Research On Unmanned Vehicle Path Planning Based On Improved Artificial Potential Field And Bat Algorithm

Unmanned vehicles,also called intelligent mobile robots,generate a collision-free efficient path from the starting point to the target location through the study of optimal path planning for unmanned vehicles to achieve the purpose of improving traffic accidents,reducing traffic congestion,reducing resource occupancy,and improving vehicle operation efficiency.In solving the optimal path planning problem,the bat algorithm has disadvantages such as easy to fall into local extrema,slow convergence speed and low convergence accuracy.Therefore,it is necessary to integrate the bat algorithm with other algorithms and further refine and improve the basic formulation of the bat algorithm.The main work of this paper are as follows.(1)Research on unmanned vehicle path planning based on improved bat algorithm.To address the problem that the bat algorithm will fall into the local optimal solution too early,this paper proposes to introduce a weighted speed update factor into the bat algorithm speed update formula,so as to balance the performance of global exploration and local exploitation of the bat population and avoid the algorithm to conduct too much local search.In the iterative process,the global difference variation strategy is added to enhance its global reconnaissance ability;and the local difference strategy is added to enhance the local search ability and improve the optimization search accuracy to maintain the diversity of the population.(2)Research on unmanned vehicle path planning based on improved artificial potential field algorithm.For the problems that the artificial potential field method is prone to local optimal solutions and unreachable targets,this paper proposes a strategy of repulsive potential field angle function to change the repulsive angle under the action of double repulsive function,so that the direction of the combined force on the unmanned vehicle is deflected and moves toward the target point.In order to avoid collision with obstacles,the traditional gravitational potential field function is replaced by introducing a distance threshold factor so that the gravitational potential field function is a segmented function about the distance between the unmanned vehicle and the target point.(3)Research on unmanned vehicle path planning based on bat algorithm-artificial potential field.According to the above analysis,the improved bat algorithm has a fast convergence speed and high accuracy in finding the optimal;the improved artificial potential field method has a strong performance in real-time obstacle avoidance,therefore,the two algorithms are fused to achieve optimal path planning for unmanned vehicles.The specific idea is to initialize the location of the bat population according to the environmental potential field model established in(2)and add it as the path planning term to the objective function of the improved bat algorithm in(1)to obtain the Bat Algorithm-Artificial Potential Field(BA-APF),and then smooth the path by the three-sample interpolation method to avoid excessive path corners.By comparing the performance of the BA-APF algorithm studied in this paper with the particle swarm-bacterial foraging algorithm,ant colony-genetic algorithm and cuckoo simulated-annealing algorithm in solving the optimal path planning problem for unmanned vehicles in MATLAB simulation experiments,it is concluded that the BA-APF is superior to the other three hybrid swarm intelligence algorithms to a certain extent.