The Application of PSO Algorithm on PenduBot Control

Particle Swarm Optimization (PSO) algorithm is a new bionics algorithm. Because of its good characteristics, the algorithm has aroused many researchers´ interests. Most researches were centered on the control abilities of the algorithm on the strongly nonlinear and naturally unstable objects. The PenduBot is the most typical experiment object in the field of automatic control. In this thesis, the state-feedback theory was chosen based on linearized PenduBot model, and the control program was finished in the MATLAB environment according to the PSO algorithm. The thesis concentrated on the position and velocity evolution equations and fitness functions selection. Compared with the simulation results on PenduBot by mature algorithm LQR, the PSO algorithm performed better control abilities. Referring to the parameters from simulation, the two-link PenduBot can stabilize itself in the physical experiment. It was a reference to the algorithm convergence and anti-interference properties.