基于SAX-ARMA模型混合模型的企业排污用电监测  

Algorithm for Pollution Emission Monitoring and Warning in Enterprises Based on Symbolic Aggregate Approximation-Autoregressive Moving Average Model Hybrid Model

作　　者：侯天玉 敬如雪 张凌薇 HOU Tianyu;JING Ruxue;ZHANG Lingwei

机构地区：[1]国网武威供电公司,甘肃武威733000

出　　处：《电力系统装备》2024年第1期163-167,共5页Electric Power System Equipment

摘　　要：随着社会的快速发展和工业化进程的加速,环境污染问题日益凸显,加强环境监管与治理成为刻不容缓的任务。为了有效地控制企业排污,提高环境质量,利用电力数据易获取、数据量大等特点,通过多种算法研究,使得企业排污用电预测更精准,实现企业排污用电的精确监测。文章旨在利用企业排污用电数据,结合ARMA算法和SAX算法的优势,构建SAX-ARMA混合模型,对企业排污用电进行监测预警。实例分析选取特定区域的排污企业用电数据作为研究对象,将预测结果与实际数据进行对比,并与设定的阈值进行对比,展示了该方法在实际应用中的效果和可行性,实现排污企业的短期排污数据预测,从而加强对重点企业污染防治的联防联控。通过对比分析ARMA模型和SAX-ARMA混合模型的预测结果,和模型的评估指标的值,发现SAX-ARMA混合模型比ARMA模型的预测值和真实值的重合程度更高,且SAX-ARMA混合模型的RMSE和MAPE的值比ARMA模型分别低0.151和3.304,综上所述SAX-ARMA混合模型更适合文章的研究内容。With the rapid development of society and the acceleration of industrialization,environmental pollution has become increasingly prominent,and strengthening environmental supervision and governance has become an urgent task.In order to effectively control enterprise emissions and improve environmental quality,the use of electricity data,which is easily accessible and has large amounts of data,can be utilized to study various algorithms to make the prediction of enterprise emissions more accurate and to monitor enterprise emissions.This article aims to use enterprise emissions and electricity consumption data,combined with the advantages of ARMA algorithm and SAX algorithm,to construct a SAX-ARMA hybrid model for monitoring and early warning of enterprise emissions.A case study was conducted using electricity consumption data from specific pollution enterprises in a certain region as the research object.The predicted results were compared with the actual data and the threshold set,demonstrating the effectiveness and feasibility of this method in practical applications,and achieving short-term prediction of emissions data for pollution enterprises,thereby strengthening the joint prevention and control of pollution for key enterprises.Through a comparative analysis of the prediction results of the ARMA model and the SAX-ARMA hybrid model,as well as the values of evaluation indicators of the models,it was found that the SAX-ARMA hybrid model had a higher degree of overlap between the predicted values and the actual values compared to the ARMA model.The RMSE and MAPE values of the SAX-ARMA hybrid model were lower than those of the ARMA model by 0.151 and 3.304,respectively.In conclusion,the SAX-ARMA hybrid model is more suitable for the research content of this article.

关 键 词：SAX-ARMA 电力大数据 环保 排污 

分 类 号：TM73[电气工程—电力系统及自动化]