Statistical Considerations in Element Value Solutions

This paper presents a method of fault isolation in electrical networks by determination of circuit element values from a sufficient set of measurements at the available or partly available terminals. The method requires that the circuit configuration be known; the measurements taken are voltage and current responses to sinusoidal stimuli of specified frequencies. The method presented is an iteration one using maximum likelihood estimation techniques at each stage, the iteration accounting for the nonlinearity of equations used with respect to the element value unknowns. A specific computational procedure is evolved. An example of a sample two-terminal passive circuit with 3 elements is given in detail, complete with computer flow charts, and numerical results.