| 摘要 |
<p>The method involves providing a multi-objective optimization algorithm (A) for determining a value of a discrete variable (Vi) that is representative of activity of a set of electricity consumption sources (mi) or electrical appliances. An error function (F) entering total measured consumption (S) of the sources is minimized. A cross population of crossed individuals chosen in a pseudo-random manner by taking census probability (qj) into account is generated, and a current population is upgraded by adding individuals selected from the crossed individuals. An independent claim is also included for a device for identification of electricity consumption sources.</p> |
