<p>The principles of inverted pendulum systems are extensively utilized in applications such as slosh modeling, humanoid robots, and space technologies. Consequently, design of effective controllers for these systems is a critical task. This study focuses on the inverted pendulum models through linearization and design of two controllers which are the linear quadratic regulator with particle swarm optimization and the linear quadratic regulator with integrator based on particle swarm optimization. The optimal controller gains are derived using particle swarm optimization, a widely used artificial intelligence technique. The proposed controllers are tested against various reference inputs including impulse, step, square, sinusoidal, and sawtooth signals on four different inverted pendulum models. The controllers are also applied to Cubli rovers and monowheel robots which utilize inverted pendulum principles. Comparative analysis is conducted with existing controllers from literature including the neuro-fuzzy controller, genetic algorithm-based linear quadratic regulator, and whale optimization-based sliding mode controller. Further evaluation involves testing the controllers on an experimental inverted pendulum setup, both on the nominal system and with an uncounted external weight or bob mounted to the nominal system, introducing unmodeled dynamics. The robustness of the controllers is assessed in this augmented system. The performance of these proposed controllers are quantified using error indices and the absolute maximum peak. Results and comparative study indicate that the proposed controllers effectively regulate system dynamics, minimize errors, and maintain system stability. These findings highlight the controllers’ versatility and efficacy in various inverted pendulum-inspired applications.</p>