Data-driven Models For Energy-efficient Heating And Cooling: Air Source Heat Pump And Desiccant Systems As Case Studies

Boilers,electric water heaters,and vapor compression systems are commonly utilized in residential and commercial buildings in China to meet the demand for cooling and heating.The current heating and cooling systems are responsible for the environmental and energy crisis.The environmental and performance problems associated with the operation of these systems,however,emphasize the need for the development of alternate environment-friendly and cost-effective substitute technologies.To overcome the problems,air source heat pumps for heating and desiccant systems for cooling are among the most appropriate solutions for energy-efficient improvements in buildings.However,the operation of an air source heat pump on heating mode in winter is associated with a relatively low heating coefficient in terms of performance and heating capacity due to impact of ambient air temperature and frosting-defrosting operation.On the other hand,the main problem with desiccant cooling systems is the reduction in the total energy performance of the system when the optimum operating conditions are not set.The utilization of properly controlled air source heat pumps and desiccant systems in the HVAC system of a building can significantly reduce energy consumption.Data-driven models have indicated powerful strength in non-linear system modeling.These models can be utilized to test and validate different control strategies and to give the basis for saving energy and optimizing the operation of heating and air conditioning systems.This thesis has four main objectives.A combination of experimental and data-driven modeling approaches is adopted to address these objectives.To enhance the thermal performance of air source heat pumps at low ambient temperatures,the first objective is to develop a general yet accurate model capable of predicting the frost characteristics over wide ranges on different plate configurations under different conditions.A parametric study is performed using the validated models to provide a good insight into this study.The results in this thesis show that the developed models predict the data points within a reasonable error band,while the other available published correlations present higher deviations using the same dataset.For making heat pumps reasonably efficient at extracting low-grade thermal energy from the air in winter and preventing frost formation,the second objective is to evaluate the performance of a cross-flow closed-type heat-source tower.To meet this objective,a test facility to test this unit under low ambient temperature conditions is designed,and a robust heuristic model is developed.Based on the validated model,the effects of the critical parameters on the performance of this unit is elaborated.For the first time in the literature,the results indicate that with the increase in the air inlet temperature,air inlet humidity ratio,and solution inlet temperature,the air outlet humidity ratio gradually increases,while the increase in the solution concentration leads to a decrease in the air outlet humidity ratio.The third objective is to experimentally evaluate the performance of different desiccant wheels and subsequently optimize them to help designers enhance the performance of desiccant-wheel dehumidifier units.Several experiments have been performed to verify the applicability of the proposed methodology and validate the obtained results.The results show that there are optimum operating conditions for each wheel,depending on desiccant material,for the efficient operation of desiccant wheels and ultimately increasing the total energy performance of the desiccant cooling systems.The fourth objective is to comprehensively investigate the behavior of air source heat pump systems and subsequently improve their performance to reach the maximum performance of the system during the winter period.For this purpose,the experimental data from four real projects in Beijing have been utilized,and a hybrid and general model coupling different intelligent models is established.Two different control strategies are further proposed and evaluated.Results demonstrate that the percentage of growth in COP is 8.18%,compared to real operation,when the indoor temperature is set between 18 ~oC and 20 ~oC.