Model Predictive Control And System Energy Saving For Liquid Desiccant Air Conditioning

HVAC(Heating,Ventilation and Air-conditioning)plays an important role in supplying a comfortable thermal and humid environment for buildings.At present,people not only pursue comfortable indoor air environment,but also pay more attention to energy efficiency.Traditional cooling dehumidification air conditioning limits the utilization of natural cooling sources and the improvement of refrigeration equipment efficiency due to the uniform treatment of sensible load and latent load of process air.LDAC(Liquid Desiccant Air-Conditioning)has been a hot spot in recent years for the reason that it is able to realize the independent control of temperature and humidity.It can also improve the low grade energy efficiency and purify indoor air environment.At present,people's requirements for indoor comfortable environment and energy saving of air conditioning make dynamic control and real-time optimization of energy consumption particularly important.The existing control strategies cannot take control performance and system energy efficiency into consideration at the same time.In this paper,dynamic control and energy efficiency optimization of LDAC system are studied.The main research work and contributions of this paper are as follows:(1)Dynamic modeling of a dehumidifier for LDAC systemConsidering LDAC's characteristics of nonlinearity and coupled thermal and wet,the nonlinear mapping relation between the inputs and outputs of a dehumidifier is established by ANFIS(Adaptive Neuro-Fuzzy Inference System)method to obtain the dynamic model.The model verification shows that the relative errors are less than 2%for outlet air temperature and less than 4%for outlet air humidity ratio,respectively.The results meet the requirement of dynamic prediction of the dehumidifier outlet temperature and humidity.(2)Control study for LDAC system by EMPC and energy efficiency analysisThe objective function including system control and energy consumption has been constructed by EMPC(Economic Model Predictive Control)strategy to obtain the finite time horizon dynamic optimization problem.The optimization problem is solved by genetic algorithm.The control performance and energy efficiency improvement of LDAC system are studied under two different work conditions.The simulation results show that the EMPC outperforms PI control with smaller tracking error,faster response,and higher energy efficiency.The set value of air state and the direction of step change have important effects on the energy-saving potential of the system.Increasing the COP(coefficient of performance)of the chiller contributes to improvement of the LDAC system energy efficiency.(3)Control study for LDAC system by DMPCThe Control performance of DMPC(Distributed Model Predictive Control)on LDAC has been studied by comparing with CMPC(Centralized Model Predictive Control)and DeMPC(Decentralized Model Predictive Control).The results show that the DMPC can achieve almost the same control performance as the CMPC when the parameters are set properly.However,the DeMPC may not achieve satisfying control effect in large complex system due to ignoring the coupling interactions among subsystems.