Research On Servo Direct Drive Pump Control Hydraulic System Based On Wavelet Neural Network

Hydraulic servo direct drive technology is a new type of transmission technology,with the continuous improvement of the servo motor control technology,the hydraulic servo direct drive technology has also been improved continuously.The hydraulic servo direct drive system can overcome the disadvantages of traditional electro-hydraulic servo valve for the high demand of the oil medium,poor anti-pollution and high loss of throttling.It has the double advantages of flexible servo motor control and large hydraulic output,the speed of the pump can be changed in real time according to the required power of the load to achieve efficient matching of power.With its advantages of low fault,high reliability,simple operation and control,small integration and environmental protection,it has been widely used in many fields,such as injection molding machine,hydraulic elevator,hydraulic press,etc.However,the dynamic response performance of such systems was low,which limits its wider application,many occasions put forward strict requirements on the control precision and dynamic performance,therefore,it is very important to study the dynamic control performance of the pressure position of the electro-hydraulic servo system of the direct drive pump control cylinder.In this dissertation,the servo direct drive pump control system theory and permanent magnet synchronous servo motor control theory were studied.the mathematical model of the motor vector control and the pump control cylinder system were established respectively,and the principle and realization method of voltage space vector control were introduced in detail.The model of servo direct drive pump control system was built in AMEsim,and the traditional control model was built in MATLAB/Simulink,through the interface setting,the joint simulation was implemented,and the influence of some parameters on the system performance was analyzed.The traditional PID control method has the disadvantage of slow response and poor stability,wavelet neural network controller was designed,the principle of wavelet neural network control and its implementation were described in detail.In MATLAB/Simulink,the wavelet neural network control model wasestablished through s-function module,finally,the joint simulation was implemented by AMEsim and MATLAB.In order to verify the correctness of the wavelet neural network control,a servo direct drive test bed was set up,the servo motor was adopted three-phase AC permanent magnet synchronous motor,DSP28335 was used as the control chip,and the experiment adopted two control methods,the PID control and wavelet neural network control,the response curve of the pressure and position under the rated load was analyzed,the results show that the response speed and the control precision of the servo direct drive pump control cylinder pressure and position control system can be greatly improved by using the wavelet neural network control.