碩士 / 國立勤益科技大學 / 冷凍空調系 / 105 / This study uses double-effect heat pump system, the cold zone controls the seawater temperature to reach the lodgeable stability condition for lobsters, the hot zone uses the hot water at 35℃-40℃ recovered by the heat pump, the recovered hot water is added in the hot water dishwasher, and the water is heated by hot water dishwasher to 80℃ for washing dishes. Several important parameters which influence the heat pump system are analyzed, including hot water recovered temperature, fish bowl cold water temperature and indoor temperature, and the hot water temperature recovered by double-effect heat pump system is predicted by back-propagation neural network integration, training, simulation and goal analysis.
This study uses artificial neural back propagation network and MATLAB for simulation, validation and analysis. In answer to the field load change, the data collected by field sensors are analyzed, simulated and validated, and an effective prediction module is built according to the classroom indoor temperature difference, and the hot water recovered temperature is predicted to decide on increasing or reducing the load of double-effect heat pump system.
An artificial neural back propagation network module is constructed to analyze and classify the historical parameters of heat pump host, so as to analyze the effect of the module on predicting various parameters. The experimental simulation of hot water temperature shows when the indoor temperature difference is greater than 1℃ during two days, the forecast result of the module is bad. Therefore, in the construction of network module, the regularity of the effect of indoor temperature shall be found, and the required modules are built within 0.5℃ of two days' indoor temperature difference, and the recovered hot water is predicted by using the module, the double-effect heat pump system is adjusted. Finally, according to the experimental results, the electric cost is lower than the original system by 10% after adjustment.
| Identifer | oai:union.ndltd.org:TW/105NCIT5680007 |
| Date | January 2017 |
| Creators | Tsung-jung Hsu, 許宗榮 |
| Contributors | Kuang-cheng Yu, 余光正 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 75 |
Page generated in 0.0022 seconds