This study introduces a PID parameter self-tuning method for non-coaxial two-wheeled vehicles using an Improved Beetle Antennae Search (IBAS) algorithm to enhance control accuracy. A mathematical model is developed based on the vehicle’s mechanics, and a Simulink simulation is constructed in Matlab. The method addresses the limitations of traditional PID control’s empirical parameter tuning by optimizing PID parameters with IBAS. Experiments across three motion states demonstrate IBAS’s superiority over BAS and PSO algorithms. In circular motion, IBAS reduces overshoot by 64.3% and ITAE by 12.3% compared to BAS, validating the optimized PID system’s effectiveness and supporting stable vehicle operation in confined spaces.