基于灰色时序组合模型的油液光谱分析预测

本资料包含pdf文件1个，下载需要1积分

以某型发动机油液光谱分析数据为基础,分别利用GM模型、AR模型和灰色时序组合模型,对实测数据进行动态建模,并对发动机润滑油中Fe元素的变化进行预测,结果对比表明,灰色模型能够模拟数据的宏观变化趋势,时序模型能够描述数据的细微随机变化,灰色时序组合模型综合两种模型的优点,达到更高的预测精度。 Based on the oil spectrum analysis data of a certain type of engine,the GM model,AR model and grey AR combination model are used to establish the dynamic model of the measured data,the content of iron in the engine＇s lubricating oil is predicted.The comparison of results shows that the gray model is able to simulate the macro trend of data,the auto-regressive model can describe the subtle random changes of data,the grey AR combination model achieves higher prediction accuracy by integrating the advantages of both models.