近年來，科技的進步使得能源的消耗大幅提升，然而生成與使用能源的同時亦對環境造成重大的污染，因此節能的議題躍升成為各界廣為討論的重點之一。根據經濟部能源局的統計資料，工業佔據臺灣37.41%的總能源消耗量與53.56%的耗電量，為臺灣最大的能源用戶。為此，許多企業致力於發展節能方法與標準以節省耗電量與用電支出，並以成為對環境友善之企業為其目標。
為降低電量的消耗，部分製造業者選擇針對其空調系統進行調整，空調系統為一用於調節辦公大樓、潔淨室、半導體廠等溫度之設備，其佔據工廠約30%耗電量。空調系統由冷卻水塔、冰水主機、AHU與水泵等設備組成，其中又以冰水主機所消耗之電量為冠，其佔據廠務設備60%耗電量。為在節能的同時不致影響生產流程，操作人員會將整體冰水系統之負載控制於一定範圍之內。在過去，冰水主機的調整多仰賴平均負載法或操作人員之經驗法則，使得人為操作差異的情況屢見不鮮，缺乏一系統化決策模式。本研究透過資料挖礦手法發展冰機節能最佳化架構，以最佳化冰機操作為目標，在不影響產能並達到外界冰水需求的前提下最小化冰水機組耗電量，此架構結合多項分析方法，包含時間序列模型(Box et al., 2015)、多變量雲型適應迴歸(Multivariate Adaptive Regression Splines; MARSplines)(Friedman, 1991)、切分點分析(Lorden, 1971)，建立找尋冰機最佳組合之機制，利用冷凍噸需求預測模型結合外氣溫度預測資料以提前預估未來現場冰水需求，透過冰機負載預測模型分析目前組合可製造之冰水量、能耗及其運作效率，而後藉由冰機健康指標模型刪去不適合調度之機台，最後模擬各可行解之效能以求得最佳解。本研究以臺灣某消費性電子產品公司之製造廠進行實證以檢驗模型效度，利用冰水主機運轉歷史數據萃取過去領域專家之專業並結合本研究架構，根據未來需求與現況供給的平衡，提供冰機操作人員加減機建議及組合表現，並持續監控機台健康狀況以事先預防重大故障發生，建立一穩健且系統化之調度標準。

In decades, advances in technology make humans rely on energy resources significantly which arises people’s attention toward energy-saving issues as well. While energies are generated and consumed, the environment is polluted inevitably. According to statistical results provided by Bureau of Energy, Ministry of Economic Affairs (2018), industries are the primary consumer of power in Taiwan, which consumed 37.41% of total energy usage and 53.56% of total electricity usage in 2017. Thus, by means of energy-saving approaches, companies can cut down on cost and accomplish eco-friendly businesses in the meanwhile.
Manufacturing industries choose to reduce electricity consumption of the air conditioning system, which consumes 30% of the total electricity usage in the factory. The air conditioning system is used to regulate the temperature of plants and offices, and it consists of cooling towers, chillers, air handling units (AHUs), water pumps, etc. Chiller is one of the most electricity-consuming parts, which accounts for 60% electricity consumption of factory facilities. To conserve electricity and costs without affecting production processes, operators are devoted to keeping the loading and the performance of the multi-chiller system within a particular range. Conventionally, operations of chillers mostly rely on equal chiller loading method (Braun et al., 1989) or engineers’ experiences, which may neglect chiller differences and bring about anthropogenic effects. In this study, chiller adjustments would be optimized to realize the effective electricity savings. The approach based on the cooling load forecasting model, the chiller system partial load ratio prediction model, and the chiller health detection to derive the optimal chiller allocation through time series model (Box et al., 2015), multivariate adaptive regression splines model (Friedman, 1991), and change-point analysis (Lorden, 1971). The primary objective of this study is to minimize the total electricity consumption of chillers on condition that the cooling load is satisfied. Regarding the practical applications, this research provides operators with the optimal adjustment suggestions. An empirical study was conducted in an electronics company to demonstrate the validity of the proposed approach.

Notation Table i
Content ii
List of Figures iv
List of Tables vi
Chapter 1 Introduction 1
1.1 Background and Motivation 1
1.2 Research Objectives 4
1.3 Organization of Thesis 4
Chapter 2 Literature Review 5
2.1 The Air Conditioning System 5
2.2 Chiller Operation Optimization Methodology 8
2.3 Machine Health 11
2.4 Methodology 12
2.4.1 Time Series Model 12
2.4.2 Multivariate Adaptive Regression Splines 13
Chapter 3 Research Framework 15
3.1 Problem Definition 17
3.2 Data Preparation 17
3.2.1 Data Collection 17
3.2.2 Data Preprocessing 18
3.3 Model Development 19
3.3.1 The Cooling Load Forecasting Model 19
3.3.2 The Chiller System PLR Prediction Model 20
3.3.3 The Chiller Health Detection 21
3.4 Model Validity 23
3.5 Evaluation and Interpretation 23
Chapter 4 Empirical Study 25
4.1 Problem Definition 26
4.2 Data Preparation 28
4.2.1 Data Collection 28
4.2.2 Data Preprocessing 28
4.3 Model Development 29
4.3.1 The Cooling Load Forecasting Model 29
4.3.2 The Chiller System PLR Prediction Model 29
4.3.3 The Chiller Health Detection 31
4.4 Model Validity 39
4.5 Evaluation and Interpretation 41
Chapter 5 Conclusion 46
5.1 Summary and Contribution 46
5.2 Future Research 47
References 49

ASHRAE guideline 14-2002 Measurement of energy and demand savings (2002), ASHRAE Standards Committee.
Bercu, S. and Proïa, F. (2013), "A SARIMAX coupled modelling applied to individual load curves intraday forecasting," Journal of Applied Statistics, Vol. 40, No. 6, pp. 1333-1348.
Bourdouxhe, J., Grodent, M., and Lebrun, J. (1997), "HVAC1 Toolkit: algorithms and subroutines for primary HVAC system energy calculations," The American Society of Heating, Refrigerating and Air Conditioning Engineers, Vol., No., pp.
Box, G. E., Jenkins, G. M., Reinsel, G. C., and Ljung, G. M. (2015), Time series analysis: forecasting and control. John Wiley & Sons.
Braun, J., Klein, S., Beckman, W., and Mitchell, J. (1989), "Methodologies for optimal control of chilled water systems without storage," ASHRAE transactions, Vol. 95, No., pp. 652-662.
Bureau of Energy, Ministry of Economic Affairs. (2018) "Statistical Monthly Report of Energy Usage." Retrieved 5/20, 2018, from http://www.moeaboe.gov.tw/
Chang, Y.-C. (2004), "A novel energy conservation method—optimal chiller loading," Electric Power Systems Research, Vol. 69, No. 2, pp. 221-226.
Chang, Y.-C., Chen, W.-H., Lee, C.-Y., and Huang, C.-N. (2006), "Simulated annealing based optimal chiller loading for saving energy," Energy Conversion and Management, Vol. 47, No. 15-16, pp. 2044-2058.
Chang, Y.-C., Lin, J.-K., and Chuang, M.-H. (2005), "Optimal chiller loading by genetic algorithm for reducing energy consumption," Energy and Buildings, Vol. 37, No. 2, pp. 147-155.
Chen, J. and Gupta, A. (1999), "Change point analysis of a Gaussian model," Statistical Papers, Vol. 40, No. 3, pp. 323-333.
Chen, J. and Gupta, A. K. (2011), Parametric statistical change point analysis: with applications to genetics, medicine, and finance. Springer Science & Business Media.
Friedman, J. H. (1991), "Multivariate adaptive regression splines," The annals of statistics, Vol., No., pp. 1-67.
Gordon, J. and Ng, K. C. (1995), "Predictive and diagnostic aspects of a universal thermodynamic model for chillers," International Journal of Heat and Mass Transfer, Vol. 38, No. 5, pp. 807-818.
Guideline, A. (2002), "Guideline 14-2002, Measurement of Energy and Demand Savings," American Society of Heating, Ventilating, and Air Conditioning Engineers, Atlanta, Georgia, Vol., No., pp.
Haberl, J. S., Claridge, D., and Culp, C. (2005), "ASHRAE's guideline 14-2002 for measurement of energy and demand savings: How to determine what was really saved by the retrofit," Vol., No., pp.
Hao, X., Zhang, G., Chen, Y., Zou, S., and Moschandreas, D. J. (2007), "A combined system of chilled ceiling, displacement ventilation and desiccant dehumidification," Building and Environment, Vol. 42, No. 9, pp. 3298-3308.
Heng, A., Zhang, S., Tan, A. C., and Mathew, J. (2009), "Rotating machinery prognostics: State of the art, challenges and opportunities," Mechanical systems and signal processing, Vol. 23, No. 3, pp. 724-739.
Huang, X., Xu, J., and Wang, S. (2010), "Operation optimization for centrifugal chiller plants using continuous piecewise linear programming," Proceedings of Systems Man and Cybernetics (SMC), 2010 IEEE International Conference on.
Hydeman, M. and Gillespie Jr, K. L. (2002), "Tools and techniques to calibrate electric chiller component models/discussion," ASHRAE transactions, Vol. 108, No., pp. 733.
Kusiak, A., Li, M., and Tang, F. (2010), "Modeling and optimization of HVAC energy consumption," Applied Energy, Vol. 87, No. 10, pp. 3092-3102.
Lee, T.-S. (2004), "Thermodynamic Modeling and Experimental Validation of Screw Liquid Chillers," Ashrae Transactions, Vol. 110, No. 1, pp.
Lee, W.-S. and Lin, L.-C. (2009), "Optimal chiller loading by particle swarm algorithm for reducing energy consumption," Applied Thermal Engineering, Vol. 29, No. 8, pp. 1730-1734.
Lehmann, E. L. and Romano, J. P. (2006), Testing statistical hypotheses. Springer Science & Business Media.
Lorden, G. (1971), "Procedures for reacting to a change in distribution," The Annals of Mathematical Statistics, Vol., No., pp. 1897-1908.
Lu, L., Cai, W., Xie, L., Li, S., and Soh, Y. C. (2005), "HVAC system optimization—in-building section," Energy and Buildings, Vol. 37, No. 1, pp. 11-22.
Mills, T. C. (1991), Time series techniques for economists. Cambridge University Press.
Pham, H. T. and Yang, B.-S. (2010), "Estimation and forecasting of machine health condition using ARMA/GARCH model," Mechanical Systems and Signal Processing, Vol. 24, No. 2, pp. 546-558.
Qureshi, T. and Tassou, S. (1996), "Variable-speed capacity control in refrigeration systems," Applied Thermal Engineering, Vol. 16, No. 2, pp. 103-113.
Soliman, S., Persaud, S., El-Nagar, K., and El-Hawary, M. (1997), "Application of least absolute value parameter estimation based on linear programming to short-term load forecasting," International Journal of Electrical Power & Energy Systems, Vol. 19, No. 3, pp. 209-216.
Sreedhara, P. and Haves, P. (2001), "Comparison of chiller models for use in model-based fault detection," Vol., No., pp.
Sezgan, O., B. Smith, and M. Moezzi (1996), “Building performance evaluation and tracking,” Berkeley, Calif.: Lawrence Berkeley National Laboratories .
Wang, W. Q., Golnaraghi, M. F., and Ismail, F. (2004), "Prognosis of machine health condition using neuro-fuzzy systems," Mechanical Systems and Signal Processing, Vol. 18, No. 4, pp. 813-831.
Wu, Y.-C., (2017), “ Power consumption saving analysis and optimization model for chillers of factory facility and the empirical study.” Master's Thesis of Department of Industrial Engineering and Engineering Management. Hsinchu: National Tsing Hua University, 48 pp.
Yan, R. and Gao, R. X. (2007), "Approximate entropy as a diagnostic tool for machine health monitoring," Mechanical Systems and Signal Processing, Vol. 21, No. 2, pp. 824-839.
Yik, F. W. and Lam, V. K. (1998), "Chiller models for plant design studies," Building Services Engineering Research and Technology, Vol. 19, No. 4, pp. 233-241.
Zhang, L., Wen, J., and Chen, Y. (2017), "Systematic Feature Selection Process Applied in Short-Term Data-Driven Building Energy Forecasting Models: A Case Study of a Campus Building," Proceedings of ASME 2017 Dynamic Systems and Control Conference.