论文部分内容阅读
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 ,采用常规诊断和基于多专家组合的精密诊断
In order to improve the ability of intelligent diagnosis system to solve complex problems, a prototype system of gear box monitoring and diagnosis was constructed based on distributed intelligence, object-oriented method and information fusion technology. The system is a multi-attribute, multi-tasking and multi-target uncertainty information processing system combining qualitative knowledge and quantitative knowledge. State monitoring task is based on multi-network collaborative reasoning method, using feature fusion and spatio-temporal decision fusion technology to improve inference efficiency. Fault diagnosis, for different diagnostic requirements, the use of conventional diagnosis and based on multi-expert combination of precision diagnosis