Cost Map-Integrated Model Predictive Control for Safe Path Following of Mobile Robots

Abstract

Autonomous navigation in mobile robots requires reliable path following and obstacle avoidance, especially in dynamic environments. Differential drive robots have limitations due to non-holonomic constraints, making accurate path tracking more challenging. This paper proposes a Model Predictive Control (MPC) for path following integrated with cost map-based obstacle avoidance within the Robot Operating System (ROS) framework. The proposed method combines global path planning using the A star algorithm with a predictive local controller that incorporates environmental information from the cost map to ensure safe navigation. The MPC formulation minimizes tracking errors while considering system constraints to produce smooth and stable motion. The navigation system is evaluated in an indoor environment with static and dynamic obstacles. Results show that the proposed approach improves path following accuracy by approximately 22.21 percent, reduces heading error by about 29.14 percent, and produces smoother and more stable control inputs compared to conventional methods. These improvements are achieved with only a marginal increase in path length of approximately 0.4 percent, indicating an acceptable trade-off between efficiency and safety. Overall, the integration of MPC with cost map-based environmental representation provides a robust and effective solution for mobile robot navigation in complex environments.