Multi-neural Networks Control of Marine Diesel Engine Generator Set

For multi-variable control system of voltage and frequency of marine diesel engine generator set, multiple CMAC neural networks (NN) controllers are designed and applied in system control according to principle of mapping connection and learning rule of CMAC NN. The CMAC NN control algorithm is provided with ability of fast learning rate on line. In control realizing process, the inverse dynamic model of object control system was obtained by multiple CMAC NN through learning. This model is a general approach model. The simulation results of the responses of generator load testing indicated that the quality is well in coordinate between two CMAC NN control loops.