Electrical Energy Consumption Estimationof Main Equipment For HVAC Based On Working Parameters Selecting

Energy shortage plays an important role in economic development, social stability and national security. Therefore, energy conservation and emission reduction have been widely concerned. Nearly 30% of total energy consumption owes to building energy consumption.Heating, ventilating and air conditioning(HVAC) energy consumption holds about 40 to 60% in the total building energy consumption which has great potential for energy savings. According to the assessment standard for green building, the electricaldevices in the building are grouped for electrical energy measuring which failed to electrical energy measuring for every device especially for large power equipment. Building automation system is applied to monitoring the working state of HVAC system in public buildings. Therefore, the modeling of main equipment electrical energy consumption estimation based on working parameters will benefit to let consumers know the energy usage situation of their main devices more clearly, realize energy saving and provide suggestions for efficient operation.On the basis of the principle of the typical HVAC system,the thesis proposes the main equipment electrical energy consumption forecasting method which based on working parameters. The main research contents are as follow:(1) The independent characteristic vectors are selected by two level pattern correlation analysis for energy consumption estimation model. The working parameters which correlate with electrical energy consumption strongly are selected by the first level pattern correlation analysis, and selected the micro-related working parameters by the second level pattern correlation analysis.(2) The modeling of main equipment electrical energy consumption estimation was built base on working parameters.The method of electrical energy consumption forecasting for chiller, secondary chilled water pump and air handling unit(AHU)used support vector machine(SVM).Comparison of the cross validation, genetic algorithm and particle swarm optimization, themost suitable optimization algorithm is used for SVM. For single AHU, the method of electrical energy consumption forecasting is based on fan speed and polynomial fit. The simulation experimentshave verified the effectiveness of the methods.(3) This paper puts up a simple physical hardware experiment to achieve energy estimation. The experiment based on STM32F407 microprocessor and MATLAB GUI, which achieved the method of electrical energy consumption for secondary chilled water pump. The SVM C source code was compiled by library function and graphical user interface was designed by MATLAB GUI.