Wang, Chao (2016-08). Dynamic Simulation and Control Design Tools for Heat Pump Systems. Doctoral Dissertation. Thesis uri icon


  • In heating, ventilating, and air conditioning applications, heat pump systems are devices that provide cooling and heating to residential and commercial buildings. Control-oriented models and control tools are essential to meet load requirements and to maximize coefficient of performance. The main challenges include complex system nonlinearities and multivariable interactions. This dissertation presents dynamic modeling and control design tools for heat pump systems. The first part introduces the dynamic modeling and development of a graphical user interface. The system-level dynamic model is developed based on component models, including compressors, expansion valves, finite control volume heat exchangers, etc. Validation results show that the models are effective tools to capture the dynamic characteristics in cooling, heating, and defrosting modes. A graphical user interface is developed for the dynamic models to facilitate parameter inputs and to generate steady-state operating conditions. The second part introduces energy optimal control and intelligent defrost control. Three energy optimal control algorithms are proposed: single-input, single-output control with fan scheduling; multiple single-input, single-output control; and multi-input, multi-output control. These strategies, with varying levels of complexity, range from decoupled proportional-integral-derivative control to linear quadratic Gaussian multivariable control. Validation results show that all three algorithms are effective for heat pump control. The goal of the intelligent defrost control is to maximize both system efficiency and human comfort during defrosting cycles. Experimental results show that there exists an optimal combination of frost growth time and defrost termination temperature. Under a higher moisture rate or lower ambient temperature, the optimal frost growth time becomes more sensitive.

publication date

  • August 2016