AUVS' DYNAMICS MODELING, POSITION CONTROL, AND PATH PLANNING USING NEURAL NETWORKS

摘要

Accurate identification of nonlinear time variant MIMO systems, especially in case of AUVs is essential for implementation of control algorithms and navigation purposes. Control problems of AUVs have also difficulties due to the nonlinear dynamics behaviors of vehicles and also unpredictable effects come from the surrounded water mass. These nonlinear effects are so complicated that bring difficulties for dynamics modeling and position control descriptions while using conventional methods. The proposed method here uses neural networks as a general idea for dynamics modeling and position control of any six-degree of freedom rigid body and are applied to an AUV, named Twin Burger 2, as an example. Supervised Learning and Unsupervised Learning are used for adjusting the neural networks' synaptic weights and the results are illustrated. Path planning of AUVs using neural network is also addressed here as of a complicated control scheme and Reinforcement Learning is used for adjusting the neural network parameters of the path planning module via some obstacle avoidance examples.