Research On Autonomous Planning Method Based On Quantum Particle Swarm Optimization For Unmanned System

Unmanned system(US)is a hotspot in the field of the intelligent machines research at which autonomous planning is the key element.The autonomous planning of US consists of two parts: one is the global path planning on the basis of the known environmental information;the other is the local planning for obstacle avoidance based on sensor data.Global planning mainly works out an optimal path for US between the start and the end according to the environmental information.Then local planning for obstacle avoidance refers to the real-time obstacle avoidance of US in accordance with information measured by sensors.This paper focuses on the autonomous planning of Unmanned underwater vehicles(UUV),combines with Quantum-behaved Particle Swarm Optimization(QPSO),completes the study of UUV's obstacle avoidance planning under partially known environment.First of all,this paper introduces the basic concept of QPSO algorithm,describes the evolution principle of the particle and the implementation steps of the algorithm.In order to improve the convergence speed and optimization ability of the algorithm,a QPSO algorithm with particle's taxonomic evolution is proposed.By the comparison with other algorithms,the results indicate that the algorithm proposed is difficult to fall into the local optimum and has faster convergence speed.Secondly,aiming at the path planning tasks of UUV,we establish an annular space model according to the known environment,which combines with the static obstacle information by using the algorithm proposed,and have path optimization in annular space.For the known dynamic obstacles,we adjust the speed and direction of UUV to avoid obstacle by optimization algorithm.In this paper,the simulation experiment certifies that this method can obtain under a variety of environments.Thirdly,for the UUV encountering unknown obstacles in autonomous motion,a method of the local planning for obstacle avoidance is proposed based on sensor detection window.When UUV detects unknown static obstacles,the algorithm plans out a new path to avoid obstacles in the optimization window.As the optimization window moves forward,new paths constantly come out until it is safe to avoid obstacles and UUV return to the global path.When UUV encounters motion obstacles,the algorithm first decide whether they will meet and if there is a risk of collision,the real-time departure methods is taken for obstacle avoidance.Finally,the simulation experiment proves that UUV can have safe navigation by this method.Finally,the methods of global planning and local planning proposed,constitutes the UUV autonomous planning system.On the basis of Qt development platform,UUV autonomous planning simulation software is designed and developed.Besides,comprehensive simulation experiment of UUV autonomous planning is done with partially known environment on the simulation platform.The experimental results prove the effectiveness of this method.