Study And Application On Automatic Voltage Control For Ningxia Power Network

The global optimization genetic algorithm is adopted and improved through introducing the historical scheme database optimization method innovatively and using the latest research achievements of genetic algorithm in this paper.Reactive optimization is a dicrete nonlinear problem with lots of control variables. The coevolution method is based on decomposition and coordination. One complicate problem is decomposed into several interaction sub-problems which coevolve with the interaction of common system model. In this paper, classified coevaluation method is adopted. Compared with commom and anneal genetic algorithm, the coevalutation method can step out from the local optimum and search for a better solution. It also has good convergence characteristic. The coevaluation method is a parallel algorithm in nature, which can be parallel to improve the computation efficiency.Based on the practical system of Ningxia network, the proper reactive optimization model is proposed according to the characteristic of power netwok and system demand. The key control variables are decided by sensitivity method. The reactive optimization is achieved mainly by adjusting generator voltage,and auxiliarily by adjusting the key station tap and parallel compensation devices. Meanwhile, the benefits of new adding capacitors are analyzed by sensitivity method. The compensation spots are sorted by benefit analysis, which provides a useful reference for equipment of parallel compensation devices. The computation results of reactive optimization can provide a feasible scheme for the reactive plan optimization of Ningxia power grid.The reactive optimization of Ningxia power grid in three typical operation modes in 2007 and 2008 are calculated and the network power loss, bus voltage and transmission lines power flow are calculated and suggestions are provided for operation. Through the reactive planning in 2010, the reactive compensation spots and capacity are proposed. Finally, the control strategy of AVC for Ningxia power network is proposed through research on application of AVC at home and abroad.The calculation results show that: the algorithm and improved measure used in this paper can solve the reactive plan optimization well. The improved algorithm is much more stable and can converge quickly. The probability search problem is overcomed. The global optimal solution or quasi global optimal solution can be obtained in greater probability.