Research On Distributed Photovoltaic Energy Storage Monitoring And Energy Management System

Distributed photovoltaic systems produce electrical energy which can effectively alleviate the shortage of power through the photovoltaic effect,and the power generation process is green and environmental.However,the distributed photovoltaic power generation has the characteristics of high volatility and high randomness.When a large scale of photovoltaic energy is connected to the grid,it may bring a great impact to the public grid.Therefore,it is of great practical significance to monitor the state of the distributed photovoltaic power generation system,and form an effective photovoltaic energy management strategy by mining the monitoring data,in order to insitu consume the photovoltaic energy.This paper mainly studies the WiFi gateway module based on STM32,the photovoltaic power monitoring software platform based on Ali cloud,and photovoltaic power forecasting and energy management technology based on depth learning algorithm on the distributed photovoltaic energy storage power station hardware platform.The main research work finished in this thesis is as follows:(1)The control scheme of photovoltaic energy storage distributed generation and energy management system was given,and the key technologies of the system were studied and analyzed which included three parts following: data acquisition technology of power generation,the monitoring software platform based on the Ali cloud,and the power forecast and energy management technology of photovoltaic power generation based on the deep learning algorithm.(2)A photovoltaic energy storage distributed generation monitoring hardware system was designed.Using STM32 as control chip,based on ZigBee and WiFi communication technology,it helped to design a solar radiation acquisition module and gateway module,which were proved to be available to implement predetermined functions through experiments.(3)A photovoltaic energy storage distributed generation monitoring software platform was developed.Based on the Ali cloud platform,the photovoltaic monitoring cloud server was set up.In addition,the cloud monitoring software and energy management strategy were also studied and tested.Besides,a local monitoring APP and a remote monitoring APP were designed on the Android equipment.With the experimental results,all the software had been tested and optimized to implement the predetermined functions.(4)A photovoltaic power prediction model based on deep belief network and an energy management method based on photovoltaic power prediction were proposed.The deep belief network prediction model was trained on Matlab and compared with the classical BP neural network model.The smallest mean absolute percentage error(MAPE)of prediction result was only 7.49%,and all MAPEs of samples data were within 15%,which proved that the prediction model had a fine prediction accuracy.(5)The experiments of the designed system were completed,based on the existing photovoltaic energy storage distributed generation hardware platform.The experimental results had shown that the designed system implemented the functions of local monitoring,remote monitoring as well as cloud monitoring of the photovoltaic station.And the proposed photovoltaic power prediction model was of high accuracy and the proposed energy management strategy was effective.