Performance Evaluation of Different Control Algorithms for Torpedo-Shaped AUVs

This paper compares Linear Quadratic Regulator (LQR) and Proportional Integrative Derivative (PID) control algorithms implemented in an in-house built Autonomous Underwater Vehicle (AUV). The AUV has actuators to control five degrees of freedom and features a non-linear dynamics model established through a Python-based simulation environment. Path following is achieved by employing the Line-of-Sight (LOS) and fully actuated path-following guidance law. The evaluation of the control algorithms is based on the trade-off between effort, endurance, time taken to reach the intended path, overshoot and steady-state error.