[학위논문]AMI 기반 공동주택 수용가의 최적전기요금제 선택모형 연구 analysis of electricity usage patterns of ami-based apartment houses and development of models for selecting optimal electricity rates for customers원문보기
In a situation where studies and demonstrations that require changes in the domestic and overseas residential electricity rates are continuously being conducted, this study derives 10 electricity usage patterns through analysis of AMI data for 5 years in apartment complexes in large cities. Pattern ...
In a situation where studies and demonstrations that require changes in the domestic and overseas residential electricity rates are continuously being conducted, this study derives 10 electricity usage patterns through analysis of AMI data for 5 years in apartment complexes in large cities. Pattern analysis used the k-Means algorithm, which is most widely used in Unsupervised Learning of machine learning. The main characteristics of each electricity rates were analyzed by applying the derived electricity usage pattern to six domestic and overseas electricity rates. Through this process, the main factors necessary for the development of an optimal electricity rate selection model were derived. Six rates were selected among domestic and international rate plans, and a model was developed so that individual households can choose the lowest rate plan among these rates. In this study, the analysis methodology was separately applied to predict and analyze the optimal electricity rate at the same time. The Random Forest model was applied to predict the total of 6 minimum rates. A Decision tree model was applied to calculate the rules for selecting individual rates. As a method to verify the accuracy of the optimal electricity rate selection model, 50% of the total data is used for model development, and the remaining 50% of the data is input into the developed model, and the selected optimal rate plan prediction result was compared with the actual value, and then the accuracy of the model was verified. As a result, Out of 9,125 data, 8,634 were predicted as optimal rates, and the accuracy was calculated as 94.7%. In addition, the analysis of the individual rate selection path of the six TOU rates showed very high prediction accuracy, and the sensitivity and precision also showed high accuracy. Based on the results of such a prediction model, The optimal rate plan selection checklist was proposed to select the optimal rate with a combination of 5 major factors for the 6 TOU rates. While thesis related to the application of the existing optional rates was first approached and studied from the perspective of government policy and utility, this paper thoroughly developed and proposed a model to select the optimal (lowest) rate from the perspective of electricity users. With the advancement of technology related to AMI metering based on electricity rate, it was possible to derive a pattern through detailed analysis of electricity usage by time period, but it is necessary to expand the research to general houses including apartments nationwide in the future. If research and demonstration taking into account the impact of climate are conducted, it is expected that the development of an optimal rates that can satisfy both utilities and customers will be made.
In a situation where studies and demonstrations that require changes in the domestic and overseas residential electricity rates are continuously being conducted, this study derives 10 electricity usage patterns through analysis of AMI data for 5 years in apartment complexes in large cities. Pattern analysis used the k-Means algorithm, which is most widely used in Unsupervised Learning of machine learning. The main characteristics of each electricity rates were analyzed by applying the derived electricity usage pattern to six domestic and overseas electricity rates. Through this process, the main factors necessary for the development of an optimal electricity rate selection model were derived. Six rates were selected among domestic and international rate plans, and a model was developed so that individual households can choose the lowest rate plan among these rates. In this study, the analysis methodology was separately applied to predict and analyze the optimal electricity rate at the same time. The Random Forest model was applied to predict the total of 6 minimum rates. A Decision tree model was applied to calculate the rules for selecting individual rates. As a method to verify the accuracy of the optimal electricity rate selection model, 50% of the total data is used for model development, and the remaining 50% of the data is input into the developed model, and the selected optimal rate plan prediction result was compared with the actual value, and then the accuracy of the model was verified. As a result, Out of 9,125 data, 8,634 were predicted as optimal rates, and the accuracy was calculated as 94.7%. In addition, the analysis of the individual rate selection path of the six TOU rates showed very high prediction accuracy, and the sensitivity and precision also showed high accuracy. Based on the results of such a prediction model, The optimal rate plan selection checklist was proposed to select the optimal rate with a combination of 5 major factors for the 6 TOU rates. While thesis related to the application of the existing optional rates was first approached and studied from the perspective of government policy and utility, this paper thoroughly developed and proposed a model to select the optimal (lowest) rate from the perspective of electricity users. With the advancement of technology related to AMI metering based on electricity rate, it was possible to derive a pattern through detailed analysis of electricity usage by time period, but it is necessary to expand the research to general houses including apartments nationwide in the future. If research and demonstration taking into account the impact of climate are conducted, it is expected that the development of an optimal rates that can satisfy both utilities and customers will be made.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.