Application of artificial neural networks in non-invasive identification of electric energy receivers

Abstract

This article concern to systems that support the manage economical use of electricity. An advisory and monitoring system equipped with artificial intelligence for non-invasive on-line identification of electrical devices was proposed. Research has been done on the basic element of the system, which is the classifier module. As data describing the objects, the electrical quantities obtained from the Elspec BlackBox 4500 analyzer were used. It implements the various configurations of artificial neural networks (ANN), to provide identification efficiency of 68%–91%. The advantages of the various classifiers due to their architecture ware pointed out.