考虑障碍物避让的水下滑翔机局部路径规划方法

Local path planning method of underwater glider considering obstacle avoidance

针对水下滑翔机"V"形轨迹的周期性运动特点,提出一种适用于水下滑翔机在三维路径中避让障碍物的粒子群优化(PSO)算法。根据水下滑翔机的运动特征,给出水下工作环境的三维空间建模方法。结合水下滑翔机能耗问题,选择用规划路径的长度、障碍物的危险度和路径平滑度来制定适应度函数。提出一种惯性权重的非线性动态调整策略,提高了算法的全局搜寻能力。将优化后的算法与基本粒子群算法比较,结果表明:优化后的算法可以生成质量更高的路径,并且可以在更少的迭代次数下收敛到全局最优,有效地提高了粒子群算法应用在水下滑翔机路径规划的算法速度和能耗效率。

Aiming at periodic motion characteristics of"V"trajectory of underwater glider,a particle swarm optimization(PSO)algorithm is proposed to avoid obstacles in the three-dimensional path of underwater glider.Firstly,according to the motion characteristics of underwater glider,a three-dimensional space modeling method of underwater working environment is proposed.Then,combined with the energy consumption problem of the underwater glider,the fitness function is formulated by the length of the planned path,the danger degree of the obstacles and the smoothness of the path.Finally,a nonlinear dynamic adjustment strategy of inertia weight is proposed to improve the global search ability of the algorithm.Comparing the optimized algorithm with the basic PSO algorithm,the results show that the optimized algorithm can generate higher-quality paths and can converge to the global optimum in fewer iterations,effectively improve algorithm speed and energy consumption efficiency of PSO algorithm applied in underwater glider path planning.