EXPERIMENTAL PLATFORM FOR ELECTRO-HYDRAULIC SERVO ACTIVE SUSPENSIONS AND METHOD FOR FAULT DIAGNOSIS

An experimental platform of the present disclosure is very close to an active suspension apparatus in a real vehicle. Fault feature data acquired on the basis of the experimental platform can well represent actual faults, so that a fault diagnosis model obtained on the basis of training of fault feature data sets has a relatively high practical value. According to the method of the present disclosure, noise reduction processing is performed on acquired vibration signals by combining with longitudinal displacement and lateral inclination angle data as well as positional relationships between vibration sensors of different hydraulic elements and an inertial sensor, and feature extraction is performed on vibration signals after noise reduction processing, so that constructed fault feature data sets are more accurate and more characteristic. The fault feature data sets are gradually expanded according to the actual faults, so that the fault diagnosis model has a certain self-learning capability.