改进粒子群算法在三维水下导航规划中的应用

The Application of an Improved PSO to the Submersible Path-Planning

基于群智能理论提出了一种改进粒子群算法.以非线性策略改变惯性权值,增强粒子群算法局部和全局寻优的调度能力,以改变迭代公式加大"优秀"粒子的影响,增强粒子群体的寻优能力.通过理论推导、验证和实验仿真,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三维路径规划中,通过对三维空间的分割降维,并进行条件约束,实现了将路径规划问题转化为路径点求解的优化问题.实验仿真获得了从起点到终点的无碰撞路径,验证了该方法的可行性.

Based on the theory of swarm intelligence, an improved particle swarm optimization （IPSO） is put forward. The IPSO can enhance scheduling ability for local and global optimal of PSO by using nonlinear strategy to change the inertial weights. It can also strengthen optimization ability of particles by changing the position iterative formula to increase effects of ＂excellent＂ particle. Through theoretical derivation, validation and experimental simulation, it is shown that the IPSO owns more excellent performance. The proposed IPSO is used in three dimensional path planning of underwater vehicle. By reducing the dimension of three-dimensional space and via segmentation adding constraint condition, the path planning problem is converted into the optimization problem about solutions of path points. Finally, through the experimental simulation, a path from start to finish point without collision is given, so the feasibility of this method is proved.