Research On Control Problem And Fault Diagnosis Of High-performance Electrohy-draulic Servo Turntable

This paper establishes the simulation model of high-performance electro-hydraulic servo turntable system and designs the controller. Based on the above, development and validation of the fault diagnosis algorithm which could be used in electro-hydraulic servo system are presented.Firstly, a mathematical model of the controlled object of middle-axis is established according to kinematics and kinetics laws. And based on this, a block diagram of the system structure is set up. And according to the actual system commissioning requirements, the frequency characteristic test program is written. Based on completion of the hardware selection of system, a good control strategy is particularly important. And it becomes the main measures to improve the performance of the system. According to the actual engineering situation, the classic control algorithm is adopted, namely velocity and acceleration feedback, position loop control, command signal differential feed-forward, disturbance input feed-forward, flutter signal link.Then, according to the failure analysis including fault pattern, cause of fault, etc of the three important components namely electro-hydraulic servo valve, hydraulic motor and angular displacement sensor in electro-hydraulic servo system, various kinds of common faults in the system are summarized. And based on the analysis above, the principles of two fault diagnosis algorithms(multi-swarm particle swarm algorithm, BP neural network algorithm) applied in electro-hydraulic servo system are introduced. In order to guarantee the safety of system, and reduce maintenance and operating costs, the exploitation and validation of fault diagnosis method by means of simulation is adopted. The corresponding parameters of structure block diagram are artificially modified to simulate typical faults of electro-hydraulic servo valve, hydraulic motor, angular displacement sensor. Referring to the different characteristics of the intelligent algorithms above, diagnostic methods are proposed. And two aspects namely diagnosis effectiveness and feasibility of the methods above are compared.