Expert system assisted identification of line faults on delta-delta distribution systems

Line faults in delta-delta distribution systems continue to pose a problem for the power industry. The authors focus on the first facet of the effort needed to implement an integrated algorithm for the detection and location of line faults in power distribution systems. This work entailed the analysis of the fault characteristics and the development of a preliminary fault identification technique. The Los Angeles Department of Water and Power, which uses Δ-Δ power transformers has identified four different single line fault types as being the most prevalent faults in their distribution systems. The analysis addresses these four single line faults. An integrated software package was developed which contained a simulation program written in FORTRAN and an expert system written in PROLOG. A laboratory experiment was performed to generate measured data to test the computer simulation program and to serve as an initial test pattern for verification of the expert system. The expert system architecture is dealt with. The results obtained using data acquired by the experiment are outlined