STOCHASTIC OPTIMAL ENERGY DISPATCH

This thesis presents a method to incorporate the effects of uncertain parameters into optimal energy dispatch of electric power systems. An efficient algorithm is presented, which is a direct extension of existing methods of analysis. The method employs an initial operating point by a second order optimization procedure. Effects of uncertainty on the optimum solution are modeled by a first order linearization. The statistics of the changes in the control vector are calculated and utilized in a Gram-Charlier series type A representation of the p.d.f. of the changes of the control vector. Various properties of the Gram-Charlier series type A are studied and a transformation is developed which improves the accuracy of the truncated Gram-Charlier series. Hence, the GCA series capabilities as a model for general p.d.f.\u27s is enhanced. Desired probabilities of events are calculated by direct application of the truncated GCA. An example demonstrates the suitability of this method for application to actual power systems