基于RBR+CBR双重推理的带式输送机驱动系统故障诊断研究

Fault diagnosis of belt conveyor drive system based on RBR+CBR dual reasoning

基于规则推理(RBR)和基于案例推理(CBR)的方法均适用于带式输送机故障诊断领域。针对二者在故障诊断中或侧重原因分析或侧重维修手段的问题,对基于RBR+CBR双重推理的推理机制进行了研究,并建立了专家系统,对带式输送机驱动系统进行了故障诊断。首先,在传统专家系统的结构基础上构建了故障诊断专家系统;然后,根据故障机理建立了驱动系统故障树,采用产生式表示法设计了故障规则库,根据故障案例与专家经验,采用矩阵表示法设计了故障案例库;基于规则推理(RBR)和基于案例推理(CBR)的方法,设计了专家系统推理机制;利用LabVIEW程序开发平台设计了专家系统人机交互界面,实现了传感器—知识库—客户端之间的串联通信;最后,为了验证上述专家系统的可行性,在某矿带式输送机运行现场进行了实验,并进行了结果分析。研究结果表明:该带式输送机专家系统的故障原因分析准确率超过95%,故障维修方案有效性超过94%;与RBR、CBR算法支持下的专家系统相比,该专家系统具有更优的收敛性、有效性和准确性,为带式输送机的故障智能诊断提供了一种有效的方法。

Both rule-based reasoning(RBR)and case-based reasoning(CBR)methods are applicable to the field of belt conveyor fault diagnosis.Aiming at the problem that both of them focus on cause analysis or maintenance means in fault diagnosis,the reasoning mechanism based on RBR+CBR dual reasoning was studied,and an expert system was established to diagnose the fault of belt conveyor drive system.Firstly,the expert system for fault diagnosis was constructed on the basis of the structure of traditional expert system.Then,the fault tree of the belt conveyor drive system was established according to the fault mechanism,the expert system fault rule base was designed by using the generative representation method,and according to the fault cases and expert experience,the expert system fault case base was designed by using the matrix representation method.The expert system reasoning mechanism was designed based on the rule-based reasoning(RBR)and case-based reasoning(CBR)methods.And the human-computer interaction interface of the expert system was designed by using the LabVIEW program development platform,which realized the series communication process between the sensor,knowledge base and client.Finally,in order to verify the feasibility of the above expert system,experiments were conducted on a belt conveyor operating site in a certain mine,and the results were analyzed.The results of the study show that,the fault cause analysis accuracy of the belt conveyor expert system exceeds 95%,and the effectiveness of the fault repair plan of the expert system exceeds 94%.In comparison with the expert system supported by RBR and CBR algorithms,this expert system has better convergence,effectiveness and accuracy,and it proposes an effective method for the intelligent fault diagnosis of the belt conveyor.