机械故障诊断系统设计与研究

发布时间：2018-04-13 04:15

本文选题：专家系统 + 故障树　；参考：《哈尔滨工程大学》2011年硕士论文

【摘要】：机械设备广泛应用于现代化大生产中,而工业大生产及科学技术的进步又使得机械设备更加复杂化、大型化,传统的机械诊断技术已经不能满足实际的诊断要求。人工智能技术的不断发展与完善,促进了机械诊断技术的改革,因此,研究与应用人工智能与机械故障诊断相结合的技术具有重要意义。本文从生产的实际情况出发,针对通常情况下生产设备工作环境的复杂性及大多数生产单位诊断专家不足的实际情况,分析了专家系统与机械故障诊断技术相结合的相关理论,并重点以生产中广泛使用的异步电动机为诊断对象,设计与研究了电动机故障诊断专家系统,并对其结构和实现的关键技术进行了分析实现。论文首先介绍了机械故障诊断专家系统的构成及其各部分的关键理论,以异步电动机为诊断对象,分析研究了其结构及其故障机理,并设计规则+框架的诊断专家系统。通过故障树分析获取知识,以建立知识库作为本文的研究重点。即将故障树知识和框架表示法相融合,并把故障树最小割集底事件融合到规则表示法的条件中,这种知识表示法不仅能很好的描述电动机的知识构成,而且可以使推理机的设计变得更简洁。故障诊断专家系统的另一个研究重点是推理机的实现技术。论文设计了以正向推理为推理方向,深度优先搜索为搜索策略的推理机。对于故障知识冲突的问题,采用故障原因重要度排序的冲突解决方法,这样可以高效的确定故障部位,推得诊断结果。采用ADO数据库访问技术,对数据库和知识库进行创建与管理,将故障树融合到知识库管理模块中,用户可以在了解电动机具体故障知识结构的前提下对知识库进行更新与管理。最后使用Visual c++ 2005程序开发工具,Access 2003数据库实现了对异步电动机故障诊断专家系统的设计,并经过电动机实验平台的测试验证,证明了其有效性与实用性。
[Abstract]:Mechanical equipment is widely used in modern large - scale production , while the industrial production and the progress of science and technology have made the mechanical equipment more complicated and large , and the traditional mechanical diagnosis technology has not met the actual diagnosis requirement . The continuous development and perfection of artificial intelligence technology has promoted the reform of the mechanical diagnosis technology . Therefore , the research and application of artificial intelligence and mechanical fault diagnosis is of great significance .

Based on the actual situation of production , this paper analyzes the related theory of expert system and mechanical fault diagnosis technology aiming at the complexity of production equipment working environment and the shortage of diagnosis experts in most production units , and analyzes the expert system of motor fault diagnosis based on the asynchronous motor widely used in production , and analyzes the key technologies of its structure and implementation .

This paper first introduces the constitution of the expert system of mechanical fault diagnosis and the key theory of its parts , and analyses the structure and its failure mechanism by using asynchronous motor as the diagnosis object , and designs the rule + framework diagnosis expert system . The fault tree knowledge and frame representation are fused , and the minimum cut - set bottom event of the fault tree is fused to the condition of the rule representation method , which not only can describe the knowledge of the motor well , but also makes the design of the inference engine more compact .

Another research focus of fault diagnosis expert system is the realization technology of inference engine . The paper designs a inference engine based on forward inference as the inference direction and the depth priority search as the search strategy . For the problem of fault knowledge conflict , a conflict resolution method is adopted for ranking the cause of the fault , so that the fault location can be determined efficiently , and the diagnosis result can be obtained .

By using ADO database access technology , the database and the knowledge base are created and managed , and the fault tree is merged into the knowledge base management module , and the user can update and manage the knowledge base on the premise of understanding the specific fault knowledge structure of the motor .

Finally , using the Visual c ++ 2005 program development tool , the Access 2003 database realizes the design of the fault diagnosis expert system of asynchronous motor and proves its validity and practicability through the test and verification of the motor experiment platform .

【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3

【参考文献】