METHOD OF EVALUATING RELIABILITY ON ESTIMATION OF DAY MAXIMUM DEMAND POWER

摘要

PROBLEM TO BE SOLVED: To perform highly accurate estimation and evaluate the cause of an error, too, in the case that the actual value is clarified and that an error is occurring. SOLUTION: An estimation model is built by a recursive model, a neutral network, etc., using the weather data for a certain period before the target day of estimation, the day maximum demand power, and the distinction data such as Saturday, Sunday, etc., (step1). Next, the day maximum demand power on the target day of estimation is estimated by this estimation model (step2). Furthermore, the past actual result data regarding the day maximum demand power are corrected into ones equivalent to the target year of estimation (step3), and this actual result data and the estimation data regarding the day maximum demand power obtained by the estimation model are superposed on each other and are indicated in a power correlation figure (step4). Hereby, the estimation data, which are plotted, being separated from the past actual result data on the power correlation figure, are judged to be ones low in reliability.