The invention provides a fault diagnosis method for electromechanical equipment based on a gray model, which comprises the steps of first, collecting operation parameters of the current electromechanical equipment; second, enabling the collected operation parameters to serve as an original modeling sequence, building a gray model, and predicting relevant data in the system by using a prediction model; third, obtaining sample data of the operation parameters of the electromechanical equipment in different fault states, building a reference model of faults of the electromechanical equipment, finding out a fault source of the system and determining a fault parameter of the system; fourth, magnifying the variation trend of the data, performing fault diagnosis, performing consistency checking on a predicted value and a solution of an analytical model, substituting the predicted result into a parameter estimation model to solve, obtaining a predicted value of the fault parameter at a futuremoment, predicting the fault parameter according to the prediction model and the parameter estimation model, and judging a fault cause according to a mechanical relationship between the variables. Thefault diagnosis method can realize diagnosis for the fault type of the electromechanical equipment during the operation.