Abstract: The vibration signals of the diesel engine are reconstructed in the phase space. Then the combination of neural networks to vibration signals approximation and time series prediction are applied. It is a two-stage system. The first stage contains two NN predictions, a multi-layer feed for work NN and functional link NN. They are trained by fuzzy back propagations algorithm. The second stage consists of a combination module to mix the two individual prediction proved in the first stage. It is trained by Karmarkar’s linear programming algorithm .It is proved available in the example.

Key words: Diesel engine, Fuzzy back propagation, Linear programming, Neural networks, Prediction, Vibration signal