Optimization of parameters for an USV autopilot

Abstract

The purpose of this research is to provide an insight into the effect of different parameters on the autopilot design. Various independent parameters of three autopilots namely, proportional integral derivative, linear quadratic regulator and model predictive controller are analyzed and evaluated to obtain optimum performance. Further these optimal parameters are employed in a controller design which is integrated with a Kalman filter and an interval Kalman filter based navigation system and a line of sight based guidance system. Overall performance of the autopilots with the optimum parameters are presented in a tabular form to enable easier comparison and to serve as a benchmark to tune autopilot parameters of an uninhabited surface vehicles.