基于一阶相关性模型的工业机器人故障诊断方法

Fault diagnosis method of industrial robot based on first-order relevant model

摘要

为提高工业机器人故障诊断实用性,减少对于产品故障机理和专家经验知识的依赖,提出一种利用一阶相关性模型理论开展工业机器人故障诊断研究的方法。该方法通过分析工业机器人内在联系和任务特点以建立典型任务模型,在此基础上采用一阶相关性列矢量法推导建立相关性数学模型,并通过数学模型优化和分割完成测试点选择、诊断树建立和故障字典编制等,为建立和优化工业机器人故障诊断策略提供基础。同时,选取焊接机器人作为典型装备,开展了机器人的相关性模型建立、模型优化和分割、建立诊断树、编制故障字典以及模拟故障试验等应用研究,故障注入模拟试验结果表明,该方法具有直观、高效的测试性建模与故障诊断能力,为提高工业机器人产品的故障诊断效率提供了新思路。

Abstract

For reducing the reliance on product fault mechanisms and expert experience knowledge, to improve the practicality of industrial robot fault diagnosis, a fault diagnosis method of industrial robots based on first-order relevant modeling theory is proposed. In this method, the typical task model is established by analyzing the internal relationship and task characteristics of industrial robots, and the first-order relevant column vector method is used to solve the relevant mathematical model. Then, through the optimization and segmentation of the mathematical model, test points selection, diagnosis tree and fault dictionary are completed, which could provide a basis for developing the product fault diagnosis strategy. Meanwhile, the welding robot is selected as a typical product to carry out the application research of relevant modeling, model optimizing and dividing, testing point selecting, diagnosis tree establishing, fault dictionary compiling and simulated fault diagnosis testing. The fault injection simulation test results show that, this method has intuitive and efficient testability modeling and diagnosis ability, and provide a new idea to improve the fault diagnosis efficiency of industrial robots.

关键词

相关性模型 / 工业机器人 / 故障诊断 / 模型分割 / 应用研究 / 故障注入模拟试验

Key words

relevant model / industrial robots / fault diagnosis / model segmentation / applied research / test verification