基于专家系统的焊接机器人故障诊断

为了找到一个针对焊接机器人故障诊断较为可行的方法,避免诊断的盲目和失误,提高诊断的效率和准确度,在经典可靠性分析方法——故障树分析法的基础上,以"焊接机器人焊枪姿态和悬浮高度设置故障"为例建立焊接机器人故障树模型,同时对所建故障树进行定性和定量分析,得出最小割集的概率重要度排序及每个底事件的概率重要度。最后在故障树中体现的树状分层逻辑关系的基础上建立专家诊断系统,设计并实现基于产生式推理框架和故障知识的搜索流程。

Fault Diagnosis of Welding Robot Based on Expert System

A fault tree model of a welding robot applied to " welding robot welding torch gesture and failure altitude setting fail-ure" which based on the classic reliability analysis method-fault tree analysis method is a more feasible method for the fault diagno-sis of welding robots, which avoids blindness and mistakes in diagnosis, and improves the efficiency and accuracy of diagnosis. At the same time, qualitative and quantitative analysis were performed on the fault tree built, and the order of probability importance of mini-mum cut sets and probability importance of each bottom event were obtained. Finally, an expert diagnosis system was established on the basis of the tree-level hierarchical logical relationship embodied in the fault tree. A search process based on a production reasoning framework and fault knowledge was designed and implemented.