Pressure control method for low-pressure vacuum systems based on mismatch-resistant Smith predictor compensation

Abstract

To achieve efficient and high-precision pressure control for low-pressure vacuum systems, a pressure control method based on mismatch-resistant Smith Predictor (SP) compensation is proposed. First, a control scheme consisting of an SP and a proportional–integral (PI) controller is designed for the internal loop of the system, combined with feedforward control to form an integrated control structure. This approach effectively addresses the issue of sluggish response in control systems caused by the significant delay characteristics of low-pressure vacuum systems. Second, to address the common issue of model parameter mismatch, which arises from inaccurate model construction and parameter identification, a model parameter mismatch compensation mechanism is designed and incorporated into the pressure feedback collection process. Proportional–derivative correction is applied to the feedback signal after SP compensation to mitigate the control signal fluctuations caused by model mismatch; thus, the pressure control performance of the low-pressure vacuum system can be enhanced. Compared to the Mac-pid and AFFPI controllers, the introduction of the SP controller leads to an average reduction in settling time of 41.16 and 12.58 seconds, the overshoot is almost eliminated, and a reduction in integral of time-weighted absolute error (ITAE) by 1304.14 and 74.09, respectively, in the simulation environment. Compared to the SP control method that does not account for model parameter mismatch, the settling time is reduced by 4.36 seconds, overshoot is reduced by 23.20%, ITAE decreases by 13.99, and the variance of the control signal is significantly reduced by 72.37%, in the real experimental environment. These results indicate that the proposed control method can effectively suppress model parameter mismatch issues caused by inaccurate modeling and parameter identification, reduce control signal fluctuations, avoid the performance degradation of traditional Smith Predictor control methods under model mismatch, and achieve efficient and high-precision pressure control in low-pressure vacuum systems.