computation needs for the maximum prediction accuracy or the best model structure interpretation. On the other hand, there is a pressing need to integrate Fog computing in manufacturing, which will reduce communication time latency and dependency on connections, improve responsiveness and reliability of the computation services, and maintain data privacy. However, there is a knowledge gap in using machine learning pipelines in Fog manufacturing. Existing offloading strategies are not effective, due to the lack of accurate prediction model for the performance of computation services before the execution of those heterogeneous computation tasks. In this paper, machine learning pipelines are implemented in Fog manufacturing. The computation performance of each sub-step of pipelines is predicted and analyzed via linear regression models and random forest regression models. A Fog manufacturing testbed is adopted to validate the performance of the employed models. The results show that the models can adequately predict the performance of computation services, which can be further integrated into Fog manufacturing to better support offloading strategies for machine learning pipelines.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"794 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117038362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216828
Jimmy Tjen, Francesco Smarra, A. D’innocenzo
In this paper an experimental setup for structural damage detection is considered and a novel sensor selection algorithm is derived, based on the concepts of entropy and information gain from information theory, to reduce the number of sensors without affecting, or even improving (as happens in our experimental setup), model accuracy. An experimental dataset is considered showing that our method outperforms previous approaches improving the prediction accuracy and the damage detection sensitivity while reducing the number of sensors.
{"title":"An entropy-based sensor selection algorithm for structural damage detection","authors":"Jimmy Tjen, Francesco Smarra, A. D’innocenzo","doi":"10.1109/CASE48305.2020.9216828","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216828","url":null,"abstract":"In this paper an experimental setup for structural damage detection is considered and a novel sensor selection algorithm is derived, based on the concepts of entropy and information gain from information theory, to reduce the number of sensors without affecting, or even improving (as happens in our experimental setup), model accuracy. An experimental dataset is considered showing that our method outperforms previous approaches improving the prediction accuracy and the damage detection sensitivity while reducing the number of sensors.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121419320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9217028
Yuan-Chih Peng, Devavrat Jivani, R. Radke, J. Wen
This paper considers image-guided assembly for large composite panels. By using fiducial markers on the panels and robot gripper mounted cameras, we are able to use an industrial robot to align the panels to sub-millimeter accuracy. We considered two commonly used visual servoing schemes: position-based visual servoing (PBVS) and image-based visual servoing (IBVS). It has been noted that IBVS possesses superior robustness with respect to the camera calibration accuracy. However, we have found that in our case, PBVS is both faster and slightly more accurate than IBVS. This result is due to the fact that the visual servoing target in the image plane is derived from a reference target, which depends on the accuracy of the camera model. This additional dependency essentially nullifies the robustness advantage of IBVS. We also implemented a simple scheme to combine inputs from multiple cameras to improve the visual servoing accuracy. Both simulation and experimental results are included to show the effectiveness of visual servoing in an industrial setting.
{"title":"Comparing Position- and Image-Based Visual Servoing for Robotic Assembly of Large Structures","authors":"Yuan-Chih Peng, Devavrat Jivani, R. Radke, J. Wen","doi":"10.1109/CASE48305.2020.9217028","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9217028","url":null,"abstract":"This paper considers image-guided assembly for large composite panels. By using fiducial markers on the panels and robot gripper mounted cameras, we are able to use an industrial robot to align the panels to sub-millimeter accuracy. We considered two commonly used visual servoing schemes: position-based visual servoing (PBVS) and image-based visual servoing (IBVS). It has been noted that IBVS possesses superior robustness with respect to the camera calibration accuracy. However, we have found that in our case, PBVS is both faster and slightly more accurate than IBVS. This result is due to the fact that the visual servoing target in the image plane is derived from a reference target, which depends on the accuracy of the camera model. This additional dependency essentially nullifies the robustness advantage of IBVS. We also implemented a simple scheme to combine inputs from multiple cameras to improve the visual servoing accuracy. Both simulation and experimental results are included to show the effectiveness of visual servoing in an industrial setting.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121256943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9216833
Ho-Jun Park, Jang-Woon Baek, Jong-Hwan Kim
This paper presents a real-time approach to count push-ups using 2D video imagery. The proposed method uses OpenPose in each frame to extract multiple joints and links of a human body. Then, it analyzes key motion features linked to counting the push-ups. Taking in consideration the push-up rules of the Republic of Korea Army, five criteria are defined and used parametrically to discriminate both correct and incorrect push-ups. A total of 147,840 samples have been collected from 220 push-up videos each in two different viewpoints: half of the videos for modeling the proposed method and the other half for testing its performance. Finally, the results shows 90.00%, 87.82%, 97.86%, and 92.57% for accuracy, precision, recall, and F-measure, respectively, demonstrating its reliability in military physical tests.
{"title":"Imagery based Parametric Classification of Correct and Incorrect Motion for Push-up Counter Using OpenPose","authors":"Ho-Jun Park, Jang-Woon Baek, Jong-Hwan Kim","doi":"10.1109/CASE48305.2020.9216833","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9216833","url":null,"abstract":"This paper presents a real-time approach to count push-ups using 2D video imagery. The proposed method uses OpenPose in each frame to extract multiple joints and links of a human body. Then, it analyzes key motion features linked to counting the push-ups. Taking in consideration the push-up rules of the Republic of Korea Army, five criteria are defined and used parametrically to discriminate both correct and incorrect push-ups. A total of 147,840 samples have been collected from 220 push-up videos each in two different viewpoints: half of the videos for modeling the proposed method and the other half for testing its performance. Finally, the results shows 90.00%, 87.82%, 97.86%, and 92.57% for accuracy, precision, recall, and F-measure, respectively, demonstrating its reliability in military physical tests.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115370301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-01DOI: 10.1109/CASE48305.2020.9217044
Bai Li, Qi Kong, Youmin Zhang, Zhijiang Shao, Yumeng Wang, Xiaoyan Peng, Daxun Yan
On-road trajectory planning is a critical module in an autonomous driving system. Instead of using a path-velocity decomposition or longitudinal-lateral decomposition strategy, this work aims to find a trajectory directly. We adopt a sampleand-search planner to get a coarse trajectory and then polish it via numerical optimization. Among the predominant sampleand-search planners, most of the sampling operations are not flexible, which inevitably lead to a solution failure if the sampling density is low, and suffer from the curse of dimensionality if the sampling density is set high. This work proposes a modified RRT* for trajectory search, aiming to promote the sampling flexibility and to get rid of the search randomness. A quadratic program (QP) based smoother is proposed to refine the coarse trajectory. Herein, the scale of the QP problem is fixed and tractable, and the feasibility of the QP problem is always guaranteed.
{"title":"On-road Trajectory Planning with Spatio-temporal RRT* and Always-feasible Quadratic Program","authors":"Bai Li, Qi Kong, Youmin Zhang, Zhijiang Shao, Yumeng Wang, Xiaoyan Peng, Daxun Yan","doi":"10.1109/CASE48305.2020.9217044","DOIUrl":"https://doi.org/10.1109/CASE48305.2020.9217044","url":null,"abstract":"On-road trajectory planning is a critical module in an autonomous driving system. Instead of using a path-velocity decomposition or longitudinal-lateral decomposition strategy, this work aims to find a trajectory directly. We adopt a sampleand-search planner to get a coarse trajectory and then polish it via numerical optimization. Among the predominant sampleand-search planners, most of the sampling operations are not flexible, which inevitably lead to a solution failure if the sampling density is low, and suffer from the curse of dimensionality if the sampling density is set high. This work proposes a modified RRT* for trajectory search, aiming to promote the sampling flexibility and to get rid of the search randomness. A quadratic program (QP) based smoother is proposed to refine the coarse trajectory. Herein, the scale of the QP problem is fixed and tractable, and the feasibility of the QP problem is always guaranteed.","PeriodicalId":212181,"journal":{"name":"2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115435716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：