电力系统可靠性原始参数的优化GM(1,1)预测

考虑到可靠性原始参数的缺乏对电力系统可靠性评估结果的真实性和有效性影响很大，用优化的GM(1,1)模型预测可靠性原始参数，开发小样本系统。优化的GM(1,1)模型在以最小二乘法优化初值的基础上，分别求取不同时间段的原始参数序列的拟合数列，再以各拟合数列与原始数列之间的模糊贴近度为权重系数对预测值进行优化加权组合。此模型既能体现数据的最新变化态势，又能体现总体发展趋势，充分挖掘原始参数包含的信息量，克服传统GM(1,1)模型预测可靠性参数随预测点推移预测精度下降较快的缺点，尤其适用于新投入元件可靠性原始参数的多点预测。

Prediction of Original Reliability Parameters of Power System by Optimized Grey Model

An optimized GM(1,1) model are used to predict the original reliability parameter of power system in order to exploiting original reliability parameter small sample systems when considering the lack of original reliability parameter has great influence to the results of power system reliability assessment. The proposed model is optimized by calculating the fitting series of original parameters of different time and integrating prediction values by fuzzy nearness between original series and fitting series as weight coefficient on the basis of using least squares to optimize the initial value. This model can reflect not only the latest changes in posture, but also the overall trend of data, therefor can fully exploit the information contained in the original parameters to overcome the defect of the prediction precision decline with prediction point. This model is applicable to the multi-point pritiction of original parameters of new electrical components.