Model predictive control based on fuzzy linearization technique for HVAC systems temperature control

The Heating, Ventilating, and Air-Conditioning systems (HVAC systems) are typical nonlinear time-variable multivariate systems with disturbances and uncertainties. A new Mamdani fuzzy model predictive control strategy based on sum-min inference was proposed to control HVAC systems in this paper. The resolution relationship of two inputs and single output variables of the Mamdani fuzzy controller was obtained by its structure analysis. Then the fuzzy linearization predictive model at k+l sampling time on base of its resolution equation was designed. And at P ahead horizon predictive models were got. The predictive control strategy based fuzzy linearization predictive model was given and the procedure to implement the control algorithm was outlined. Finally simulation test results showed that the proposed fuzzy model predictive control approach is effective in HVAC systems temperature control applications. Compared with the conventional PID control, this fuzzy model predictive control algorithm has less overshoot and shorter setting time.