灰色模型在不确定性疲劳寿命预测中的研究

Prediction Method for Fatigue Life Based on Grey Model

考虑影响疲劳寿命因素的不确定性,应用灰色理论进行疲劳寿命预测。提出了基于线性GM(1,1)模型和非线性灰色Verhulst模型预测构件疲劳寿命的新方法。利用试验数据进行分析和计算,结果表明:基于传统Miner方法的疲劳寿命预测误差为61.4%;基于线性GM(1,1)模型的预测误差为24.1%;基于非线性灰色Verhulst模型的预测误差降低到17.5%。基于灰色模型预测的结果均偏向安全,说明灰色模型在不确定性疲劳寿命预测中具有较好的预测精度和潜在的工程实用价值。

Considering the uncertainty factors of the fatigue life,the grey system theory was used to predict the fatigue life.A new algorithm for predicting the fatigue life based on GM（1,1）model and non-linear grey model Verhulst was proposed.The results of analyzing and calculating experimental data validate that the prediction error which is based on the traditional Miner method is 61.4%,and the prediction error which is based on the linear GM（1,1）model is 24.1%,and the prediction error which is based on the non-linear grey Verhulst model is decreased to 17.5%.The prediction which is based on the grey model is safe.Therefore,the analysis of the examples indicates that using grey forecasting to predict the fatigue life has good forecasting accuracy and excellent applicability.