Load frequency control considering very short-term load prediction and economic load dispatch using neural network and its application

Abstract

The paper presents a new technique for the load frequency control (LFC) of interconnected power system. The LFC system monitors, at a minimum, power system frequency, generator output, net interchange schedule and tie-line power flows. It compares the actual frequency and tie-line values to the desired or scheduled values and generates an error value called the area control error (ACE). ACE is the instantaneous estimate of load demand in the area. However, due to relatively fast area load demand fluctuations and the relatively slow area generation response rate, only the instantaneous estimate of load demand cannot provide us with a good dynamic response. For this purpose, a look-ahead load forecasting feature is needed for effective LFC. To improve performance for LFC strategies neural network use for very short-term load prediction. From this forecasted load every 5-min ahead find out load change or error of the system can be recorded. This total load change divided all the units by concept of the unit’s participation factor. Because of load is always changing so a fixed integral controller is not suitable for LFC, a fuzzy logic controller is used for this purpose.