Research On Adaptive Control For Hydraulic Servo System Of Rolling Mill With Input Saturation(Engineering Design)

Due to its advantages of fast response and strong load capacity, the hydraulic servosystem has been extensively used in many industrial applications such as rolling mill,robot，aerospace, and etc. However, it should be noted that the hydraulic servo system hasthe inherent nonlinearity characteristic. In particular, the control input is saturation due tophysical limits of the actuator. At the same time, the system has uncertainties includingparameters perturbation and external load disturbance. In addition, the system parameterschange abruptly when the rolling mill driven by the hydraulic servo system contacts theload. All these bring certain difficulties to the control design with high performancerequirements. Consequently, it is very desirable to further search the hydraulic servoposition system of rolling mill with above issues, and the main contents of this thesismainly include the following aspects.Firstly，a backstepping control scheme based on neural networks disturbance observer isproposed for the hydraulic servo system with parameter uncertainties, disturbance andinput saturation. When the input saturation does not occur, the neural network disturbanceobserver is designed to estimate the composite disturbance of the system, and compounddisturbance is compensated in the backstepping design. Then the conclusion is extended tothe situation that the control input is saturation. Simulation results show the designedcontroller can make the system output track accurately the desired position.Secondly, a robust adaptive control scheme is developed combining the dynamicsurface control with L2-gain control for the hydraulic servo system with parameteruncertainties, external disturbance and input saturation, the adaptive control is used toapproximate the unknown uncertainties, the robust control ensure that L2-gainperformance can be achieved. When the control input is saturation, the auxiliary system isdesigned to reduce the effects of input saturation. It can be proved that the closed-loopsystem is bounded stable, and has stronger robustness to parameter uncertainties.Finally, an adaptive backstepping control method based on multi-model switching ispresented for the hydraulic servo systems of rolling mill with parameters uncertainties, change abruptly and input saturation. Based on the multiple model sets of the hydraulicservo system, the adaptive backstepping controllers are designed for all subsystem, andthe auxiliary system is designed to reduce the effects of input saturation. With the help ofcommon Lyapunov function, the proposed method can guarantee the multi-modelswitched system is stable. Finally the effectiveness of the proposed algorithm is verifiedby simulation results.