| Summary: | 碩士 === 國立臺北科技大學 === 電機工程系 === 106 === The electric bill is higher and higher led environmental consciousness to rise, the office building or general small and medium-sized industrial and commercial users use a high proportion of air conditioners, and the air conditioning load in summer can be as high as 40-50% of the total electrical load of the building in our country, there is worth exploring issues to saving energy and maintain indoor environment temperature at the same time, because the air conditioners power consumption accounts for 48% of the buildings total electricity consumption. Therefore, this paeper propose small size air conditoners control strategy to saving energy but maintain resident’s comfort, first through the Recurrent Neural Networks and recursive strategy to build the multi-step indoor temperature forecasting model, and thorugh the Adaptive Particle Swarm Optimization to optimize compressor’s duty cycle, the objectives function was added weights, through ajust the weights to change output result, make the optimiaztion system fexible can handle different environment and different season achieve saving energy and maintain indoor environment temperature at the same time.
|
|---|
