Research On Hydraulic Cylinder Fault Pattern Recognition Method Based On Information Fusion

Hydraulic systems have many leading features,such as light-weight structures, easy implementation of poleless speed-adjusting and remote control, and moreover, their operations are stable. Therefore they are widely applied in numerous engineering fields and play a key role in numerous engineering equipment, including drivering and control. The timely and accurate implementation of hydraulic system's fault diagnosis and its normal operation is playing a significant role in improving efficiency, reducing maintenance costs and reducing unnecessary economic losses.As the execution element of hydraulic system, hydraulic cylinder'failure will directly influent the hydraulic system or even overall unit.Information fusion technology is used in hydraulic cylinder" fault diagnosis based on the analysis of current research. Redundancy and complementary information of each sensor are made full use, the decision's accuracy and robustness of system are improved.The failure mechanism of hydraulic cylinder' internal leakage was first introduced in this paper,test bench was built to simulate hydraulic cylinder" internal leakage. The suitable hardwares were choosed and the data acquisition were programmed to realize the signal acquisition.After Signal characteristics were acquired, information fusion diagnosis method was given based on the BP neural network and was tested in hydraulic cylinder' internal leakage. Due to the influence of uncertainty factors, information fusion diagnosis method based on the BP neural network has certain shortcomings in the fault diagnosis of hydraulic cylinder's internal leakage. D-S evidence theory 'advantage in dealing with uncertain knowledge makes up this deficiency, but still can't achieve complete diagnosis. After the above methods were explored,comprehensive diagnosis method based on the BP neural network and the D-S evidence theory was given and was tested effective. This study provide a new way for improving the accuracy rate in hydraulic cylinder' fault diagnosis.