Adaptive Control of Ball and Beam System Using Knowledge-Based Particle Swarm Optimization

Abstract

A knowledge-based Particle Swarm Optimization (PSO) algorithm is used to achieve more optimized control of Ball and Beam System (BBS) adaptively. It adopts an improved nonlinear inertia weight, an adaptive strategy and a fitness function combining prior knowledge and one traditional performance criterion. Comparing four classic performance criteria, the simulation results indicate that Integral of Time multiply Absolute Error (ITAE) is better, and it is combined with prior knowledge. Based on the response curve of advanced correction, Ziegler-Nichols, basic PSO algorithm and knowledge-based PSO algorithm through experimental simulation, knowledge-based PSO algorithm is more effective to BBS.