Distributed Photovoltaic Scenario Generation Based on Edge-Cloud Collaborative Architecture

With the expansion of the power system and the growth in renewable energy penetration, the installed capacity of distributed photovoltaics in the power system will increase significantly in the future, and the amount of related data will increase sharply in the future. Traditional centralized cloud management is difficult to afford the massive data processing business to guarantee service quality. This paper proposes an edge-cloud collaboration architecture based on Kubernetes. At the edge, we cluster the distributed photovoltaics to the nearby centralized photovoltaic power plants with strong correlation and synthesize as a virtual station. Subsequently, we use historical meteorological and power data to predict the next day power curve. At the cloud, we generate the scenarios for the centralized station based on the day-ahead power data uploaded by the edge, then we obtain the scenarios for distributed photovoltaic cluster by converting the scenarios of centralized station with the capacity conversion formula.

I. Introduction

With the continuous expansion of the power system and the penetration of renewable energy, the uncertainty of the power system increases. The access of a large number of distributed photovoltaics in the future will bring a large amount of data processing requirements to the system. The traditional centralized cloud is difficult to process and manage massive data, and the data transmission process will easily cause network bandwidth congestion. Therefore, how to timely process and analyze the large amount of data generated by a large number of renewable energy devices connected to the system in the future is of great significance for the uncertainty analysis of the power system. The edge-cloud collaborative architecture, relative to the traditional centralized cloud, is more like the sinking of cloud computing capability from the center to the edge. It can effectively reduce the delay and transmission cost, and is suitable for the occasions where the preliminary data uploaded by the aggregated terminal needs to be processed [1].