Intelligent Self Tuning Controller for A 6 DOF AUV

In this paper, an intelligent neural network method for controlling of a variable mass underwater vehicle is presented. The control scheme is capable of learning and adapting to changes in the vehicle dynamics and parameters, and consists of a gain tuning neural network and a variable gain PID controller. This neural network is trained so that the error between the plant output and reference signal is minimized. For updating the neural network weights, the system jacobian must be known. Because of the complexity and unknown governing equations of the system, jacobian calculation is too difficult and sometimes impossible. To overcome this problem another neural network (NNM) was trained for modeling the system and estimating the system jacobian. The results of this control scheme are compared with a constant gain PID controller. It is shown that the presented control scheme is better and more robust against disturbance than the conventional controller.