摘要
针对化工泵在高腐蚀、高负载工况下易发轴承、密封与电机故障的问题,提出基于多源数据驱动的故障识别与维修优化方法。通过采集温度、振动、电流与压力等关键参数,提取频域能量、波动特征与非稳态指标,结合XGBoost与LSTM模型实现故障类型识别与趋势预测。根据识别结果,制订轴承、密封、电机的专属维修策略,明确报警阈值、工艺流程与控制参数,提出标准化实施路径。该方法构建了识别一响应一执行闭环机制,显著提升了故障处理的主动性与精准性,为化工泵设备预测性维修体系的建立提供了技术支撑与工程实践依据。
A fault identification and maintenance optimization method based on multi-source data-driven approach is proposed to address the issue of bearing,seal,and motor failures in chemical pumps under high corrosion and high load conditions.By collecting key parameters such as temperature,vibration,current,and pressure,extracting frequency domain energy,fluctuation characteristics,and nonstationary indicators,and combining XGBoost and LSTM models to achieve fault type identification and trend prediction.Based on the recognition results,develop exclusive maintenance strategies for bearings,seals,and motors,clarify alarm thresholds,process flow,and control parameters,and propose standardized implementation paths.This method constructs a recognition response execution closed-loop mechanism,significantly improving the initiative and accuracy of fault handling,providing technical support and engineering practice basis for the establishment of a predictive maintenance system for chemical equipment.
出处
《今日自动化》
2026年第1期64-66,共3页
Automation Today
关键词
化工泵
故障识别
多源数据
预测性维修
智能建模
chemical pump
fault identification
multi-source data
predictive maintenance
intelligent modeling
