A high-speed uv motion control method

Abstract

The invention discloses a high-speed autonomous underwater vehicle (AUV, Autonomous Underwater Vehicle) motion control method. The method integrates the radial basis neural network and the boundary layer method into the traditional sliding mode control to improve the speed of the AUV during high-speed motion. control accuracy, and effectively suppress the chattering of the control system. The high-speed AUV motion control method involved in the present invention includes the simplification of the high-speed AUV motion model and the design of the motion control method. Compared with the traditional AUV model simplification method, an adaptive parameter model is added when the model is simplified, so that all The established model is closer to the actual situation; compared with the traditional AUV sliding mode control method, the present invention adds the compensation of the radial basis neural network in the sliding mode control, and improves the switching surface of the sliding mode through the boundary layer method, The sliding mode control system can maintain high control precision and suppress chattering.

Description

Technical field [0001] The present invention relates to the field of autonomous underwater robots (AUV, Autonomous Underwater Vehicle), and more particularly to a high-speed AUV motion control method. Background technique [0002] Autonomous Underwater Vehicle, AUV) is one of the powerful tools of exploring marine space. AUV has become more and more widely used in military marine technology, oilfield survey, seabed salvage, pipeline maintenance, seabed survey. However, traditional underwater robots generally have problems with low speed, environmental adaptation capacity, and the urgent needs of high-speed AUVs in terms of rapid emergency search, underwater environment, and explore the motion control method of high-speed AUV is important. direction. [0003] PID Control, Back Stepping Control, Fuzzy Control, Sliding Mode Control, Neural Network Control, etc. is some of the commonly used controls today Method, and in recent years, deep learning (Deep Learning) has also greatly pro...

AI Technical Summary

Problems solved by technology

PID control can only be applied to simple control of some AUVs under weak maneuvering, but it is sensitive to changes in environmental parameters, and tuning optimization is troublesome; backstepping control relies on accurate mathematical models, but it is difficult to accurately model this type of new high-speed AUV Obtained; although the sliding mode variable structure control has the characteristics of fast response, it is easy to cause chattering; the fuzzy controller relies on prior knowledge; although the neural network control has a strong nonlinear approximation ability, its network layers and The number of nodes in each layer is difficult to determine; deep learning has powerful complex nonlinear modeling capabilities, but its training is time-consuming, and the verification of model correctness is complex and cumbersome

Images