Short-term Operation Planning in Electric Power Systems

Abstract

A practical application of operational research methods is shown here for the short-term operation planning in an electric power system. Two interactive application programs are presented: a short-term load forecasting program and a unit commitment optimisation program. For the load prediction within the load forecasting program a four-component weather dependent load model is used, whereby for the weather-dependent load component an econometric model, and for the time-dependent load component a seasonal multiplicative ARIMA-model is applied. The unit commitment is optimized using dynamic programming. With this method the start-up and shut-down costs can easily be taken into consideration. Special features are necessary to allow the optimal scheduling of a large number of generators to minimize storage requirements and execution time of the program. For every cost-optimal unit commitment determined the reliability of the energy supply is then calculated.The success of the application software depends to a large extent on the man-machine interface. Special attention has to be paid to the man-machine dialogue. The dialogue has to be structured in such a way that the user has the facility to control the programs as easily as possible.All input data has to be checked and in case of a false input an error message has to be displayed and a new input requested. The data-base has to be protected against wrong actions.