Research On Fault Simulation Of Hydraulic System Of Shield Machine Based On Neural Network

The hydraulic system plays an important role in the whole process of the shield machine.The hydraulic system is more closed and easy to make fault.When the fault is varied and difficult to repair.With the extensive application of shield machine in urban construction,the hydraulic failure of shield machine becomes a serious subject,and its diagnosis becomes more and more important.Based on the working principle of the hydraulic system of shield machine,the faults and causes of hydraulic pump,hydraulic valve,hydraulic cylinder and hydraulic motor are analysed in work.Signals are collected,the sensor is determined in detecting position distribution.Based on Microsoft Visual Studio platform,parameters data is classified and storaged to MySQL database.Hydraulic system data collection center of shield machine and Internet remote display interface are made to data management,including screw conveyor system and cutterheadderdrive system,foam system,segment assembling system and hydraulic thrust system.Starting from the principle of the hydraulic system,based on the AMESim simulation environment,cutterhead driving system,propulsion system,screw conveyor and segment assembling system are simulated,and it is found that when the system flow is 200r/min,the rotation speed is 886 RPM,and driving motor speed is stable;When the shield body is in high voltage stability in the process,the host of the displacement is always more than oil cylinder,and also constantly drivage,and the difference is on the rise obviously,and at the top of the cylinder head,hydraulic rod stroke is 1.901m,the host displacement is 1.988m.the difference value of 0.087m;The rotational speed of the spiral shaft is 23r/min,which is in line with the actual results,indicating that the AMESim model of the spiral conveyor is successfully constructed.At the same time,the higher the flow rate,the higher the rotation speed.While the flow phase,the load is not the same,the inlet pressure is basically the same,which proves that the assembly system has good speed regulation performance,and different quantity of pipe has little impact on the rotation speed.On the basis of analysis of the parameters of the collected,Matlab are binding to database.Based on the BP neural network analysis of shield machine hydraulic system fault diagnosis,fault convergence error can reach 0.01,meetting the requirements.