基于数字孪生的智能工厂能耗优化模型研究

打开文本图片集

关键词：数字孪生；能耗；模型；数据采集；约束条件

中图分类号：TP278 文献标志码：A 文章编号：1003-5168(2025)24-0008-05

DOI:10.19968/j.cnki.hnkj.1003-5168.2025.24.001

Research on Energy Consumption Optimization Model of Smart Factory Based onDigital Twin

SUN Liping(Bureau of Industry and Information Technology of Luozhuang District,Linyi 276O17,China）

Abstract:[Purposes] Traditional factory energy consumption management mostly relies on manual statistics and empirical regulation, which has problems such as lagging data collection,one-sided energy consumption analysis,and lack of dynamic adaptability of optimization strategies.To address the above challenges,an energy consumption optimization model for intelligent factories based on digital twins has been designed.[Methods] In this paper,electricity，heat,fluid and environmental data are firstcollected through Schneider PM8OOO power monitors,Siemens SITRANS flowmeters and environmental sensor nodes. Then, multi-source heterogeneous data preprocessing is carried out by using wavelet denoising,Kalman filtering and time alignment algorithm.Subsequently，a feature engineering system was constructed,including time series statistical features,FFT frequency domain features and EMD decomposition features.Key features were screened through mRMR and L1 regularization,and a feature matrix was constructed by using PCA dimensionality reduction.Furthermore,a high-fidelity geometric solid model is established based on BIM,combined with the energy consumption mechanism models of the power,compressed air and cooling water systems,to achieve synchronous mapping of virtualphysical dataand form a digital twin environment that can interact and predictin real time.Finally,a multi-objective optimization function is designed to quantify industrial constraints,and combined with LSTM prediction and the NSGA-Il optimization algorithm，the rolling time-domain energy consumption optimization is achieved.[Findings] The experimental results show that the digital twin optimization model has reduced the average daily power consumption of the workshop by 28.65% ,the power consumption of the press and the transmission line by 27.42% and 42.09% respectively,and the energy consumption of the central air conditioning and other systems by 13.46% and 39.07% respectively. The production plan achievement rate has been raised to 99.71% , the product qualification rate to 99.23% the no-load running time of the press has been shortened to O.83 hours,and the number of fault shutdowns of the transmission line within3Odayshas been reducedto5times.Itcan beseenfromthis that this model demonstrates significant advantages in reducing energy consumption,enhancing production efficiency and improving the operational performance of equipment.[Conclusions] The energy consumption optimization method based on digital twins can provide eficient and visualized decision support for smart factories.Atthe same time,itoffers theoretical and practical references for industrial energy conservationand carbon reduction,intelligent operation and maintenance of equipment,and the construction of future digital factories.It can be further expanded to directions such as multi-factoryarea linkage optimization and new energy integration dispatching.

Keywords: digital twin; energy consumption; model; data collection; constraints

0 引言

在"双碳”战略目标深化推进与制造业智能化转型加速的双重背景下，智能工厂作为工业4.0时代的核心载体，其能耗管理水平不仅直接关系到生产效率的提升，更成为衡量企业可持续发展能力的关键指标。（剩余7369字）