Time Series Analysis Module for the Energy Prediction and the Energy Saving Equipment

• Main Authors: Hui-I Su, 蘇慧倚
• Other Authors: Huang-Mu Lo
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/h5j2pb

碩士 === 朝陽科技大學 === 環境工程與管理系碩士班 === 96 === Abstract People in Taiwan have been with the economical growing and the increasing of the Gross Domestic Product (GDP) in recent years, so was the amount of relying on energy. With the increased numbers of the College students enrollment, the College administrative decision of updating the necessary research equipment and the dormitory was made. As a result, the energy and the water consumption have become two major issues of the College expenditure. The energy demand and its cost were greater than the water demand. Hence, this research was mainly focused on maintaining the energy usage under regular base without overly output concern. The paper investigates the energy demand of the National Chi Nan University with three major goals: (1) Collect historical data from 2004-2007 at 45 of them to analyze on campus energy consumption demand, in order to build up the monitoring system for the amount of the energy usage and its distribution infrastructure. (2) Implement the Time Series Analysis method for the amount of the energy prediction. Furthermore, utilizes the energy predictive function to optimize and sets up the goals of the proper amount of the energy usage in future. (3) Organize the management approach of the measures of the energy saving and further energy output control for the optimization the energy saving equipment. 45 of the total energy output data have been utilized for the energy prediction from this paper by the energy prediction of the multiple models of the time series analysis. This result of the average of the MAPE was 4.09%. And, the RMSPE was 5.11%. The R² of the energy demand was 0.93. The additive mode of the time series of the average MAPE was 4.36%. Moreover, the average RMSPE was 5.61%. The coefficient of the relationship R² between the practical and the predicable electricity was 0.92. According to the above R² data, the MAPE electricity data, and the RMSPE data, the evidence between the prediction and the real electricity demand were highly coherent. The result has shown that it was beneficial for the preventive control of the electricity by increasing offline computer calculation. In addition, the combination of the energy demand monitoring system with the energy saving equipment on campus can provide the effective information for the decision makers.