摘要
基于相关的人工智能理论,对墙面喷浆抹平机器人的故障原因进行了分析,并综合专家系统和神经网络特点,提出了基于2种方法相互融合的墙面喷浆抹平机器人故障诊断系统,实现了故障的快速、准确诊断及处理.设计了墙面喷浆抹平机器人故障诊断系统总体结构,采用故障树分析法,以墙面灰浆不均匀为例,分析了墙面喷浆抹平机器人出现此故障的可能原因,并创建了常见故障代码表.以Visual Basic为基础语言,开发了状态检测和故障诊断的上位机界面.该研究为喷浆抹平机器人的不断完善和其智能故障诊断系统的开发提供了理论依据.
In this study, the fault formation of wall-guniting robots is analyzed on the basis of artificial intelligence theories. By synthesizing expert systems and the neural networks, a wall-guniting diagnosis system is first proposed for fast and accurate fault diagnosis and solving. Then, the architecture of fault diagnosis sys- tem is designed. In the ease of uneven mortars, the possible fault causes of wall-guniting robots are detected using the fault tree analysis. Next, a common fault coding table is generated. Finally, the host computer interfaces are developed via Visual BasicTM for condition monitoring and fault diagnosis. To this end, this approach establishes a theoretical basis to progressively improve the intelligent system.
出处
《中国工程机械学报》
2008年第3期354-358,共5页
Chinese Journal of Construction Machinery
基金
山东省教育厅科技发展计划基金资助项目(J0D10)
关键词
喷浆抹平机器人
BP神经网络
故障诊断系统
专家系统
guniting-floating robot
backpropagation neural network
fault diagnosis system
experts system