为了提高城市交叉口通行效率,在车路协同环境下对无信号交叉口优化控制方法进行了研究。设计并采用基于时延Petri网(Timed Petri Net,TdPN)的无信号交叉口优化控制方法,利用TdPN建立无信号交叉口控制模型,并依此建立交叉口车辆最快消散...为了提高城市交叉口通行效率,在车路协同环境下对无信号交叉口优化控制方法进行了研究。设计并采用基于时延Petri网(Timed Petri Net,TdPN)的无信号交叉口优化控制方法,利用TdPN建立无信号交叉口控制模型,并依此建立交叉口车辆最快消散目标函数,采用递归方式求解车辆最优通过序列;利用Q-Paramics构建基于车路协同环境下的无信号交叉口仿真平台,分析该方法在不同交通流量下对交叉口平均延迟、平均停车次数、平均排队长度和平均速度4个交通参数的影响,并将其与传统的信号控制方法进行对比。研究结果表明:基于TdPN的无信号优化控制方法能够在一定程度上缓解中小交通流量下的交叉口通行问题,并且其控制结果明显优于传统的感应控制方法。展开更多
The goal of this paper is to propose a unique control method that permits the evolution of both timed continuous Petri net （TCPN） and T-timed discrete Petri net （T-TDPN） from an initial state to a desired one. Mod...The goal of this paper is to propose a unique control method that permits the evolution of both timed continuous Petri net （TCPN） and T-timed discrete Petri net （T-TDPN） from an initial state to a desired one. Model predictive control （MPC） is a robust control scheme against perturbation and a consistent real-time constraints method. Hence, the proposed approach is studied using the MPC. However, the computational complexity may prevent the use of the MPC for large systems and for large prediction horizons. Then, the proposed approach provides some new techniques in order to reduce the high computational complexity; among them one is taking constant control actions during the prediction.展开更多
文摘为了提高城市交叉口通行效率,在车路协同环境下对无信号交叉口优化控制方法进行了研究。设计并采用基于时延Petri网(Timed Petri Net,TdPN)的无信号交叉口优化控制方法,利用TdPN建立无信号交叉口控制模型,并依此建立交叉口车辆最快消散目标函数,采用递归方式求解车辆最优通过序列;利用Q-Paramics构建基于车路协同环境下的无信号交叉口仿真平台,分析该方法在不同交通流量下对交叉口平均延迟、平均停车次数、平均排队长度和平均速度4个交通参数的影响,并将其与传统的信号控制方法进行对比。研究结果表明:基于TdPN的无信号优化控制方法能够在一定程度上缓解中小交通流量下的交叉口通行问题,并且其控制结果明显优于传统的感应控制方法。
基金supported by the region Haute-Normandie Project(Nos.CPER-SER-DDSMRI 2013,2014 and CPER-SER-SEL 2015)
文摘The goal of this paper is to propose a unique control method that permits the evolution of both timed continuous Petri net （TCPN） and T-timed discrete Petri net （T-TDPN） from an initial state to a desired one. Model predictive control （MPC） is a robust control scheme against perturbation and a consistent real-time constraints method. Hence, the proposed approach is studied using the MPC. However, the computational complexity may prevent the use of the MPC for large systems and for large prediction horizons. Then, the proposed approach provides some new techniques in order to reduce the high computational complexity; among them one is taking constant control actions during the prediction.
