predictive model, rolling optimization and feedback correction. It chooses unit step response as its predictive model, so that the modeling process is simplified. In addition, good control and decoupling effects could be possessed by means of selection suitable parameters.
In this paper, the environmental test device is introduced briefly and the existing problems are showed. Then multivariable predictive control algorithm is presented particularly, including multivariable auto-balance system predictive control algorithm and multivariable auto-unbalance system predictive control algorithm. Next, system modeling process and corresponding system model are proposed. Further, the multivariable predictive control algorithm is applied to the temperature and humidity control system of the environmental test device. Finally, the simulation results are compared.
Results of the simulation show that multivariable predictive control algorithm could be used in those multivariable system like the temperature and humidity control system of the environmental test device and the control result would be more satisfactory than that of the routine PID control.