Reactive power control by least squares minimization

An effective method is presented for performing reactive power control in a bulk power supply system by optimally rescheduling generator terminal voltage magnitudes and/or transformer tap settings. The optimization procedure is the familar least squares minimization technique and is suitable for on-line application in energy control center computers. Sensitivity factors provide the basis for the control function and are developed from a linearized reactive power model. Important features of this method include a model approximating the reactive power flow through a load-tap-changing (LTC) transformer and a Norton's equivalent circuit model approximating the reactive power output of a regulating generator. As a result of the linear models developed, optimally ordered triangular factorization is fully exploited to enhance computational speed and efficiency. Test results are presented to demonstrate the effectiveness of the method. A linearized active power model is briefly presented to illustrate its dual relationship to the linearized reactive power model. This is an important relationship, enabling both active and reactive power control by application of the same general model.