配电网快速高容错性故障定位的线性整数规划方法

Integer Linear Programming Based Fault Section Diagnosis Method With High Fault-tolerance and Fast Performance for Distribution Network

研究快速高容错性馈线故障定位方法对于提高配电网运行可靠性和自愈性有重要作用。该文基于代数运算关系,提出配电网故障定位逼近关系描述的非逻辑建模方法,并构建具有高容错性能的故障定位绝对值新模型。依据馈线状态编码特征,通过对绝对值模型的等效变换,建立了仅含0-1离散变量的线性整数规划故障定位新模型。基于代数关系的配电网故障定位新模型优点在于:能够克服已有间接故障定位方法对群体智能算法的依赖,可利用常规优化算法进行决策。3个仿真算例表明,采用新模型和线性整数规划方法进行配电网故障定位时不仅具有高的容错性能,能够实现多信息畸变下馈线故障区段的准确辨识,且具有实现方便、故障辨识效率高、数值稳定性好的优越性。

Fast and high-tolerance feeder fault location method plays an important role in improving reliability and self-healing of distribution network. The novel fault location absolute value model for distribution network with high tolerance was first presented on algebra descriptions based approximation relationship approach. Then,the integer linear programming fault location model with 0-1 discrete variables was constructed by through equivalent transformation of absolute value model for distribution network based on feeder state encoding. Notable feature of the novel model was that linear integer programming can be directly applied to make decisions for identification of fault location,which can avoid over-reliance on swarm intelligence algorithm. Three simulation examples were executed and results show that the novel fault location method for distribution network not only has high fault tolerance performance,but also can realize the accurate identification of feeder fault section under multi information distortion, and has the advantages of high efficiency and good numerical stability.