面向矿用机电设备数字孪生模型的故障特征提取与识别技术  

Fault feature extraction and identification technology for digital twin model of mine electromechanical equipment

作　　者：李丁卯 罗珍平 LI Dingmao;LUO Zhenping(School of Electrical Engineering and Automation,Henan Polytechnic University,Jiaozuo 454003,China;Shanxi Luneng Hequ Electric Coal Development Co.,Ltd.,Xinzhou 036500,China)

机构地区：[1]河南理工大学电气工程与自动化学院,河南焦作454003 [2]山西鲁能河曲电煤开发有限责任公司,山西忻州036500

出　　处：《现代电子技术》2025年第8期173-178,共6页Modern Electronics Technique

基　　金：国源电力科信2023年度科技项目及技术标准项目(GSKJ-23-65)。

摘　　要：为满足矿用机电设备的智能化故障诊断需求,基于数字孪生模型提出了一种故障特征提取与识别技术方案。该方案主要包括机电设备的数字孪生建模和故障特征提取与识别两方面。通过卷积神经网络(CNN)和长短期记忆(LSTM)网络的混合模型完成数字孪生的建模;使用数据可视化技术和Z-score标准化方法对数据进行处理和筛选,利用小波变换方法进行故障特征提取,并设计一种基于CNN的故障识别算法。相比于传统方法,所提出的故障识别算法能够有效提高故障识别的准确率和实时性。实验测试结果表明:所构建的数字孪生模型能够准确地模拟和表征设备运行情况,验证了所提方法的正确性和有效性;而且故障识别准确率高于同类技术模型,在提高故障诊断效率方面的工程应用效果良好。In allusion to the demand for intelligent fault diagnosis of mine electromechanical equipment,a fault feature extraction and recognition technology schemes based on digital twin models is proposed.This scheme mainly includes digital twin modeling of electromechanical equipment and fault feature extraction and recognition.The modeling of digital twins was completed by means of the hybrid model of convolutional neural network(CNN)and long-short term memory(LSTM)network.The data visualization techniques and Z-score standardization methods are used to process and filter the data,the wavelet transform method is used to extract fault features,and a CNN based fault recognition algorithm is designed.In comparison with traditional methods,the proposed algorithm can effectively improve the accuracy and real-time performance of fault recognition.The experimental testing results show that the constructed digital twin model can accurately simulate and characterize the operation of the equipment,which verifies the correctness and effectiveness of the proposed method.The average accuracy of fault identification is higher than that of similar technical models,and the engineering application effect in improving the efficiency of fault diagnosis is good.

关 键 词：煤矿机电设备 数字孪生模型 故障特征提取 故障识别算法 卷积神经网络 长短期记忆网络 诊断精确度 

分 类 号：TN929-34[电子电信—通信与信息系统] TP277[电子电信—信息与通信工程]