Abstract
Household cogeneration systems using proton exchange membrane fuel cells (hereinafter abbreviated as PEMFC systems) have attracted attention for their high energy efficiency. To bring out a PEMFC system's performance, it is crucial to use effectively both of electric power and hot water produced from a PEMFC system. Therefore, the optimal scheduling problem of a PEMFC system is very important. However there are difficulties due to the uncertainty of the household energy demand. A difficulty, to which little attention has been paid so far, lies in the difference between the scheduled electric power output and the actual electric power output of a PEMFC system caused by rapid fluctuations of the electric power demand. In this paper, we propose a new method to cope with this problem. In our method, we treat the energy demand as a random variable and calculate its probability density at each time step. Then a constraint that takes into acount the energy demand fluctuations is defined through the probalility density and is cooperated in the original optimal scheduling problem of the PEMFC system. Using our method, we can deal with rapid energy demand fluctuations under the setting of the scheduling problem with long time intervals.
