Design And Implementation Of Photovoltaic Management And Control Cloud Platform Based On Hadoop

During the 14 th Five-Year Plan period,under the favorable policy of the country’s formulation of dual carbon goals,renewable energy generation gradually occupies the core position in the new power system,in which photovoltaic power generation is considered to be the most reliable choice of the whole renewable energy,and all countries have issued policies to support the photovoltaic industry.With the gradual increase in the scale of construction and grid-connection of photovoltaic power plants in various provinces,most photovoltaic power plants are built in remote places with harsh environments,and each photovoltaic power plant is equipped with a set of independent local monitoring,which can only be managed and controlled independently.The actual operation and maintenance management still adopts the traditional method of paper or Excel spreadsheet.There are problems such as difficulties in data statistics,and the functions are relatively insufficient,so that the group cannot know the power station information in time.Therefore,it is necessary to study a set of photovoltaic management operation and maintenance cloud platform realistic meaning.In view of the integrated operation and maintenance of photovoltaic power plants and the group’s demand for displaying real-time power station information,this dissertation develops a Hadoop-based photovoltaic management and control cloud platform,which centrally manages and controls all power plants,and applies the core technologies and concepts of big data to the system.Developed and designed a cloud platform for photovoltaic tube control based on Spring Cloud micro-service framework,and the collected data is applied in engineering.The thesis mainly involves works as follow:1.Conduct on-site requirement research on operation and maintenance of photovoltaic power plant,design and develop two sets of cloud platform systems for photovoltaic tube control.The platform is implemented by front-end and back-end separation thought and mainstream micro-service architecture,covering 15 modules,such as asset management,equipment point management,patrol management,report statistics,file management,real-time monitoring,system management,etc.2.Research and use Hadoop and other big data storage technologies to store the collected massive device point data into Hive,and establish a historical data warehouse to deal with the problem of large-scale data storage and acquisition.3.The power and related data collected by the platform are deeply studied,the photovoltaic power prediction model based on K-means similar day clustering and RBF neural network is established,and the simulation test is carried out.The research shows that the prediction algorithm has high accuracy and can meet the requirements of engineering applicationBased on the demonstration project of the company,the system is developed by using micro-service framework and big data technology.The machine learning algorithm is used to build the photovoltaic power prediction model,and the requirements for information-based production operation and maintenance management and control are initially realized.After many project research,construction,development and testing,the sub-stations under the company’s jurisdiction have been put into use,and the functions meet the needs of power station.