The Research Of Optimal Design Of Autonomous Underwater Vehicle Fuzzy Neural Network Controller

Fuzzy neural network control, which is suitable for the control of strong nonlinear system with complicated kinematic property, is widely used in the field of intelligent control. Underwater vehicle, the precise kinetic equations of which are usually hard to acquire, is a nonlinear system with strong coupling motions of each freedom, so fuzzy neural network control is widely used in the motion control of underwater vehicles. However, the control rules are often big problems and lots of experience of the designer is needed to give the right rules.Based on the space motion equations of six-freedom of the AUV, a four-freedom dynamic simulation model of the multifunctional AUV is constructed combined with the engineering demands. And a kind of control method using fuzzy neural network to control the AUV is presented.Owing to the control rules of the fuzzy neural network controller which are difficult to determine all by hands, intelligent algorithms are used to design the optimal rules automatically. And the theories of genetic algorithm (GA), immune algorithm (IA), particle swarm optimization (PSO) algorithm are investigated and used in the optimal design of control rules of the fuzzy neural network controllers. Additionally, HPSO algorithm which brings the idea of immune algorithm into PSO algorithm is proposed and also used in the optimal design. Finally, the controllers optimized by those algorithms above are tested on the simulation platform of the AUV.The simulation results show that it's feasible to design the control rules in terms of intelligent algorithms.