NARMAX Self-Tuning Controller for Line-of-Sight-Based Waypoint Tracking for an Autonomous Underwater Vehicle

In this brief, a constrained self-tuning controller (CSTC) is developed for the heading and diving motions of an autonomous underwater vehicle (AUV) considering the parameter variation and practical realization of the algorithm. Parameters in the AUV dynamics may vary due to change in payload or physical structure. A Nonlinear Auto-Regressive Moving Average eXogenous (NARMAX) model is designed using the significant regressors to identify the AUV heading and diving dynamics, respectively. The parameters of the NARMAX model are updated using a recursive extended least square algorithm at each time instant. Furthermore, using the identified model, a CSTC law is designed to track desired waypoints using line of sight guidance law. The controller gains are updated at each instant satisfying the actuator constraint and computational efficiency is studied in view of achieving the practical implementation of the developed algorithm. The efficacy of the developed NARMAX-based CSTC algorithm to track a given reference is verified in an experimental environment with payload variation and external disturbance. From the obtained results, it is concluded that the developed control algorithm is effective for an AUV to track desired waypoints or heading/diving reference.