基于模糊聚类的配电网电能质量分级预警

Hierarchical early warning of power quality of the distribution network based on the fuzzy clustering method

现有国家标准中对于电能质量各指标只能判断其是否超标,且各指标限值固定不变,不能根据配电网自身的运行情况和特点进行动态预警。为解决上述问题,本文提出一种电能质量稳态指标分级预警方法。根据不同场景、时间对配电网电能质量的要求,通过将相同电压等级的预警对象进行聚类的方法动态设置其预警阈值,并根据预警阈值的不同,将电能质量分为若干等级。其次,提出基于BP神经网络算法的电能质量预测方法,提高了电能质量稳态指标的预测精度。最后,针对10 kV和35 kV两个电压等级的配电网进行了仿真分析,验证了所提出方法的有效性和实用性。分析证明,该预警方法的实施可提前预知配电网的电能质量情况,以便于提前采取措施,保证配电网安全运行。

In the existing national standards,the indicators of power quality can only be judged whether they exceed the standard.Because the limit value of each index is fixed,dynamic early warning cannot be made according to the running condition of the distribution network.In order to solve the above problems,this paper proposes a hierarchical early warning method for power quality steady indicators.Firstly,according to different requirements for the power quality of distribution network in different scenarios and time,the early warning threshold values are dynamically set by the method of clustering early warning objects having the same voltage level,and the power quality is divided into several grades according to the different early warning threshold values.Secondly,a power quality prediction method based on BP neural network algorithm is proposed to improve the prediction accuracy of power quality steady state indicators.Finally,the simulation analysis of the distribution network with two voltage levels of 10 kV and 35 kV is carried out to verify the effectiveness and practicability of the proposed method.The analysis proves that the implementation of the early warning method can predict the power quality of the distribution network in advance,so that measures are taken in advance to ensure the safe operation of the distribution network.