Auto-tuning of a modified L1-adaptive controller with genetic algorithms for dynamic positioning of a remotely operated vehicle under marine currents

Abstract

In this article, a modified L1-adaptive controller with auto-tuning using a genetic algorithm is presented for dynamic positioning of remotely operated vehicles (ROVs) under marine currents, based on a six-degree-of-freedom nonlinear model of an ROV. To enable tuning of some of the parameters of the controller, a cost function related to the error of the steady state positions of the system is minimised with the use of the genetic algorithm. A series of simulations are conducted to ascertain the performance of the system with the implemented controller, taking into consideration the vehicle position, orientation, and control signals sent as commands to the thrusters. The simulations are carried out with noise levels representative of those encountered by the standard underwater instrumentation on an ROV, as well as with underwater current velocities. In addition, the results are compared with those of a classical controller to verify the improvements offered by the controller proposed in this paper.