Research in this area has led to the development of a new and efficient method for computing the sensitivities of power flows, bus voltages, etc., with respect to transmission network parameters, for example, line admittances or tap settings. Termed the adjoint method, it computes the partial derivatives of the power system responses from the results of at most two load-flow analyses (one of the original system and the other of the topologically equivalent adjoint system) regardless of the number of variable parameters and without pertubing them. The procedure is an exact one and the computational effort is independent of the number of parameters.
This method is especially suited to system optimization and design, where sensitivities are needed to predict parameter changes. We have developed a number of computer codes which use the adjoint method in transmission expansion planning and in the solution of the optimal load flow problem. In addition, we are currently using the adjoint method to rank the severity of network contingency outages for the on-line security assessment problem.
The following papers illustrate our work in power system sensitivities:
1. "Efficient Computation of Optimal Load Flow Sensitivities", R. Fischl, R.G. Wasley, IEEE Canadian Comm. Power Conference, IEEE publication, CH 1373-0, pp. 401-4, 1978
2. "Efficient Method for Computing Electrical Power Transmission Network Sensitivities", R. Fischl, W.R. Puntel, Paper No. C72-167-0, IEEE Winter Power Meeting, New York, N.Y., January 1972.
