The maintenance and forecast expert system of equipment based on Artificial Neural Network is composed of control, measure, failure forecast, execution, data processing module and database. The data processing module obtains the change of the controlled objects′ structure and parameters, then takes correspondent measures according to the examination and diagnosis information. The failure forecast module finds the control system fault, separates the fault symptom location, tells the fault kind, estimates the magnitude and time of the fault, and finally makes evaluation and decision.