基于改进GM(1,1)模型的装备故障预测

Equipment Fault Prediction Based on Improved GM(1,1) Model

针对装备故障数据量小、传统灰色预测模型GM(1,1)的精度受背景值影响大的问题,提出了一种基于背景值优化的改进GM(1,1)装备故障预测方法。该方法基于背景值的几何意义,从背景值与发展系数之间数量关系的角度出发,通过最小二乘法对二次参数进行估计,还原得到原始参数估计值,并结合新陈代谢对模型进行改进。导弹装备的故障预测实例表明,改进GM(1,1)模型较传统模型有更高的预测精度。

In view of the small amount of equipment fault prediction data and the great influence of background value on the accuracy of traditional grey prediction model GM(1,1),an improved GM(1,1)equipment fault prediction method based on background value optimization is proposed.Based on the geometric meaning of background value,from the point of view of the quantitative relationship between the background value and the coefficient of development,the least squares method is used to estimate the two original parameters,and combined with metabolism to improve the model.The failure prediction example of missile equipment shows that the improved GM(1,1)model has higher prediction accuracy than the traditional model.