426 ADVANCED DATA ANALYSIS APPLIED TO ENERGY EFFICIENCY IN DISTRICT HEATING(International Session)

概要

論文の詳細を見る
Energy usage in District Heating Networks raises concerns such as mid season overheating, transients heat losses detection, optimal regulation strategies, and efficiency drifts of heat exchangers under actual running conditions. This paper aims at improving the operating conditions in order to avoid energy waste with the implementation of statistical data analysis; herein various algorithms are to be benchmarked according to their ability to bring out key parameters, such as abnormal fouling factor or heating control law, accounting for abnormal running conditions. In order to challenge the developed algorithms, we first derive a fully controlled reference dataset derived from an engineering modeling, along with artificial abnormal conditions specifically designed for the desired proof. Those data accounts for one year operation of a district heating that would supply fifty dwellings. Two separate approaches are used to identify the key parameters: a time domain model identification problem for the single building fault detection question, and a multi-site energy performance classification problem. We deal here with one mathematical model based on Vapnik's theory.
2008-07-09