点击次数：

论文类型：会议论文

发表时间：2008-04-06

收录刊物：EI、CPCI-S、Scopus

页面范围：1900-1905

关键字：automatic generation control; frequency bias coefficient; multi-target optimum technology; very short-term load prediction

摘要：In order to implement object-controllability and pre-control of AGC, a method of estimating frequency bias coefficient by using multi-objective optimization technology and very short-term load prediction is presented. For practical implementation, the method is designed based on discrete-time system. It introduces control factors, which artificially control the importance of each performance requirement according to area characteristic or operator's intention. In the meantime, using the results of very short-term load prediction, an optimal algorithmic program is run before every forecast period, thus each area could obtain object-controllable B coefficient forward, and the AGC units could regulate output under the guidance of load prediction. The method is examined by digital simulation with a three-area system model. The results showed that the method is accurate and effectual for estimating B coefficient, and the performance of interconnected power systems is improved by using the object-controllable and predictive frequency bias coefficient.