With the development of artificial intelligence technology, more and more parks have realized automated delivery by unmanned vehicles or deployed self-driving sightseeing vehicles. Research on the control and optimization of traffic flow in the park is of great help to improve the efficiency of logistics vehicles in the park and the rationality of tourist vehicle routes. Because of the complexity of the traffic environment in the park and the uncertainty of personnel and obstacles, it is difficult to use conventional mathematical models to summarize and optimize the analysis of rules. In order to solve this problem, this paper proposes a logistics vehicle scheduling algorithm based on PPO. Through the reinforcement learning method driven by data, the traffic flow of intelligent agents in the park is optimized and controlled to achieve path planning and intelligent speed regulation. In this paper, the scenario of Lanzhou University campus is constructed in the SUMO simulator and simulation experiments are carried out. Compared with the traditional path planning algorithm, the logistics vehicle scheduling algorithm has a great improvement in time efficiency.
Build Simulation Platform in Real Logistics Scenario and Optimization Based on Reinforcement Learning
2022-05-04
2066349 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Automated reinforcement learning scenario variation and impact penalties
Europäisches Patentamt | 2024
|Multi-agent Reinforcement Learning-Based UAS Control for Logistics Environments
Springer Verlag | 2022
|