抄録
In an actual power system, there are various types of controllers such as AVR, PSS and GOV which influence the stability of power system remarkably. It is necessary to properly tune the parameters of such controllers to stabilize a power system; eigenvalue sensitivity is a useful tool for this purpose. This paper presents an efficient algorithm for evaluating eigenvalue sensitivities in large power systems. In the proposed method, the sparsity of the system state matrix is fully exploited so that the sensitivity can be obtained with far less computational effort compared with the existing methods. Furthermore, we propose a useful technique to automatically calculate a derivative coefficient matrix which is necessary to evaluate the sensitivities. By using this technique, the eigenvalue sensitivities with respect to any parameters of various controllers can be evaluated easily. The efficiency of the proposed method is illustrated by applying it to a very large power system which has 220 generating units and 3000 state variables.
