Intelligent Dynamic Control for Underwater Vehicles Based on Fuzzy CMAC

The operation of an underwater vehicle (UV) is very complicated due to the multi-input multi-ouputs structure, numerous operational requirements and constraints, so the workload of the operator is very heavy. how to dreduce the workload of the operator while maintaining the fault tolerance capability is a challenging problem. We propose to develop an intelligent dynamic control systems for UVs based on the Fuzzy Cerebellar Model ARithmetic Computer (FCMAC) neural network. FCMAC is an expert knowledge in its rule base. When some actuators fail, it can reconfigure the control system using the remaining control devices to continue the mission or return to the base safely. With this intelligent controller, the UV operator can concentrate on guidance and navigation, and the controller will minimize the risk of vehicle loss and maximize the chance of completing missions automatically. This controller can also assist a moderate skilled operator, so he can perform as well as experienced operators. This intelligent dynamic controller can be applied to unmanned UVs as well. The deliverables will include detailed documents of the design, performance and reliability analysis, and a prototype fo the controller.