Research On Path Smoothing Based On Improved Particle Swarm Optimization Algorithm

As the performance of mobile robots continues to improve,they are widely used in industrial,agricultural,medical,and service industries.Traditional path planning algorithms plan global paths that are composed of multiple folds,so there is an inevitable problem of sharp turns(large changes in curvature),especially in environments with dense obstacles.In practical application scenarios most mobile robots required high smoothness of paths,such as warehouse logistics robots,which need to carry the weight of goods when moving,so their driving paths should be as smooth as possible in order to ensure the safety of goods and the mobile robot itself.In response to these problems,this paper conducts research from the following two aspects:(1)A method based on the improved particle swarm optimization algorithm is proposed to optimize the polynomial curve parameters for reducing energy consumption caused by excessive turning points in the original path for mobile robots.Firstly,the path coordinates of the mobile robot motion are set artificially,and the path before smoothing can be regarded as a fold line connected with adjacent path points.Second,the shortest distance from the path point to the smoothed path and the safe avoidance of obstacles are chosen as the penalty function,and the shortest length of the whole smoothed path is taken as the cost function.Again,Particle swarm optimization has strong generality and a simple principle with few parameters,making it an ideal optimization algorithm for this study.The size of inertia weights directly affects the global and local search ability of particles in the particle swarm algorithm,so in order to balance the global and local search ability of particles,nonlinear inertia weights are introduced to improve the particle swarm algorithm.Finally,experimental simulations are conducted by improving the particle swarm algorithm with several different inertia weights,as well as by improving the particle swarm algorithm for different order polynomial curve smoothing path simulation experiments in the environment without obstacles and with obstacles.The experimental results show that the nonlinear inertia weights proposed in this paper improve the global and local search ability of particles.By comparing the experimental results of several polynomial smoothing paths of different orders in an obstacle-free environment and an obstacle-bearing environment,the quadratic polynomial curve smoothing path is more ideal and meets the driving conditions of the mobile robot under the optimization effect of the improved particle swarm algorithm.(2)Polynomial smoothing of robot paths may result in missing important path points,which can cause the robot to fail to complete its task.Therefore,a method based on improved particle swarm optimization algorithm is proposed for multi-segment path connection and smoothing.The path obtained by the path planning algorithm is divided into several sub-paths,and each sub-path is smoothed using a cubic Bezier curve.The smoothed adjacent sub-paths are connected and the conditions for satisfying the continuity at the connection points are derived using the second-order derivatives of the third-order Bezier curves.The optimization problem for a complete path with several variables is transformed into an optimization problem for two variables by the geometric continuity relation at the connection points.Under the conditions of safe obstacle avoidance and minimum path curvature,the shortest path length after smoothing is guaranteed,so the path smoothing problem is transformed into a parametric optimization problem by the particle swarm optimization algorithm.The particle algorithm is easily caught in the local optimum in optimization,so a perturbation strategy is applied to the particle velocity during particle search to make it jump out of the local optimum,thus enhancing the optimization-seeking ability of the particles in the smoothed path.Finally,the experimental simulation of several different perturbation strategies to improve the particle swarm algorithm and the adaptive simulation experiments of the path smoothing method with multi-segment path articulation in different path scenarios in both obstacle and obstacle-free environments are carried out.The experimental results show that the perturbation strategy proposed in this paper can help the particles to jump out of the local optimum.In complex without obstacle environments and obstacle environments with different distribution densities,the path smoothing method based on improved particle swarm algorithm for multi-segment path articulation proposed in this paper is feasible and effective.