Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560503
Meng Yu, X. Zuo, Xinchao Zhao, Chunlu Wang
Single row layout problem is widely implemented in manufacturing systems. In this paper, we study a single row layout problem with consideration of additional clearances between adjacent facilities. The additional clearance is used to store work in process and allow technician access to the side of the facility and may be shared by any two adjacent facilities. A hybrid solution approach that integrates a tabu search with a mathematical programming is proposed to address it. The purpose of tabu search is to optimize the sequence of facilities and mathematical programming is used to find the optimal additional clearances. The hybrid approach is applied to a number of problem instances involving a range of facilities. Experimental results show that the hybrid approach can find the optimal solutions for most of small-size problem instances and good quality solutions for large-size problem instances.
{"title":"Hybridizing tabu search with mathematical programming for solving a single row layout problem","authors":"Meng Yu, X. Zuo, Xinchao Zhao, Chunlu Wang","doi":"10.1109/COASE.2018.8560503","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560503","url":null,"abstract":"Single row layout problem is widely implemented in manufacturing systems. In this paper, we study a single row layout problem with consideration of additional clearances between adjacent facilities. The additional clearance is used to store work in process and allow technician access to the side of the facility and may be shared by any two adjacent facilities. A hybrid solution approach that integrates a tabu search with a mathematical programming is proposed to address it. The purpose of tabu search is to optimize the sequence of facilities and mathematical programming is used to find the optimal additional clearances. The hybrid approach is applied to a number of problem instances involving a range of facilities. Experimental results show that the hybrid approach can find the optimal solutions for most of small-size problem instances and good quality solutions for large-size problem instances.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"27 1","pages":"974-980"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84832257","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560482
Chao-Bo Yan, Xingrui Cheng, F. Gao, X. Guan
Machines consume intensive energy in some production systems. Although substantial efforts have been devoted to performance analysis, continuous improvement, and design of production systems, the research on reduction of energy consumption in these systems is limited. In this paper, we formulate a nonlinear programming problem with production rate constraint to minimize machines' energy consumption. Structural characteristics of the problem and optimality conditions are analyzed to establish two optimality equations. An effective algorithm based on binary search method is developed to solve the optimal solution from the optimality equations.
{"title":"Energy Consumption Optimization in Two-Machine Bernoulli Serial Lines","authors":"Chao-Bo Yan, Xingrui Cheng, F. Gao, X. Guan","doi":"10.1109/COASE.2018.8560482","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560482","url":null,"abstract":"Machines consume intensive energy in some production systems. Although substantial efforts have been devoted to performance analysis, continuous improvement, and design of production systems, the research on reduction of energy consumption in these systems is limited. In this paper, we formulate a nonlinear programming problem with production rate constraint to minimize machines' energy consumption. Structural characteristics of the problem and optimality conditions are analyzed to establish two optimality equations. An effective algorithm based on binary search method is developed to solve the optimal solution from the optimality equations.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"8 1","pages":"1302-1307"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84162339","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560576
A. Fazlirad, R. Brennan
This paper is an updated review of the state of the art in distributed multiagent based scheduling as one element of intelligent manufacturing. It re-examines the most relevant previous reviews and details open research questions and directions proposed by other researchers. It highlights some of the recent publications in multiagent based scheduling in terms of these previously identified research areas and questions. The paper is concluded by outlining the extent to which recent literature addresses these areas to provide directions for future work.
{"title":"Multiagent Manufacturing Scheduling: An Updated State of the Art Review","authors":"A. Fazlirad, R. Brennan","doi":"10.1109/COASE.2018.8560576","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560576","url":null,"abstract":"This paper is an updated review of the state of the art in distributed multiagent based scheduling as one element of intelligent manufacturing. It re-examines the most relevant previous reviews and details open research questions and directions proposed by other researchers. It highlights some of the recent publications in multiagent based scheduling in terms of these previously identified research areas and questions. The paper is concluded by outlining the extent to which recent literature addresses these areas to provide directions for future work.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"14 1","pages":"722-729"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82059863","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560509
F. Acerbi, G. Nicolao, Josef Obiltschnig, Patrick Richter, Cristina De Luca
The optimal energy management of multiple chiller systems calls for the construction of mathematical models of chiller energy efficiency. The existing grey- or black-box models include parameters that have to be estimated from experimental data. So far, the predictive capabilities of alternative models have been assessed and compared on data sets created by laboratory tests or provided by chiller manufacturers. In an Industry 4.0 context, the continuous monitoring and collection of field data discloses new opportunities but raises also robustness issues that are herein addressed. Herein, exploiting an extensive experimental dataset collected over a six-month period, four literature models and a new machine learning approach are compared. The second objective is assessing the robustness of the five models against covariate shifts, i.e. variations in the statistical distribution of the input variables that occur across different months. The grey-box Gordon-Ng model, though less accurate in nominal conditions than the Bi-quadratic and Multivariate polynomial models, proves however more robust against covariate shifts. The best performances, both in term of accuracy and robustness, are however provided by the two machine learning methods, with the Gaussian Process model performing better than the MLP artificial neural network.
{"title":"Accuracy and Robustness Against Covariate Shift of Water Chiller Models","authors":"F. Acerbi, G. Nicolao, Josef Obiltschnig, Patrick Richter, Cristina De Luca","doi":"10.1109/COASE.2018.8560509","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560509","url":null,"abstract":"The optimal energy management of multiple chiller systems calls for the construction of mathematical models of chiller energy efficiency. The existing grey- or black-box models include parameters that have to be estimated from experimental data. So far, the predictive capabilities of alternative models have been assessed and compared on data sets created by laboratory tests or provided by chiller manufacturers. In an Industry 4.0 context, the continuous monitoring and collection of field data discloses new opportunities but raises also robustness issues that are herein addressed. Herein, exploiting an extensive experimental dataset collected over a six-month period, four literature models and a new machine learning approach are compared. The second objective is assessing the robustness of the five models against covariate shifts, i.e. variations in the statistical distribution of the input variables that occur across different months. The grey-box Gordon-Ng model, though less accurate in nominal conditions than the Bi-quadratic and Multivariate polynomial models, proves however more robust against covariate shifts. The best performances, both in term of accuracy and robustness, are however provided by the two machine learning methods, with the Gaussian Process model performing better than the MLP artificial neural network.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"139 1","pages":"809-816"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81763393","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560553
Nan Tian, Benjamin Kuo, X. Ren, Michael Yu, Robert Zhang, Bill Huang, Ken Goldberg, S. Sojoudi
We present a cloud-based human-robot interaction system that automatically controls a humanoid robot to mirror a human demonstrator performing flag semaphores. We use a cloud-based framework called Human Augmented Robotic Intelligence (HARI) to perform gesture recognition of the human demonstrator and gesture control of a local humanoid robot, named Pepper. To ensure that the system is real-time, we design a system to maximize cloud computation contribution to the deep-neural-network-based gesture recognition system, OpenPose, and to minimize communication costs between the cloud and the robot. A hybrid control system is used to hide latency caused by either routing or physical distances. We conducted real-time semaphore mirroring experiments in which both the robots and the demonstrator were located in Tokyo, Japan, whereas the cloud server was deployed in the United States. The total latency was 400ms for the video streaming to the cloud and 108ms for the robot commanding from the cloud. Further, we measured the reliability of our gesture-based semaphore recognition system with two human subjects, and were able to achieve 90% and 76.7% recognition accuracy, respectively, for the two subjects with open-loop when the subjects were not allowed to see the recognition results. We could achieve 100% recognition accuracy when both subjects were allowed to adapt to the recognition system under a closed-loop setting. Lastly, we showed that we can support two humanoid robots with a single server at the same time. With this real-time cloud-based HRI system, we illustrate that we can deploy gesture-based human-robot globally and at scale.
{"title":"A Cloud-Based Robust Semaphore Mirroring System for Social Robots","authors":"Nan Tian, Benjamin Kuo, X. Ren, Michael Yu, Robert Zhang, Bill Huang, Ken Goldberg, S. Sojoudi","doi":"10.1109/COASE.2018.8560553","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560553","url":null,"abstract":"We present a cloud-based human-robot interaction system that automatically controls a humanoid robot to mirror a human demonstrator performing flag semaphores. We use a cloud-based framework called Human Augmented Robotic Intelligence (HARI) to perform gesture recognition of the human demonstrator and gesture control of a local humanoid robot, named Pepper. To ensure that the system is real-time, we design a system to maximize cloud computation contribution to the deep-neural-network-based gesture recognition system, OpenPose, and to minimize communication costs between the cloud and the robot. A hybrid control system is used to hide latency caused by either routing or physical distances. We conducted real-time semaphore mirroring experiments in which both the robots and the demonstrator were located in Tokyo, Japan, whereas the cloud server was deployed in the United States. The total latency was 400ms for the video streaming to the cloud and 108ms for the robot commanding from the cloud. Further, we measured the reliability of our gesture-based semaphore recognition system with two human subjects, and were able to achieve 90% and 76.7% recognition accuracy, respectively, for the two subjects with open-loop when the subjects were not allowed to see the recognition results. We could achieve 100% recognition accuracy when both subjects were allowed to adapt to the recognition system under a closed-loop setting. Lastly, we showed that we can support two humanoid robots with a single server at the same time. With this real-time cloud-based HRI system, we illustrate that we can deploy gesture-based human-robot globally and at scale.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"5 1","pages":"1351-1358"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83438704","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560542
Manfred Smieschek, Timo Hinrichs, André Stollenwerk, S. Kowalewski, Rudiger Preub
Heavy-duty slow-rotating roller chains are expensive and time-consuming to replace. Inspections to assess the wear on a chain can only be carried out periodically and usually lead to a standstill of production, which comes with a loss of earnings. Therefore, the incentive to maintain them only when needed, and to use them as long as possible is high. Thus, the use of condition monitoring for roller chains is justified. It can continuously determine the remaining wear margin of the chain so that maintenance can be scheduled when necessary and production holds can be minimized. For slow-rotating roller chains the usual frequency-based approaches are not applicable. We propose the average torque of the driving motor for the roll-in of each chain link as a meaningful condition indicator. The angle and torque can be read out from the motor drive and encoder, respectively, and the average torque can easily be calculated. On data from a highly worn chain of a demonstrator, the roll-in of certain chain links lead to a 25% increased average torque of the driving motor. We expect the average torque to gradually increase with less remaining wear margin.
{"title":"A New Condition Indicator for Slow-Rotating Roller Chains based on the Angle and Torque of the Driving Motor","authors":"Manfred Smieschek, Timo Hinrichs, André Stollenwerk, S. Kowalewski, Rudiger Preub","doi":"10.1109/COASE.2018.8560542","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560542","url":null,"abstract":"Heavy-duty slow-rotating roller chains are expensive and time-consuming to replace. Inspections to assess the wear on a chain can only be carried out periodically and usually lead to a standstill of production, which comes with a loss of earnings. Therefore, the incentive to maintain them only when needed, and to use them as long as possible is high. Thus, the use of condition monitoring for roller chains is justified. It can continuously determine the remaining wear margin of the chain so that maintenance can be scheduled when necessary and production holds can be minimized. For slow-rotating roller chains the usual frequency-based approaches are not applicable. We propose the average torque of the driving motor for the roll-in of each chain link as a meaningful condition indicator. The angle and torque can be read out from the motor drive and encoder, respectively, and the average torque can easily be calculated. On data from a highly worn chain of a demonstrator, the roll-in of certain chain links lead to a 25% increased average torque of the driving motor. We expect the average torque to gradually increase with less remaining wear margin.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"24 1","pages":"642-644"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83956300","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560479
S. Cho
Recently, impulse radio-ultra wideband is widely used for highly accurate location estimation of robots and objects in indoor space for building automation. For their accurate positioning, calibration must be performed after the positioning system is installed, and a compensation technique of the multipath signals should be performed in real-time. In this paper, a real-time calibration and error compensation method is proposed. A filter that estimates channel common error and channel-specific error sequentially is designed based on cubature Kalman filter and the fast and accurate estimation performance of this filter is verified experimentally.
{"title":"CKF-based Fast Error Compensation Filter Design for IR-UWB Indoor Positioning System for Building Automation","authors":"S. Cho","doi":"10.1109/COASE.2018.8560479","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560479","url":null,"abstract":"Recently, impulse radio-ultra wideband is widely used for highly accurate location estimation of robots and objects in indoor space for building automation. For their accurate positioning, calibration must be performed after the positioning system is installed, and a compensation technique of the multipath signals should be performed in real-time. In this paper, a real-time calibration and error compensation method is proposed. A filter that estimates channel common error and channel-specific error sequentially is designed based on cubature Kalman filter and the fast and accurate estimation performance of this filter is verified experimentally.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"30 1","pages":"668-670"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89163548","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560358
Ishara Paranawithana, Hsieh-Yu Li, S. Foong, U-Xuan Tan, Liangjing Yang, Terence Sey Kiat Lim, F. Ng
Percutaneous Nephrolithotomy (PCNL) is a minimally invasive percutaneous surgical procedure used for large kidney stone removal under ultrasound and fluoroscopy guidance. During the surgery, precise control of handheld 2D ultrasound probe is required but highly challenging as it depends on operator's experience, judgement and dexterity. To complicate the problem, kidney stone moves away from its 2D ultrasound image plane due to respiratory movement of the patient. This makes locating the kidney stone extremely challenging, if not impossible, further limiting the success of the initial needle puncture. Therefore, there is a need to bring automation to the intraoperative workflow to compensate out-of-plane motion of the kidney stone. Maintaining simultaneous control of appropriate contact force during visual tracking is also essential to ensure accurate percutaneous access to the target calyx. This work proposes a visual servoing framework to address the aforesaid problems. Our proposed visual servoing framework comes in the form of two stages namely; pre-scan and realtime visual servoing. Probe holding robotic manipulator firstly scans a small region around the target to construct 3D volume data, followed by out-of-plane target tracking using image correlation-based block matching algorithm. A position based admittance control scheme is developed to address the latent need of maintaining an appropriate contact force between the probe and patient's body during visual servoing. Experimental results show that proposed framework is able to track out-of-plane motion of kidney stone with a position error of only one frame while regulating the environment force feedback with a maximum error of 0.2N. By incorporating automation to existing surgical workflow, we hope to positively impact the way minimally invasive surgeries are performed.
{"title":"Ultrasound-Guided Involuntary Motion Compensation of Kidney Stones in Percutaneous Nephrolithotomy Surgery","authors":"Ishara Paranawithana, Hsieh-Yu Li, S. Foong, U-Xuan Tan, Liangjing Yang, Terence Sey Kiat Lim, F. Ng","doi":"10.1109/COASE.2018.8560358","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560358","url":null,"abstract":"Percutaneous Nephrolithotomy (PCNL) is a minimally invasive percutaneous surgical procedure used for large kidney stone removal under ultrasound and fluoroscopy guidance. During the surgery, precise control of handheld 2D ultrasound probe is required but highly challenging as it depends on operator's experience, judgement and dexterity. To complicate the problem, kidney stone moves away from its 2D ultrasound image plane due to respiratory movement of the patient. This makes locating the kidney stone extremely challenging, if not impossible, further limiting the success of the initial needle puncture. Therefore, there is a need to bring automation to the intraoperative workflow to compensate out-of-plane motion of the kidney stone. Maintaining simultaneous control of appropriate contact force during visual tracking is also essential to ensure accurate percutaneous access to the target calyx. This work proposes a visual servoing framework to address the aforesaid problems. Our proposed visual servoing framework comes in the form of two stages namely; pre-scan and realtime visual servoing. Probe holding robotic manipulator firstly scans a small region around the target to construct 3D volume data, followed by out-of-plane target tracking using image correlation-based block matching algorithm. A position based admittance control scheme is developed to address the latent need of maintaining an appropriate contact force between the probe and patient's body during visual servoing. Experimental results show that proposed framework is able to track out-of-plane motion of kidney stone with a position error of only one frame while regulating the environment force feedback with a maximum error of 0.2N. By incorporating automation to existing surgical workflow, we hope to positively impact the way minimally invasive surgeries are performed.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"48 1","pages":"1123-1129"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88520947","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560578
Zhengcai Cao, Sikai Gong, Meng Zhou, Kaiwen Liu
The bi-objective reentrant hybrid flow shop problem (BRHFSP) is a typical NP-hard scheduling case in a semiconductor wafer fabrication. In this paper, a self-braking Symbiotic Organisms Search algorithm (SSOS) is proposed to minimize the total tardiness and makespan of this problem. A discrete multi-objective Symbiotic Organisms Search is selected to reduce the wasted time of adjusting excessive parameters in most evolutionary algorithms. This algorithm has a brief structure with no control parameters. Moreover, the entropy-based termination criterion is added to multi-objective Symbiotic Organisms Search to decrease the computation burden. In this way, an entropy-based dissimilarity measure criterion is generated to help our algorithm stop automatically with the increase of iterations. Numerical test results in many cases demonstrate that SSOS is effective for BRHFSP.
{"title":"A Self-braking Symbiotic Organisms Search Algorithm for Bi-objective Reentrant Hybrid Flow Shop Scheduling Problem","authors":"Zhengcai Cao, Sikai Gong, Meng Zhou, Kaiwen Liu","doi":"10.1109/COASE.2018.8560578","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560578","url":null,"abstract":"The bi-objective reentrant hybrid flow shop problem (BRHFSP) is a typical NP-hard scheduling case in a semiconductor wafer fabrication. In this paper, a self-braking Symbiotic Organisms Search algorithm (SSOS) is proposed to minimize the total tardiness and makespan of this problem. A discrete multi-objective Symbiotic Organisms Search is selected to reduce the wasted time of adjusting excessive parameters in most evolutionary algorithms. This algorithm has a brief structure with no control parameters. Moreover, the entropy-based termination criterion is added to multi-objective Symbiotic Organisms Search to decrease the computation burden. In this way, an entropy-based dissimilarity measure criterion is generated to help our algorithm stop automatically with the increase of iterations. Numerical test results in many cases demonstrate that SSOS is effective for BRHFSP.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"7 1","pages":"803-808"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89034556","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 : 2018-08-01DOI: 10.1109/COASE.2018.8560386
Hankun Zhao, Andrew Cui, Schuyler A. Cullen, B. Paden, Michael Laskey, Ken Goldberg
To facilitate automation of urban driving, we present an efficient, lightweight, open-source, first-order simulator with associated graphical display and algorithmic supervisors. FLUIDS can efficiently simulate traffic intersections with varying state configurations for the training and evaluation of learning algorithms. FLUIDS supports an image-based birdseye state space and a lower dimensional quasi-LIDAR representation. FLUIDS additionally provides algorithmic supervisors for simulating realistic behavior of pedestrians and cars in the environment. FLUIDS generates data in parallel at 4000 state-action pairs per minute and evaluates in parallel an imitation learned policy at 20K evaluations per minute. A velocity controller for avoiding collisions and obeying traffic laws using imitation learning was learned from demonstration. We additionally demonstrate the flexibility of FLUIDS by reporting an extensive sensitivity analysis of the learned model to simulation parameters. FLUIDS 1.0 is available at https://berkeleyautomation.github.io/Urban_Driving_Simulator/.
{"title":"FLUIDS: A First-Order Lightweight Urban Intersection Driving Simulator","authors":"Hankun Zhao, Andrew Cui, Schuyler A. Cullen, B. Paden, Michael Laskey, Ken Goldberg","doi":"10.1109/COASE.2018.8560386","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560386","url":null,"abstract":"To facilitate automation of urban driving, we present an efficient, lightweight, open-source, first-order simulator with associated graphical display and algorithmic supervisors. FLUIDS can efficiently simulate traffic intersections with varying state configurations for the training and evaluation of learning algorithms. FLUIDS supports an image-based birdseye state space and a lower dimensional quasi-LIDAR representation. FLUIDS additionally provides algorithmic supervisors for simulating realistic behavior of pedestrians and cars in the environment. FLUIDS generates data in parallel at 4000 state-action pairs per minute and evaluates in parallel an imitation learned policy at 20K evaluations per minute. A velocity controller for avoiding collisions and obeying traffic laws using imitation learning was learned from demonstration. We additionally demonstrate the flexibility of FLUIDS by reporting an extensive sensitivity analysis of the learned model to simulation parameters. FLUIDS 1.0 is available at https://berkeleyautomation.github.io/Urban_Driving_Simulator/.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"29 1","pages":"697-704"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81705025","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：