基于最大熵的电压暂降幅值概率分布评估

Evaluation on Voltage Sag Amplitude Probability Distribution Based on Maximum Entropy

暂降幅值概率分布评估是电压暂降评估中的重要环节。针对电网中监测装置配备滞后,监测样本不足等问题,运用最大信息熵原理,提出一种适用于小样本的电压暂降随机评估方法。根据电压暂降样本数据,构建基于各阶中心矩约束下暂降幅值分布的最大信息熵模型,约束条件中充分考虑了数据典型特征量的影响,并进一步研究模型参数的设定,对模型约束条件中中心矩最高阶数进行详细分析。利用最大信息熵原理求解暂降幅值的概率分布,该方法无需多次模拟,计算时间短,具有客观性,能够避免主观误差。最后以IEEE39节点系统为例进行电压暂降评估,与基于蒙特卡洛法的评估结果对比,证明本文方法的有效性。

In allusion to problems of lag equipment of monitoring devices in power grids and insufficient monitoring samples, this paper uses principle of maximum information entropy to present a kind of random evaluation method for voltage sag （VS） suitable for small samples. According to VS sample data, it constructs a maximum information entropy model based on VS amplitude distribution under the constraint of central moment of each order. It fully considers influence of typical char- acteristic quantity of data in constraint conditions and further studies setting of model parameters to analyze the highest or- der of central moment. Based on maximum information entropy principle, it can solve probability distribution of VS ampli- tude. There is no need for repeated simulation and the calculation costs little time. This method has objectivity which means capability of avoiding subjective error. It takes IEEE 39 node system for evaluation on VS and then compares the result with that of Monte-Carlo method, which proves validity of this method.