| With the continuous expansion of the influence range of artificial intelligence,intelligent transportation has become the key development direction of the future transportation field,and the unmanned driving technology is one of the important manifestations.In order to meet the increasing demands of users,improve the efficiency of unmanned travel tasks,improve the rationality of route selection,respond to complex dynamic traffic environments,ensure safety,and accelerate the intelligent process of actual transportation operations.In this paper,combined with the overall trend of the development of unmanned driving,the path planning module is studied with unmanned driving technical support as the background,and an improved optimization method is proposed.Specific research contents include the following:1.Investigate the overall development of unmanned driving technology and the current research status of path planning,introduce the basic composition of unmanned driving system architecture,and research and design the key technical architecture of unmanned driving in conjunction with the application of path planning.Subsequent path planning module research provides corresponding technical support.2.The path planning of driverless cars faces a complex and changeable traffic environment.In order to more comprehensively evaluate the select indicators of path to plan a more reasonable path,and better solve the impact of dynamic changes of road section environment on planning results.A dynamic path planning algorithm-RDMA*(Real-time Dynamics of Multiple influencing factors AStar)algorithm is considered.With A*(AStar)algorithm as the core,use two-way search mode and sorting based on heap structure,improve the cost function by adding traffic evaluation factors with multiple influencing factors,comprehensive consider the distance,traffic congestion,road smoothness and other influencing factors,use AHP to determine the relative weight of each influencing factor,take the comprehensivecost value as the evaluation index for path planning;Through GPS,radar,camera and other equipment,using fusion sensing technology to obtain relevant road environment information.According to the obtained global and local traffic environment data information,using real-time dynamic update strategy to solve the path planning problem in dynamic environment and plan the optimal path in real time.3.Combined with the RDMA* path planning algorithm,it is applied to the actual transportation scheduling tasks of unmanned vehicles.Based on the comprehensive generation value of path selection and transportation penalty costs,a mathematical model for multiobjective-multi-target unmanned vehicle transportation scheduling has been proposed.The genetic algorithm with improved multi-layer coding is used to solve the problem,and the global optimal unmanned vehicle transportation scheduling scheme between target points is obtained.4.Based on the Prescan simulation platform,the simulation realizes the unmanned driving technology architecture and related configuration.Based on the transportation background of cold chain logistics in the Pearl River Delta region,a case was constructed to verify the effectiveness of the RDMA* algorithm.The results show that the RDMA* path planning method can obtain a lower total travel time than traditional path planning methods and improve the efficiency of travel tasks.Moreover,under the condition of encountering special events in the dynamic traffic environment,it can provide an optimal feasible path with the minimum comprehensive substitute value and the least time for the driverless vehicle dynamics.Finally,by constructing a multi-target supplier-multi-target demand supplier’s transportation task scenario,on the premise of RDMA* path planning,the improved multi-layer coding genetic algorithm is used to solve the problem of transportation scheduling between multi-target points,and it is demonstrated that this method can obtain the overall situation Optimal solution and maintain a fast search efficiency. |
PDF Full Text Request