Robot Path Planning and Experiment with an Improved PSO Algorithm

An improved PSO (particle swarm optimization) algorithm is proposed for robot path planning, aiming at the shortcomings of PSO algorithm, such as the low optimization accuracy and the premature. Firstly, an improved model of particle velocity update is proposed according to the principle that the variables change in the direction of negative gradient in gradient descent method. Then, an adaptive particle position update coefficient is added in order to improve the efficiency and the accuracy of particle search. Finally, an improved PSO algorithm is designed by introducing ε-greedy strategy. The experimental results on some optimized test functions show that the search accuracy of the proposed algorithm is at least twice that of other algorithms, and the convergence speed is also greatly improved. The proposed algorithm and the improved DC-HPSO (dynamic clustering hybrid PSO) algorithm are applied to the simulations and the actual experiments of static path planning. Results show that the proposed model has the advantages of high accuracy, high efficiency and high success rate.