The temperature control system is required to have both cooling and heating actuators to achieve a high-precision thermostatic indoor air environment of industrial buildings. The coupling of the two actuators and the generation of redundant energy consumption leave a sufficient gap to be optimized. Combining the Linear Quadratic Regulator (LQR) approach with the sensitivity decoupling structure, we construct a dual-stage PI controller for this type of dual-input single-output system. Meanwhile, a proposed energy management technique for lowering energy consumption can alter the controlled quantities adaptively and synergistically. Accordingly, the multi-objective teaching–learning-based optimization algorithm is implemented to determine the optimal parameters and establish a balance between temperature stability and energy consumption. Simulation and experimental platforms are built to verify the design effect. Compared with four typical single-stage PI controller tuning algorithms, the proposed method has better output temperature stability with an improvement of the index by at least 14%. And the results reveal that the control input is decreased by 45.6% when the energy management technique is implemented, while the stability index of the final output temperature is maintained at about the same level.
Read full abstract