基于灰色神经网络的装备计量预测研究与实现

Research and Implementation of Equipment Metrological Forecasting Based on Grey Neural Network

摘要:

基于装备计量数据历史样本数据较少的特点,将适合小样本的灰色理论GM(1,1)模型应用于基于计量数据的装备状态预测,同时为提高GM(1,1)模型精度,提出了基于RBF神经网络优化GM(1,1)传统模型的灰色神经网络模型。装备计量数据实例应用分析表明,上述模型均可获得该装备计量数据的合理预测值,且相对于GM(1,1)传统模型,GM(1,1)优化模型具有更优的模型精度和预测效果,基于MATLAB开发的装备计量预测软件,实现了GM(1,1)传统及优化模型下装备计量状态预测及比较的可视化操作,为装备计量保障提供了可参考的技术方案。

Abstract:

In order to realize the forecasting of equipment technical status based on metrological data which with less historical sample data, the GM (1,1) model of grey theory which suitable for less sample data was applied. And a grey neural network model which based on GM (1,1) traditional model optimized by RBF neural network was proposed, which in order to improve the accuracy of GM (1,1) model. The application analysis of equipment metrological data shows that the models all can obtain the reasonable forecasting value, and compared with the GM (1,1) traditional model, the GM (1,1) optimization model has better model accuracy and forecasting effect. The software of equipment metrological forecasting which developed by MATLAB, which realized the visualization operation of equipment metrological forecasting and comparison which in GM (1,1) traditional and optimization model, provides a reference technical scheme for equipment metrological support.