abstract: The United States building sector was the most significant carbon emission contributor (over 40%). The United States government is trying to decrease carbon emissions by enacting policies, but emissions increased by approximately 7 percent in the U.S. between 1990 and 2013. To reduce emissions, investigating the factors affecting carbon emissions should be a priority. Therefore, in this dissertation, this research examine the relationship between carbon emissions and the factors affecting them from macro and micro perspectives. From a macroscopic perspective, the relationship between carbon dioxide, energy resource consumption, energy prices, GDP (gross domestic product), waste generation, and recycling waste generation in the building and waste sectors has been verified. From a microscopic perspective, the impact of non-permanent electric appliances and stationary and non-stationary occupancy has been investigated. To verify the relationships, various kinds of statistical and data mining techniques were applied, such as the Granger causality test, linear and logarithmic correlation, and regression method. The results show that natural gas and electricity prices are higher than others, as coal impacts their consumption, and electricity and coal consumption were found to cause significant carbon emissions. Also, waste generation and recycling significantly increase and decrease emissions from the waste sector, respectively. Moreover, non-permanent appliances such as desktop computers and monitors consume a lot of electricity, and significant energy saving potential has been shown. Lastly, a linear relationship exists between buildings’ electricity use and total occupancy, but no significant relationship exists between occupancy and thermal loads, such as cooling and heating loads. These findings will potentially provide policymakers with a better understanding of and insights into carbon emission manipulation in the building sector. / Dissertation/Thesis / Doctoral Dissertation Civil, Environmental and Sustainable Engineering 2018

本論文一冊，六章，二十四節，七萬字。 第一章：緒論：介紹本論文之研究背景、研究動機與目的、研究限制。 第二章：文獻探討：探討有關非計量多元尺度與聯合分析之文獻；並介紹國內相關論 文之研究。 第三章：研究方法：介紹問卷之設計與其分析之方法；說明抽樣之方法與實際抽樣之 過程；觀念架構。 第四章：結果分析（一）：解釋非計量多元尺度方法分析之結果，並以不同事前區隔 之群體來分析其聯合空間圖。由此資料可了解不同區隔之消費者在各屬性上，對各廠 牌冷氣機之印象。 第五章：結果分析（二）：解釋應用聯合分析方法分析資料之結果，並以不同之區隔 來分析冷氣機各屬性之成份效用值，企圖找出最佳化之產品設計。 第六章：結論與建議：由競爭分析與產品屬性分析之結果，歸納出消費者對各品牌冷 氣機之印象及各屬性之重視程度，進而建議各廠牌在行銷策略上應注意之事項。

This Ph.D. thesis deals with problems about voltage dips and short voltage interruptions, generally named as voltage events, which origin in power supply networks and have a negative influence for all connected electric appliances. In this thesis single phase appliances are considered. These problems closely relate with area of electromagnetic compatibility which solve all questions about correct operation of different types of electric appliances during electromagnetic disturbances impact. Voltage events are ones of the many types of electromagnetic disturbances. The connection between disturbance sources and sensitive electric appliances is created by power supply lines. The immunity of electric appliances to voltage dips and short interruptions is currently tested via voltage dips with strictly defined parameters which are intended by class of electromagnetic environment in which the usage of electric appliance is recommended. During immunity tests the rectangular shape of voltage dips is preferred. The main descriptive parameters of testing voltage events are the residual voltage and the event time duration. But voltage dips and short interruptions defined by this way do not closely relate with parameters of real voltage dips and interruptions occurred in public supply system where parameters of voltage dips are variable. Moreover in the power supply system there are many of others voltage parameters which can have a significant influence to immunity level of connected electric appliances. This Ph.D. thesis also deals with finding more voltage event parameters. Although the voltage events occurrence in the power supply system is not limited and voltage events are considered only as informative voltage parameter it is important to monitor voltage events occurrence. The monitoring device has to be able to operate for ling time period and it has to detect parameters of voltage events with adequate accuracy. The accuracy of detected event parameters and the detection delay depends on the detection algorithm characteristics. That is why the part of this thesis relates with a comparison of several detection algorithms and their abilities to correct detection of voltage event parameters. The main purpose of this thesis is the proposal of connection between classification of voltage dips and short interruptions occurred in power supply system with the classification of electric appliances immunity to these voltage events. On the base of many of provided electric appliances immunity tests and also on the base of long time period voltage events monitoring the special compatibility levels are proposed in this thesis. The observation of proposed compatibility levels will bring the increasing level of reliable operation of all connected electric appliances.