远程协同诊断的现场信息预处理研究

On Local Information Preprocessing for Remote Collaborative Diagnosis

面对远程协同诊断存在的信息传输延迟、网络阻塞等问题,提出了现场诊断信息预处理概念。在其框架介绍基础上,重点研究了现场信息预处理策略控制和深层信息预处理;运用统计分析和SVM回归预测的集成工况预测方法作为预处理策略控制的依据,而基于粗集的时序诊断信息提取方法则是深层信息预处理的核心。最后给出了现场信息预处理在工程装备上的应用示例。

In order to solve the bottleneck problems in remote collaborative diagnosis, such as information transmission barrage and net delay, an iD-LIP （intelligent Diagnosis-Local Information Preprocessing）was proposed. We first presented an iD-LIP function framework. Then we studied the managing strategy of preprocessing and deep in- formation preprocessing. The managing strategy of preprocessing was realized by an integrated equipment performance prediction method which composed of statistics analysis and SVM （ support vector machine） regression prediction. An information reduction method based on rough set theory was adopted in deep information preprocessing. As an example, an application of iD-LIP to engineering equipments was given. Results show that the iD-LIP strategy is feasible and efficient.