Study on Carbon Emission Pattern Derived from Electricity Data for Rural Area—A Case Study of Yushan Island

The purpose of this research is to build a workflow discovering building time-series energy usage valuable information decreasing carbon emission efficiently and establish a program realizing intelligent decision. Research done so far mainly concentrates on total power utilization instead of time-series data. In this paper, a k-shape time-series algorithm is harnessed for clustering, while some regression algorithms are utilized building classifier model. The studied object regards Chinese Yushan island as the database sample. The result shows that all village carbon emission patterns derived by electricity consist of V, M and line. Each mode corresponds different inhabitants. Thereinto, fisherman illustrates classical islet energy consumption feature and several carbon lesson methods could be used in their vacant houses during fishing period. At last, the conclusion indicate that this studied workflow could efficiently identify the energy usage characteristic and provide targeted carbon reduction strategies for users quickly.