デマンドレスポンスを考慮した需要予測へのJITモデリングの適用

概要

論文の詳細を見る
This paper presents a demand response forecasting model by Just-in-Time modeling. It is planned to introduce the renewable energy on a large scale into the power system in order to achieve the low carbon society. However, the renewable energy has a negative effect on the power system. In recent years, demand response is expected to control the load as the reserve power. Thus, it is necessary to forecast load with demand response for unit commitment. The proposed JIT modeling uses the decision tree to analyze demand factors and the load correcting method of similar day. The former has functions to extract if-then rules and evaluate the variable importance. The latter corrects the load of similar day with demand rate and weather condition. The effectiveness of the proposed method is demonstrated using actual data of demand response about the adjustable Critical Peak Pricing field test at Kitakyushu Smart Community Project. The proposed method with the decision tree can extract similar day more accurately than a conventional method based on meteorological condition and day type. Moreover, the proposed load correcting method can correct the extracted load accurately.