Study On Optimal Control Of Ground Source Heat Pump System Based On Matlab

For the investment of manipulating Ground-Coupled Heat Pump (GCHP)System is far more than the investment of initial equipment, the optimal control ofGCHP system is the main focus of GCHP system researches. GCHP system haveseveral characters as nonlinear, strong coupling, uncertainty and large energyconsumption. Based on the real time data of Olympic Forest Park, this thesis studiedthe energy saving optimal control of GCHP system.The basic thinking of GCHP system optimal control are build the energyconsumption model of every equipment and ensure the coupling relationship amongthem, and moreover, established the constrain of system energy. This thesis followingthe load exchange of user side air conditioning, used a proper optimal algorithm tominimize the complex energy consumption of GCHP system.First, this thesis analysis the basic principle of GCHP system in building2A03ofOlympic Forest Park. Energy consumption model of every part is correctlyestablished and the coupling relationship among these models are ensured. Throughthe study of energy consumption model, the key parameters which affect energyconsumption character. Because the coupling relationship among those models, it isnecessary to analysis the system energy consumption synthesized.Second, aiming at the nonlinear and uncertain characters, the multi-variablescontrol problem of GCHP system is realized. There are several disturbance variablesduring the manipulation of the system, as user side load, Borehole wall temperatureand so on. This thesis choose the water mass fow rate in buried side, the water massfow rate in user side and the temperature of water supply to user side as optimizationvariables and provide an effective multi-variables control method.Third, the neural network (NN) predictive model of GCHP system is establishedaccording to the real time data collected by experimental platform, facing thenonlinear, strong coupling and complexity problems of the GCHP system modeling.The NN model is established based on the nearest neighbor clustering method and thehidden neural number of NN is self-organized. The advantages of this NN model aresimple structure, fast convergence speed, low calculation complexity and completelyensure the modeling accuracy.Fourth, facing the energy consumption problem of GCHP system and considering the relationship between energy consumption and key parameters,construct the steady state optimization model and constrains of the system and anadaptive partial swarm optimization (APSO) algorithm is properly applied for theoptimization problem. After the GCHP system is modeled by RBFNN, the water massfow rate in buried side, the water mass fow rate in user side and the temperature ofwater supply to user side are optimized based on APSO via RBFNN model and realtime parameters.This thesis aiming at saving the energy consumption, via the study of energyconsumption model, RBFNN predictive model and system steady state optimizationmodel of GCHP system, designed a predictive optimization control method based onAPSO and nearest neighbor cluster method. The designed method is applied in theoptimal control of GCHP system, the results show that the energy consumption isreduced effectively under constrains of user side load.