Research on UGV Path Planning in Tunnel Based on the Dijkstra*-PSO* Algorithm

Abstract

Nowadays, the researches on UGV path planning mainly are mainly related to the interwoven road environment of the city or the relatively flat and broad environment in the field. As a kind of obstacle models that can not be ignored in the process of movement, the research on UGV path planning in the tunnel has considerable necessity. In this paper, aiming at the internal path planning problem of obstacles such as tunnels, a motion path planning scheme of unmanned ground vehicles (UGVs) based on Dijkstra- particle swarm optimization (Dijkstra-PSO) is proposed. Firstly, the appropriate obstacle curve function is selected according to the characteristics of the obstacle model; Then the Dijkstra algorithm is improved to realize the initial screening of the number of path nodes through a node screening rule, and the initial optimized path can be obtained by using the improved algorithm; By using the dynamic step size to adjust the value spacing and taking the maximum available step size as the influence factor to evaluate whether the optimization of tentative point is stagnant or not, the PSO algorithm is further improved to avoid generating redundant optimized nodes. Finally, an improved PSO algorithm is proposed to re-optimize the initial optimized path. The results show that both the Dijkstra algorithm and the hybrid algorithm can realize path planning without collision and obtain the optimal path from the starting point to the end point. Compared with only using one algorithm, the hybrid algorithm (the Dijkstra*-PSO* algorithm) can effectively shorten the path length and improve the quality of the path searching. In the experiment, compared with the experimental results of the single improved algorithm (the improved Dijkstra algorithm and the traditional PSO algorithm) and the double improved algorithm (the improved Dijkstra algorithm and the improved PSO algorithm), the path length obtained by the latter one is shorter and the number of path nodes is smaller, which indicates that the double improved algorithm can meet the requirements of the later application of UGV to carry out collision-free movement in similar tunnel obstacles.