The External Characteristics Simulation System of Diesel Engine Based on Levenberg-Marquardt Algorithm

摘要

A control system which can analog the external input and output characteristics of the diesel engine is aimed to be designed in this paper. Proposed a scheme that using the method named "Adaptive neural network based on model reference" to control asynchronous motor in order to simulate the external characteristics of the diesel engine. The simulation model of diesel engines, asynchronous motors and control systems based on MATLAB /SIMULINK are established. The results show that, the improved BP neural network algorithm can improve the efficiency of the training of neural network significantly, the asynchronous motor under the control of adaptive neural network based on model reference can approach the output characteristics of the diesel engine accurately under the same input.