Short-term PV generation forecasting based on a multi-object AVC strategy assessment system based on modified radar chart method

摘要

The AVC system is a critical automatic control system which makes contribution to securing the voltage and reactive power operation of the power network. During the closed-loop detection of the AVC system which is to be applied in sub-transmission network, much problems have been found about AVC strategies, including their failing in optimizing voltage level, power factor of 220kV transformer and active power loss, while in some occasion the switching constraints is not satisfied. The existed AVC assessment index is usually designed based on power system operation assessment index, which is not so suitable for AVC assessment. Aiming at this problem, this essay present a three-layer AVC strategy assessment index. The first two layer of the index gives assessment on the satisfaction and the effect of AVC strategy, while the third layer makes analysis about the regulation potential of the network. The fuzzy membership function is used in the designing of the first layer of index to depict the relationship between index value and satisfaction degree. Finally, a modified radar chart method is employed to illustrate the assessment result, while two evaluation functions are presented to compare different strategies. The example shows that this method is credible.