【摘要】 针对低压台区线损情况,提出一种数据挖掘技术的台区线损计算方法,根据样本的电气参数,提出台区K-Means聚类算法,该算法将台区线损值进行集中分类,解决了数据分散的问题,之后通过LM算法将BP神经网络优化,拟合出电气参数与样本线损率的关系,得到线损的变化规律,并根据拟合算法编程实现。结果表明,LM算法优化后得到的神经网络模型计算结果精度高,能够准确计算出台区的线损率,且更具合理性。更多还原

【Abstract】 According to the situation of line loss in low voltage station area, a calculation method of line loss in low voltage station area is proposed. According to the electrical parameters of the sample, the K-Means clustering algorithm in station area is proposed, which classifies the line loss value in station area centrally and solves the problem of data dispersion. Then the BP neural network is optimized by LM algorithm, and the relationship between electrical parameters and line loss rate of sample is obtained, and the change law of line loss is obtained. It is programmed according to the fitting algorithm. The results show that the neural network model optimized by LM algorithm has high accuracy, can accurately calculate the line loss rate of the issued area, and is more reasonable.更多还原