Distributed Parallel Motion Control For The Multi-thruster Autonomous Underwater Vehicle

The motion of Autonomous Underwater Vehicle (AUV) is strong nonlinear,coupling of every degree of freedom and is hard to model. The motion control is acomplex question of AUV, and it will directly influence the task implementation ofupper layer.With the development of distributed control system, the architecture is also morecomplex, so we should choose more accurate control algorithm. Now many newcontrol algorithms have been gradually introduced to the controlling area of AUV,such as PID control, self-adaptive control, sliding mode control, fuzzy control, neutralnetwork control and the combination of the above methods.The main target of this paper, based on Variable-Structure Control with SlidingMode (SMC), is to design a kind of Distributed Parallel Motion Control (DPMC)system for the Multi-thruster AUV. Based on the prototype of C-RANGER, the paperintroduced the disposal of thruster and designed driving system with the method ofSMC. To resolve the creeping phenomenon of low-speed mechanical servo system,fuzzy estimation was used in the friction of thruster with the method of controlcompensation.A distributed parallel motion control system merged with distributed control andfiled bus technology, used the parallel hardware architecture and decoupling controlalgorithm skillfully. The paper analyzed the relationship between the thruster anddegree of freedom according to the kinetics model of AUV. In addition, we also set upa model of sliding mode controller and apply simulation research to the DPMCalgorithm under the MATLAB circumstances. To decrease chattering phenomenon inthe sliding mode control effectively, learning function of neural network is utilized toadjust sliding surface variable online. The results of simulation show that DPMCalgorithm can complete the expectant routine, and is highly reliable and flexiblelayout.