Optimization And Software Development For Air Handling Subsystem In The Heating Ventilating And Air-Conditioning System

Building energy consumption accounts for 30% of total energy consumption, and HVAC system energy consumption accounted for 60% of building energy consumption. While the energy cost of the air handling subsystem is about 35 percent of the HVAC system. It can efficiently reduce the energy consumption of the HVAC system.In general, the design of the HVAC system is based on the assumption that system is working on the largest loads, yet the control of HVAC in the building automation is run under the most optimal set-point of the rated loads. But the key problem is to how to deal with the optimal set-point of scattered air handlers which own plenty of parameters under the situation of varied external conditions and internal loads.In this thesis, the method of modeling was studied by HVAC air handling system and the process aspect of the energy consumption characteristics of the key equipment. According to table modeling, air handler engineering model was built. Furthermore exact precision of the model obtained, the modeling of air handler was built by generalized regression neural network. The library simulation showed that it not only reflected the heat transferred character of the air handler, but also received the exact precision optimal set-points of air handler, such as supplied-air temperature and supplied-air pressure.On this basis, this thesis made engineering modeling and neural network modeling of air and then the online optimization was operated with table search optimization and EP with penalty function. Those methods solved the most optimal problem of air handler with multi-variable, non-linear, equality and inequality constraints.On based VC++6.0 and SQL server2000 Database, the software modules of modeling and optimization were developed. And the module of the table modeling and the table search optimization was completed. At the same time the neural network modeling and evolutionary programming optimization method were validated. The library simulation showed that EP method could decide the most optimal control effect of the air handler under varied loads.