基于分布式智能的齿轮箱监测和诊断原型系统设计

A Prototype of the Intelligent System for Gearbox’s Monitoring and Diagnosis using Distributed Intelligence

为提高智能诊断系统解决复杂问题的能力 ,基于分布式智能思想、面向对象的方法和信息融合技术 ,构造齿轮箱装置监测和诊断的原型系统。该系统是一个定性知识与定量知识相结合的多属性、多任务、多目标的不确定性信息处理系统。状态监测任务由基于多网络协同推理的方法完成 ,采用特征融合和时空判决融合技术提高推理效率。故障诊断时 ,针对不同诊断要求 。

In order to improve the complicated problem solving ability, a prototype of the intelligent system for gearbox's condition monitoring and fault diagnosis, based on the combinations of experts, is constructed in this paper using the object-oriented methodology , the distributed intelligence, and the data fusion.The system is a multiple task, multiple attribute, multiple object information processing system with uncertainty, which combines qualitative and quantitative knowledge.The technique which combine the feature compression, time-space decision fusion and multiple network cooperative inference, is used to improve the inference efficiency in the monitoring.The hierarchical diagnostic strategies, the convention diagnosis and the combinations of experts, are employed to meet different diagnostic requirements.