Synthesis of Automatic Motion Control Systems of an AUV Based on Fuzzy Logic Methods with Neural Network Setting of Parameters

The article deals with the problem of controlling the stabilization of the depth of an autonomous underwater vehicle with a search type. The main results in this area were obtained for torpedo-shaped underwater vehicles, the dynamics of which significantly differs from the dynamics of the object under study. An algorithm is presented; methods of fuzzy logic are used for neural network adjustment of control parameters so that the result of the operation of a fuzzy controller is reproduced. This algorithm can be applied to playback and other regulators, which makes it a universal tool for building automatic control systems. The results of modeling the controlled movement of an underwater vehicle in the vertical plane are presented, showing the highly efficient functioning of the adaptive neuro-fuzzy control in comparison with the proportional-differential controller.