Control oriented real time model of marine powerstation diesel engine based on neural network

To build a precise and real time model is a key issue in fulfilling on optimal control of marine power station diesel engine in the hardware-in-the-loop simulation system. In this paper, a control oriented real time model for marine power station diesel engine is proposed based on neural network. After establishing the quasi-steady model of diesel, the hybrid model based on the compensation of neural networks, which raises precision of diesel engine model, are obtained. A NN model is obtained from the hybrid one to meet with the requirement of real time simulation. Simulation studies are conducted, and results show that the model built is precise and real time, and can be used for developing and analyzing engine control algorithms in the hard-in-the-loop simulation system.