基于Hybrid CBR的混合动力推土机故障诊断专家系统研究

发布时间：2018-05-23 09:36

本文选题：故障诊断 + 推土机　；参考：《山东大学》2017年硕士论文

【摘要】：混合动力推土机是混合动力技术在工程机械领域的创新应用,混合动力推土机由于系统多、结构复杂、运行工况恶劣,其故障的产生原因较为复杂,因此使用传统建模的方法进行故障诊断较为困难。近年来智能理论的应用发展十分迅速,以历史数据为基础的故障诊断方法也在故障诊断领域中发挥了十分重要的作用,本文就是基于此趋势进行混合动力推土机的故障诊断研究。基于规则的推理技术(RBR,Rule-based Reasoning)在故障诊断方面有着重要应用,它对前人经验进行系统化总结并模拟专家解决疑问的思维方式对当前问题进行求解。基于实例的推理技术(CBR,Case-based Reasoning)同样在故障诊断方面应用较广,它基于通过历史、监控和实验采集到的实际案例对当前问题进行求解。本文通过远程监控系统收集到车辆各系统数据后形成故障诊断数据库并以此为基础将RBR和CBR进行集成,最终建立基于Hybrid CBR的混合动力推土机故障诊断专家系统。此系统可以迅速找到故障发生的位置、发生的原因以及处理的方法,帮助操作人员对问题及时处理,对提高设备可靠性、缩短维修周期、降低维修成本有着十分重要的研究意义和应用价值。基于规则的推理技术包括规则总结、规则推理、规则管理三大部分,其中规则总结是RBR中最为关键的技术,也是RBR技术研究的关键点。而规则总结的目标就是将领域内的知识进行总结形成可以用于故障诊断的规则库。本文使用大数据关联规则分析中的Apriori算法对通过远程监控系统采集到的数据进行挖掘,寻找出各特征项与故障之间的联系,形成基于规则推理的故障诊断规则库。并对利用此规则库进行RBR诊断进行了详细介绍。基于实例的推理技术中实例检索是最为关键的技术,其检索算法的使用直接关系到CBR检索模块和整个系统的效率。本文使用K最临近(K-NN)算法进行实例检索,在此基础上使用AHP层次分析法对特征项进行加权,使其能够突出重点,更精确的进行检索。系统的CBR模块将检索分为两步,首先使用数据库的SQL语言进行初步检索,再使用K-NN算法进行精确检索,从而在保证检索精确度的同时提高了系统效率。本文将RBR和CBR两种方式有机的串联起来,构建了基于Hybrid CBR的混合动力推土机故障诊断专家系统。本系统使用Visual Studio 2010作为软件开发环境,使用SQL Server 2008作为数据库管理工具,并在系统不同模块的开发中用到C#、R以及SQL语言作为编程语言。
[Abstract]:Hybrid power bulldozer is an innovative application of hybrid power technology in the field of engineering machinery. Due to many systems, complex structure and bad operating conditions, hybrid bulldozer is complicated, so it is difficult to use traditional modeling method to diagnose fault. In recent years, the application of intelligent theory is very rapid. The fault diagnosis method based on historical data also plays a very important role in the field of fault diagnosis. This paper is based on this trend to study the fault diagnosis of hybrid bulldozer. RBR (Rule-based Reasoning) has an important application in fault diagnosis, and it has been applied to the previous experience. CBR (Case-based Reasoning) is also widely used in fault diagnosis. It is based on the actual cases collected through history, monitoring and experiment. This paper is collected through remote monitoring system. After collecting the data of the vehicle system, the fault diagnosis database is formed and the RBR and CBR are integrated on this basis. Finally, the fault diagnosis expert system of hybrid power bulldozer based on Hybrid CBR is established. The system can quickly find the location of the fault, the cause and the method of the reason, and help the operator to get the problem in time. It has very important research significance and application value for improving equipment reliability, shortening maintenance cycle and reducing maintenance cost. Rule based reasoning technology includes rule summary, rule reasoning, rule management three parts. Rule summary is the most critical technology in RBR, and it is also the key point of RBR technology research. The goal is to sum up the knowledge in the field to form a rule base which can be used for fault diagnosis. In this paper, the Apriori algorithm in the analysis of large data association rules is used to excavate the data collected through the remote monitoring system and find out the connection between the features and the fault, and form a rule based fault diagnosis rule. RBR diagnosis is introduced in detail. Case retrieval is the most critical technology in case based reasoning. The use of the retrieval algorithm is directly related to the efficiency of the CBR retrieval module and the whole system. In this paper, the K nearest (K-NN) algorithm is used for instance retrieval, and on this basis, the AHP hierarchy is used. The CBR module of the system is divided into two steps. First, the system uses the SQL language of the database to carry out the preliminary retrieval, and then uses the K-NN algorithm for accurate retrieval, thus improving the system efficiency while guaranteeing the accuracy of the retrieval. This paper puts the RBR and CBR two kinds of parties in this paper. The Hybrid CBR based fault diagnosis expert system of hybrid power bulldozer based on Hybrid CBR is set up in series. The system uses Visual Studio as the software development environment, uses SQL Server as the database management tool, and uses C, R and SQL language as the programming language in the development of different modules of the system.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU623.5

【相似文献】