纺纱车间能源监控平台研究与实现  

Research and implementation of an energy monitoring platform for spinning workshops

作　　者：姜国强 袁逸萍[1] 晁永生[1] JIANG Guoqiang;YUAN Yiping;CHAO Yongsheng(School of Mechanical Engineering,Xinjiang University,Urumqi 830017,China)

机构地区：[1]新疆大学机械工程学院,乌鲁木齐830017

出　　处：《现代纺织技术》2025年第2期107-117,共11页Advanced Textile Technology

基　　金：新疆维吾尔自治区科技计划项目-重点研发专项(2022B01057-2)。

摘　　要：针对传统纺纱车间能耗信息不透明、能源管理粗放等问题,设计一种基于物联网和大数据技术的车间能源监控平台。基于智能仪表、LoRa无线传输技术等建立能耗数据物联监测网络;基于大数据技术实现海量能耗数据协同集成、分类存储和分布式计算;基于VMD-PSO-LSTM算法构建能耗数据预测模型,该模型相比于BP、LSTM、PSO-LSTM、VMD-LSTM等实现了更高精度的设备能耗多步预测;基于Java语言、SpringBoot框架、Vue框架等开发了前后端分离架构的能源监控系统。研究实现了车间能耗数据的及时采集、全面监测、准确预测和有效管理,对于传统纺纱企业低碳减排、数字化转型有一定应用价值。Against the backdrop of the accelerating global industrialization process and the increasingly scarce energy resources,the textile industry is undergoing a critical transformation phase.Driven by policy guidance and market competition,the textile industry is moving towards digitalization,intelligence and greening.However,many traditional spinning enterprises still rely on manual meter reading and analyzing and calculating energy consumption data.This method is not only inefficient but also prone to errors,making it difficult to accurately grasp workshop energy consumption and restricting the energy management capabilities and production efficiency of these enterprises.To solve this problem,this paper proposes an energy monitoring platform for spinning workshops based on Internet of Things(IoT)and big data technology.The platform establishes an energy data IoT monitoring network through smart meters equipped with LoRa wireless transmission technology,which realizes the comprehensive and timely collection of energy data.Big data technologies,including HDFS,HBase,Spark and other components in the Hadoop ecosystem,are used to provide a multi-tiered storage and analysis architecture for energy consumption data,which ensures fast integration,efficient computation and flexible querying of data.A multi-step energy consumption prediction model based on variational modal decomposition(VMD),particle swarm optimization(PSO),and long-short-term memory neural network(LSTM)is proposed,and the power consumption data and relevant operational data of a spinning enterprise in Xinjiang were selected for verification is validated.The experimental results show that the model outperforms traditional algorithms such as BP,LSTM,PSO-LSTM,and VMD-LSTM in single-step,three-step,and six-step energy consumption prediction.Specifically,its single-step average prediction error is reduced by 74.50%compared to LSTM and 39.07%compared to VMD-LSTM,and the R 2 values for single-step,three-step,and six-step predictions are 0.9820,0.9642,and 0.9151,re

关 键 词：能源监控 大数据技术 纺纱车间 LoRa VMD LSTM 

分 类 号：TS111.8[轻工技术与工程—纺织材料与纺织品设计]