Optimal Design of Hybrid Optimization Technique for Balancing Inverted Pendulum System

Abstract

Inverted Pendulum system is one of the most exciting problems in control theory. In this research work, a new variant of Grey Wolf optimizer (GWO) via Particle Swarm Optimization (PSO) based on Adaptive Constants (AC) is proposed. The proposed technique (GWO/PSO-AC) is tested via twenty-three benchmark functions and compared to GWO based on PSO without adaptive constants (GWO/PSO). The suggested technique shows superiority in determining the optimal solutions for the well-established benchmark test functions with high computing performance compared to alternative techniques. The proposed GWO/PSO-AC technique, is employed to tune the parameters of the Variable Structure Adaptive Fuzzy (VSAF) controller in addition to the Reduced Linear Quadratic Regulator (RLQR) suggested by the authors. Both controllers are used to stabilize the cart position and to swing up the pendulum angle. The RLQR has an advantage over regular LQR, which is, the numberof the required parameters to obtain the required LQR gains is reduced. The proposed technique is compared with two optimization techniques. The proposed technique achieves high performance for both the cart position and the pendulum angle. The attained results are very promising.