Research On Web-based Photovoltaic Data Visualization Monitoring System

In recent years,the global photovoltaic industry has developed rapidly,and the number of photovoltaic power plants has continued to increase.However,photovoltaic power plants are usually built in places with inconvenient transportation and harsh environments,and they have been running unattended for a long time.It is difficult for operation and maintenance personnel to monitor,manage and analyze photovoltaic data.In addition,many photovoltaic power plants have problems such as cumbersome system operation,chaotic data management,difficult data analysis,inconsistent databases,and low operation and maintenance efficiency.They usually provide monitoring data in traditional ways,and there is no correlation between data,and data retrieval efficiency is low.This paper combines visualization technology with Web technology,adopts browser/server(B/S)mode to design and develop a set of photovoltaic data visualization monitoring system,constructs a scientific photovoltaic database structure,and visually presents photovoltaic data to operation and maintenance personnel in the form of visual charts and rich interactive operations.It is convenient for them to quickly understand the characteristics of the data,quickly find and solve problems,and perform an in-depth analysis of the data through data mining algorithms.The main tasks completed in this paper include:(1)Analyzed the requirements of the visual monitoring system for photovoltaic power generation data,and design the system architecture using the B/S model.Investigate the project overview and photovoltaic data types of the Emaozhai photovoltaic power station in Pan County,Guizhou,and put forward the overall design goals and implementation plans of the system.(2)Carry out the analysis of the asynchronous communication mechanism for the realtime refresh of photovoltaic data,and study the low-coupling layered operation and maintenance mode of the Web back-end photovoltaic data monitoring system,analyze the realization method of the photovoltaic data time series visualization on the Web,and established the photovoltaic database object relational mapping(ORM)management model.A prediction method of photovoltaic power generation based on support vector regression(SVR)is proposed.(3)The visualization process was designed according to the data requirements of the Emaozhai photovoltaic power station,the mainstream visualization tools were analyzed,and the Web front-end photovoltaic data visualization framework was established based on SVG and Canvas technology,and the functional module design of the front-end photovoltaic data visualization system was completed.(4)Based on Django and MySQL,the functional design of the back-end photovoltaic data monitoring system was completed,the demand analysis of the photovoltaic database,and the design of the database table were carried out,and the design of the back-end management module of the Web system was completed.The photovoltaic data preprocessing process was completed,the data dimensionality reduction was completed using random forest regression(RFR),and the photovoltaic power generation power prediction model was established based on SVR and the error was verified.(5)Cloud server deployment of the system and testing of various functional modules.The test results show that the web photovoltaic data visualization monitoring system designed and implemented in this paper has normal functions,simple and practical interface design,good visualization effects,and effectively improves the efficiency of operation and maintenance.The effect is in line with expectations.