智能化视角下造纸企业能耗预测神经网络算法优化研究  

Research on Optimization of Neural Network Algorithm for Energy Consumption Prediction of Paper Making Enterprises from the Perspective of Intelligence

作　　者：王小春 张宏甫[1] WANG Xiaochun;ZHANG Hongfu(Xi’an Aeronautical Polytechnic Institute,Xi’an 710089,China)

机构地区：[1]西安航空职业技术学院,陕西西安710089

出　　处：《造纸科学与技术》2025年第4期98-101,105,共5页Paper Science And Technology

基　　金：陕西省教育科学规划课题(SGH20Y1656)。

摘　　要：造纸行业是高能耗产业,能源成本在制造成本中占比较高。准确预测能耗对造纸企业精细化管控、节能降耗具有重要意义。基于此,分析了造纸企业能耗预测的重要性以及制浆、抄纸、废纸利用等影响因素,提出优化能耗预测神经网络算法的五大策略:融合制浆造纸工艺流程数据优化样本、针对造纸能耗特点设计网络结构、结合制浆造纸生产节奏改进训练算法、契合造纸企业能效管控需求设计评价指标、面向制浆造纸多工序联动进行模型集成。通过算法优化,提升造纸企业能耗预测的准确性,为智能化节能管控提供数据支撑,推动造纸行业高质量发展。The paper industry is a high energy consumption industry,and energy costs account for a relatively high proportion of manufacturing costs.Accurate prediction of energy consumption is of great significance to fine control,energy saving and consumption reduction of papermaking enterprises.This paper analyzes the importance of energy consumption prediction in papermaking enterprises and the influencing factors such as pulping,papermaking and waste paper utilization,and puts forward five strategies to optimize the neural network algorithm for energy consumption prediction:Optimize the sample by integrating the pulp and paper process data,design the network structure according to the characteristics of paper energy consumption,improve the training algorithm combined with the pulp and paper production rhythm,design the evaluation index to meet the energy efficiency control needs of paper enterprises,and integrate the model for the linkage of pulp and paper multi-process.Through algorithm optimization,we can improve the accuracy of energy consumption prediction of paper enterprises,provide data support for intelligent energy saving management and control,and promote the high-quality development of the paper industry.

关 键 词：智能化 造纸企业 能耗预测 神经网络 算法优化 

分 类 号：TS78[轻工技术与工程—制浆造纸工程]