Research And Implementation Of Electricity Consumption Behavior Analysis Platform For Industrial And Commercial Users Based On Electric Power Big Data

With the finalization of the 14th Five-Year Plan,digital new infrastructure,as an important content of "new infrastructure,new urbanization initiatives and major projects",marks that all walks of life will usher in the wave of digital transformation.As a key industry related to the national economy and the people’s livelihood,electric power industry is also accelerating the deep integration of digital technology and traditional grid industry.Big data is the core of digital transformation,and data mining technology is the key technology to help grid digital transformation.Utilizing big data mining and intelligent analysis technology to mine users’ electricity consumption behavior can enable grid companies to better design energy management plan and ensure the stable supply of electricity.Compared with ordinary residential users,industrial and commercial users have a large demand for electricity,so they are the key service customers in the transformation and upgrading of the grid industry.Therefore,this paper design and implement electric consumption behavior analysis platform for industrial and commercial users based on the electric power big data,which can identify the typical load pattern curves of industrial and commercial users in various industries and carry out real-time load forecasting.This platform can assist the electric power department in intuitively and quickly monitoring and analyzing the electricity consumption behavior of industrial and commercial users,providing data and platform support for the electric power department to formulate relevant energy management plans,and enterprise users to formulate energy storage plans.The mining of electricity consumption behavior of industrial and commercial users in this paper is mainly divided into two aspects:extracting typical load pattern curve based on evolutionary clustering algorithm,and performing load forecast for industrial and commercial users based on load forecasting algorithms.To overcome the problems of sensitivity to initial values and inapplicability of high-dimensional data in traditional clustering algorithms,the evolutionary clustering algorithm VAE-DenStream proposed in this paper first compresses high-dimensional electricity consumption data into the low-dimensional hidden space based on the inference network of VAE model to extract electricity consumption features,and then uses DenStream clustering algorithm to cluster the extracted electricity data features.The evolutionary clustering algorithm proposed in this paper is a clustering method based on incremental computing,which can achieve clustering quality comparable to static clustering algorithms while reducing the storage of historical data;At the same time,compared with other benchmark evolutionary clustering algorithms such as the evolutionary k-means algorithm,the evolutionary clustering algorithm proposed in this paper has better current clustering quality and time smoothness.Meanwhile,this paper proposes a load forecasting method based on typical load pattern mining.This load forecasting algorithm effectively classifies historical load data based on similarity by inferring the load pattern category of the day to be predicted.Then,historical load data with the same load pattern is used as input to the load prediction model,weakening the impact of electricity consumption data from different load patterns on the prediction results.The experimental results prove the effectiveness and accuracy of the short-term load forecasting method proposed in this paper.Based on the research results of electricity consumption behavior mining for industrial and commercial users,this paper aims to design and implement an electricity consumption behavior analysis platform for industrial and commercial users based on electric power big data.In order to realize the platform,this paper first analyses the requirements of the system according to the specifications of software engineering,and defines the functional requirements and non-functional requirement that the system needs to meet.Then,this paper provides an overall and detailed design of the system,and establishes a web-based platform for analyzing the electricity consumption behavior of industrial and commercial users.Finally,a series of tests prove the availability and stability of the system.The platform enables the electric power department to monitor the load curve of industrial and commercial users flexibly and intuitively,analyze the typical load patterns and evolution rules,and carry out real-time load forecasting,which is of great significance for the electric power department to grasp the overall electricity consumption of commercial users,ensure the safe and stable supply of electricity,and promote the green and low-carbon development of energy.