Design And Implementation Of Electricity Big Data Analysis Platform Subsystem For Industrial And Commercial Customers

The widespread use of smart grids has enabled power companies to grasp a huge amount of electricity load data.Using big data analysis technology to analyze the electricity consumption behavior of customers and obtain the potential value hidden in it is an urgent problem for power companies to solve.Although there are some studies on electricity consumption behavior analysis methods in the academic field,most of them still use some traditional data analysis tools,which are insufficient for processing and calculating massive data,and the function structure of traditional algorithms is relatively single,which has certain limitations for in-depth and comprehensive research on customers’ electricity consumption behavior.At present,the customers of power grid companies mainly include residential customers and industrial and commercial customers,among which industrial and commercial customers account for a larger proportion of revenue and have higher requirements for power supply quality than residential customers.The ultimate goal of this paper is to design and implement a subsystem of the power big data analysis platform for commercial and industrial customers,providing a complete set of data storage,data analysis and visualization solutions for customers.For the objective of designing and realizing the analysis platform subsystem,this paper first investigates the technologies related to electricity consumption behavior analysis and clarifies the feasibility of system implementation,then conducts requirement analysis for the system functions,divides the specific functional requirements according to the research objectives,designs two functional requirements of customer electricity consumption rationality analysis and price based demand response,and clarifies the key issues to be solved in the process of subsystem implementation.The key issues studied in this paper include electricity consumption rationality analysis and price-based demand response for industrial and commercial customers.In order to realize the rational analysis of electricity consumption of users,this paper proposes an improved k-Shape-based evolutionary clustering algorithm,which innovatively introduces selfencoder and shape-based k-Shape clustering to make the clustering algorithm more adaptable to the characteristics of electricity data of industrial and commercial users.The comparison experimental results show that the proposed method effectively improves the clustering accuracy of user load data compared with that before the improvement;in order to In order to realize the price-based demand response,this paper innovatively proposes the kernel lasso-based forecasting algorithm and the reinforcement learning-based dynamic pricing algorithm,the kernel lassobased forecasting method introduces the kernel function based on the lasso model for improvement,and the experimental results show that the improved method performs better than other comparison methods in commercial and industrial load forecasting,and the reinforcement learning-based dynamic pricing method The innovative use of reinforcement learning algorithm divides the pricing model into two parts:customers and electricity sellers,and determines the retail electricity price through learning strategies during the interaction between the two.Experimental results show that the method can effectively reduce the load consumption of customers during peak electricity consumption periods.After solving the key issues,this paper designs and implements the subsystem according to the results of requirement analysis,divides the system into several modules and designs the key modules in detail,completes the construction of the subsystem,and designs and completes 12 groups of unit tests and integration tests.Finally,the main work of this paper is summarized,and the shortcomings and possible improvement directions of the analysis platform subsystem are proposed.