PARTICLE SWARM OPTIMIZATION USED FOR MECHANISM DESIGN AND GUIDANCE OF SWARM MOBILE ROBOTS

Abstract

This chapter presents particle swarm optimization (PSO) based algorithms. After an overview of PSO’s development and application histor y also two application examples are given in the following. PSO’s robustness and it s simple applicability without the need for cumbersome derivative calculations make it an attractive optimization method. Such features also allow this algorithm to be adjusted for engineering optimization tasks which often contain problem immanent equality and inequality constraints. Constrained engineering problems are usually treated by sometimes inadequate penalty functions when using stochastic algorithms. In this work, an algorithm is presented which utilizes the simple structure of the basi c PSO technique and combines it with an extended non-stationary penalty function approach, called augmented Lagrangian particle swarm optimization (ALPSO). It is used for the stiffness optimization of an industrial machine tool with parallel kinematics. Based on ALPSO, we can go a further step. Utilizing the ALPSO algorithm together with strategies of special velocity limits, virtual detectors and others, t he algorithm is improved to augmented Lagrangian particle swarm optimization with special velocity limits (VLALPSO). Then the work uses this algorithm to solve problems of motion planning for swarm mobile robots. All the strategies together with basic PSO are corresponding to real situations of swarm mobile robots in coordinated movements. We build a swarm motion model based on Euler forward time integration that involves some mechanical properties such as masses, inertias or external forces t o the swarm robotic system. The results show that the stiffness of the machine can be optimized by ALPSO and simulation results show that the swarm robots moving in the environment mimic the The content of this paper is exactly the same as the final published, except the formalization. For correct citation please visit my website.