针对复杂环境下传统蝙蝠算法(BA)在无人车路径规划过程中存在的规划效率低、可达性差等问题,同时考虑无人车可行性及安全性要求,提出了一种多种策略改进的蝙蝠优化算法(improved bat algorithm,IBA).改进算法同时引入了混沌映射和截断均值稳定策略,每次迭代后用新蝙蝠替换旧蝙蝠中最差的蝙蝠,以增加种群多样性,从而使算法在复杂环境下快速收敛同时保持稳定运行.此外,利用B样条曲线平滑生成的行驶路径,减少转向代价.仿真结果表明,与其他算法相比,IBA算法在路径长度、运行时间指标上均更短,规划成功率更高,更有利于无人车行驶.
Path planning of unmanned vehicle based on IBA algorithm
In order to solve the problems of low planning efficiency and poor accessibility in traditional bat algorithm(BA)for unmanned vehicle path planning in complex environments,and considering the feasibility and safety requirements of unmanned ve-hicles,an improved bat algorithm(IBA)with multiple strategies was proposed.The improved algorithm also introduces chaotic mapping and truncated mean stability strategy,replacing the worst bats in the old bats with new bats after each iteration to increase population diversity,so that the algorithm can converge quickly and maintain stable operation in complex environments.In addi-tion,the smooth driving path generated by B-spline curve is used to reduce the turning cost.The simulation results show that,com-pared with other algorithms,IBA algorithm has shorter path length,running time indicators,higher planning success rate,and is more conducive to the driving of unmanned vehicles.
bat algorithmB-spline curveunmanned vehiclepath planning
万方数据
