The invention relates to the technical field of intelligent transportation, in particular to an unmanned bus dispatching simulation system based on reinforcement learning, which comprises a cloud dispatching simulation system, a station system, a station system, a vehicle end system and a 4G/5G mobile communication network, and the station system, the field station system and the vehicle end system communicate with the cloud scheduling simulation system through the 4G/5G mobile communication network. Compared with a traditional method, the technical scheme is based on the DQN algorithm, the deep reinforcement learning simulation model is constructed, the purpose of the scheduling system is to improve the score of the reward function due to different passenger appearing and destinations in the environment, so that the scheduling system does not form a fixed vehicle scheduling route, and the vehicle scheduling efficiency is improved. The optimal reward function scores corresponding to different passenger appearing places and destinations and corresponding implementation modes do not have fixed ways, the advantages of a scheduling algorithm based on reinforcement learning are embodied, the operation cost is reduced, and the average passenger carrying rate and the comprehensive operation index are improved in the process of reducing cost and increasing efficiency.
本发明涉及智能交通技术领域,具体涉及一种基于强化学习的无人巴士的调度仿真系统,包括云端调度仿真系统、站点系统、场站系统、车端系统以及4G/5G移动通讯网络,所述站点系统、所述场站系统、所述车端系统均通过所述4G/5G移动通讯网络与所述云端调度仿真系统通讯。相较于传统方法,本技术方案基于DQN算法,构建深度强化学习仿真模型,由于环境中的乘客出现及目的地不同,调度系统的目的为提高奖励函数得分,因此调度系统不会形成固定的车辆调度路线,不同的乘客出现地点及目的地对应的最佳奖励函数分数及对应的实现方式无固定途径,体现了基于强化学习的调度算法的优势,降低运营成本,在降本增效的过程中,提高平均载客率,提高综合运营指标。
Unmanned bus scheduling simulation system based on reinforcement learning
一种基于强化学习的无人巴士的调度仿真系统
2023-12-22
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06Q Datenverarbeitungssysteme oder -verfahren, besonders angepasst an verwaltungstechnische, geschäftliche, finanzielle oder betriebswirtschaftliche Zwecke, sowie an geschäftsbezogene Überwachungs- oder Voraussagezwecke , DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES |
AIRCRAFT SCHEDULING SIMULATION SYSTEM USING REINFORCEMENT LEARNING AND METHOD THEREOF
European Patent Office | 2024
|Reinforcement Learning in Scheduling
British Library Conference Proceedings | 1993
|Reinforcement learning in scheduling
NTRS | 1994
|Reinforcement Learning Based Communication Security for Unmanned Aerial Vehicles
Springer Verlag | 2021
|