Study on model predictive control to minimize temperature change at multi positions in vertical plate with varying heat generation

Precise temperature control under noise-temperature change is required in high-precision manufacturing process. We analytically study the model predictive control to minimize temperature change at multi object positions in a two-dimensional vertical plate with varying noise-heat-generations. The noise-heatgenerations are that the heating-ON and OFF every 300 s, and it makes temperature change of 2.5° at the object positions in the plate without control. Control-heaters in the plate are controlled with the model predictive control method to minimize temperature change at the object positions. In this work, we study the method to minimize temperature change at the multi object positions in the plate without increasing number of the control-heaters. We found that temperature changes at the multi positions can be minimized, when heating rate distribution of the control-heaters is determined so that the steady temperature distribution at the control-heaters for only heating of the control-heat-generation comes close to that for only heating of the noise-heat-generation. Change of the object temperature at the multi positions using this method is 0.03°, which is 1/5 times smaller than that with uniform heating rate of the control-heaters (0.16°) and 1/80 times smaller than that without control (2.5°). We also study the effect of artificial error in the dynamic predictive model for one object position. Error in the dynamic predictive model causes various patterns of temperature change at the object position such as periodical error and oscillation error.