Research On Intelligent Power Utilization Strategy Of Industry Customers Based On Deeply Mining Of Multivariate Data

In the background of the continuous promotion of the power system reform,the main body of the electricity market has become more diversified.The electricity market in China has also entered a new stage with the demand of products as the guidance and the customer satisfaction as the purpose.According to the theoretical method of cluster analysis,this paper makes deeply mining of the daily distribution characteristics and the annual distribution characteristics of the industry power consumption.Then aimed at the peak type industry customers,the temperature is corrected by the influence of the urban micro meteorology and the relevance between the load and the corrected temperature is analyzed.At last,based on the above research results,the formulation process of the power use package of the industry customers is designed.The main contents of this paper are as follows:(1)Based on the data mining algorithm,the daily load and annual load of different industry customers are deeply excavated and analyzed,and the daily distribution characteristics and the annual distribution characteristics of industry power use are obtained.Among them,the industry customers' load are divided into four categories including peak type,peak-averting type,high-load-factor type and continuous type by the daily distribution characteristics of industry power consumption.Besides,the annual distribution characteristics of industry power consumption are analyzed from three aspects of the industry's annual power consumption mode,open or shutdown state and holiday sensitivity.Among them,the analysis of open or shutdown state is divided into full-start mode,intermittent-shutdown mode and complete-shutdown mode according to the downtime curve.And about the analysis of holiday sensitivity,the customer load is divided into holiday-insensitivity type,holiday and weekdays balance type and holidaysensitive type according to the defined index of holiday occupation.(2)Taking the industry customers of peak type as the research object,the relevance between the industry customers' load of peak type and the corrected temperature is analyzed through the study of the temperature correction considering the impact of urban micrometeorology to screen out industry customers who are closely related to the temperature.Then for the target customers,the temperature correction model considering the temperature and humidity effect and the temperature cumulative effect is established,and the effectiveness of the temperature correction model is illustrated by an example,which provides the basis for the customization of intelligent power use package below.(3)Taking the industry customers screened out of peak type and temperature sensitive as the research objects,the establishment of the benchmark peak-valley time-of-use price and the superimposed price is studied.And then the formulation process of the power use package of the industry customers is designed.In the study of the establishment of the benchmark peak-valley time-of-use price,based on the user response model of consumer psychology,the target function is set according to the method of similarity measure.Finally,the optimization model of peak-valley time-of-use price is established,and the time section of peak,flat and valley is divided according to the semi trapezoid fuzzy membership grade function.Besides,in the study of the superimposed price model,the influence of holiday and temperature on customer load is reflected by holiday occupation index and temperature sensitivity index,and the superimposed price of peak and flat section is finally obtained by calculating the difference of customer load in the peak and flat period.