Optimization of Linear Quadratic Regulator for Reaction Wheel Inverted Pendulum using Particle Swarm Optimization: Simulation and Experiment

This paper presents an optimization approach for the Linear Quadratic Regulator (LQR) applied to a Reaction Wheel Inverted Pendulum (RWIP) system, utilizing Particle Swarm Optimization (PSO). The study involves both simulation and real-world experimental verification. A mathematical model of the system is first developed using the Euler Lagrange method, and the LQR controller is designed to stabilize the highly nonlinear system, specifically a Single Input-Multiple Output (SIMO) system. PSO is employed to fine-tune the LQR parameters, optimizing performance metrics such as overshoot, settling time, and steady-state error. Simulation results, performed in MATLAB, are compared with experimental results obtained using an STM32F407 microcontroller-based hardware setup. PSO optimized LQR demonstrates significant improvements in stability and response time, outperforming standard optimization. The results confirm the efficiency of PSO in optimizing control systems for nonlinear dynamics, with potential applications in balancing robotics and self-stabilizing vehicles.