With the increasing size of the construction machinery,the hydraulic system is also more complex,so the hydraulic system fault diagnosis becomes very difficult.The hydraulic cylinder is the actuator of the hydraulic system,and if it fails,it may cause the entire hydraulic system to be paralyzed.Therefore,timely and rapid diagnosis of hydraulic cylinders running to ensure the normal operation of engineering machinery and equipment is very important.In this paper,based on MATLAB software,through analysing pressure signal of the hydraulic cylinder in time domain and frequency domain,feature parameters reflecting the leakage failure are extracted.Then,a neural network system is established,and time-domain feature parameters are used as sample data to train the neural network.After the performance of neural networks meets certain requirements,it could be used as a platform of fault diagnosis for the hydraulic cylinder.Finally,based on MATLAB GUI,a human-machine interface is created in order to achieve simple operation and the purpose to understand easily. |