改进粒子群算法的潜器导航规划

针对粒子群算法的寻优搜索能力强和已有的一些导航算法存在收敛速度慢、迭代时间长的缺点，提出一种基于粒子群算法的潜器导航算法.利用群智能理论，对基本粒子群算法进行改进:提出一个含突变因子的可变调的惯性权值策略，从而达到增强粒子群算法局部和全局寻优的调度能力.通过实验仿真验证，证明了改进粒子群算法具有更优的性能.在此基础上，将该算法应用到水下潜器的路径规划中，通过对环境的建模分析进行条件约束，最终将路径规划问题转化为路径点求解的优化问题.实验仿真结果获得了从起点到终点的无碰撞路径，收敛速度也较快，验证了该方法的有效性和可行性.

Submersible path-planning based on an improved PSO

To overcome the shortcomings in slow convergence speed and long iteration time in the existing algorithms, take advantage of the searching ability of a particle swarm optimization (PSO), a new navigation algorithm was created based on PSO. First, the theory of swarm intelligence was used to improve the basic PSO algorithm: proposing a transformable inertia weight which contains a mutation factor to improve the PSO, and thereby enhancing the local and global optimum capability of PSO. Through experimental simulation and validation, the better performance of the improved PSO (IPSO) was proven. On this basis, the IPSO was used in path planning of an underwater vehicle. By using modeling analysis of space and adding constraint conditions, the path planning problem was converted into an optimization problem concerning solutions of path points. Finally, through experimental simulation, a collision free path from start to finish was given, proving the validity and feasibility of this method.