Research On Multi-vehicle Dynamic And Static Mission Planning Strateg

With the rapid development of artificial intelligence and intelligent driving technology,the research on intelligent vehicle and robot has become more mature.The leap from single intelligence to group intelligence is the inevitable trend of future development.When there are a large number of tasks in the multi-vehicle system,reasonable task allocation is a key link of the multi-vehicle system.Considering the problem of task allocation,this paper carries out the following research work:Firstly,a special multi-vehicle task allocation problem was proposed based on the actual requirements.The objective was to assign the optimal route for each vehicle,and the problem was expressed as a multi-objective electric vehicle routing problem with time windows and cooperative tasks.A multi-objective integer programming mathematical model was established,with the total time to complete tasks and total penalty value as the optimization objective.After developing the research route,the hierarchical experimental system of task allocation-path tracking is set up,and the specific contents and functions of decision planning layer,tracking control layer and remote communication layer are described.Secondly,the 8-neighborhood extension A* algorithm is used to carry out path planning based on high-density grid map.Then the curve smoothing process is carried out based on Bessel curve,and the obstacle boundary extension strategy is proposed to avoid collision between curve and obstacles.The path tracking controller was designed based on pure pursuit algorithm,and the strategy of adaptive previewing distance and speed was proposed to enhance the tracking ability of turning conditions.The performance of the path planning and tracking algorithm is verified by experiments.Next,a hybrid non-dominated sorting genetic algorithm-II based on variable neighborhood search is proposed to solve the task allocation mathematical model in static environment.The algorithm combines the local search strategy of variable neighborhood search algorithm,puts forward the feasible recovery strategy for the phenomenon of infeasible solutions flooding in the population,and introduces the concept of immigrant population.Simulation and experimental results show that this algorithm can solve the problem effectively and has good robustness in different complexity scenarios.Finally,considering the possible environmental changes in a dynamic environment,a dynamic task reallocation strategy based on tabu search algorithm is proposed.The strategy changes the dynamic task reallocation problem into sub-route optimization problem,based on which the algorithm flow under different changes is established.The feasibility and rationality of the reallocation strategy is verified by applying it to the actual multi-vehicle experimental system.