Design And Implementation Of Short-Term Power Consumption Prediction System Based On Grey System Theory

In recent years,with the development of power system reform and market economy of China,all provinces and autonomous regions have set up power trading centers in response to national policies,electricity trading has gradually entered the marketization process.Industrial enterprises with large electricity consumption will directly participate in electricity market transactions by means of "annual long-term agreement" and "monthly bidding".Due to the difference of electricity consumption in each region and the existence of deviation assessment in electricity consumption declaration,enterprises must face the problem of electricity consumption declaration whether using "annual long-term association" or "monthly bidding".Therefore,this paper studies and designs a short-term electricity consumption prediction system,which is applicable to the actual situation of enterprises in industrial parks,to help local enterprises accurately report monthly electricity consumption to avoid the penalty of deviation assessment.This thesis analyzes the functional requirements of the short-term electricity consumption prediction for the enterprises in the industrial park in Guangxi ASEAN economic and technological development zone.This thesis gives the functional structure of the short-term electricity consumption prediction system for the enterprises in the industrial park,expounds the functions of the system modules,and establishes four database table structures including user information tables,historical data tables of electricity consumption,local meteorological data tables of the industrial park and data tables of electricity consumption forecast value.This thesis also designs the E-R chart of short-term electricity consumption forecast of enterprises in the industrial park.Meanwhile,this thesis designs a short-term power consumption prediction model and corresponding algorithm of industrial park enterprises based on the grey system theory,and optimizes the short-term power consumption prediction algorithm of industrial park enterprises by using residual correction method and combining climate factors of the industrial park with specific arrangement of enterprise production plan.Based on the Windows operating system,using Python language and SQLite3 database management system,this paper develops and realizes a short-term power consumption prediction system for industrial park enterprises.The experimental results on real data show that,compared with the basic forecasting algorithm,the short-term power consumption prediction results,which is obtained by the optimized prediction algorithm based on residual correction method while combining with the local meteorological factors of industrial park and the specific arrangement of enterprise production plan,are better.The implemented short-term power consumption prediction model system can accurately predict the future power consumption of industrial park enterprises,it plays a guiding role in formulating production and electricity consumption plans for enterprises in industrial parks.