Power estimation method of low-voltage distributed photovoltaic generation based on similarity aggregation

Abstract

Regarding to the issue that there is no automatic data acquisition for low-voltage distributed photovoltaic generation, a power estimation method for low-voltage distributed photovoltaic generation based on similarity aggregation is proposed in this paper. Firstly, based on the time series power curve data characteristics of middle-voltage distributed photovoltaic generation and L2 norm distance, a middle-voltage distributed photovoltaic aggregation model is established, and a solution method of middle-voltage distributed photovoltaic aggregation based on K-means clustering is proposed; Secondly, based on the aggregation results of middle-voltage distributed photovoltaic, a parameter calculation method for the power estimation of low-voltage distributed photovoltaic based on geographical distribution and electricity data is proposed; Thirdly, considering the time series data characteristics of distributed photovoltaic changing with date, a strategy to determine whether to re-calculate the middle-voltage distributed photovoltaic aggregation and the power estimation parameters according to the cumulative error is proposed. Finally case studies are accomplished and the effectiveness of the proposed method are verified. The aggregation of middle-voltage distributed photovoltaic and the power estimation of low-voltage distributed photovoltaic, which is helpful to master the overall power generation of regional distributed photovoltaic, realize the observability of low-voltage distributed photovoltaic power generation, improve the safe operation level of distribution network, and has important theoretical and practical significance for serving the orderly development of distributed photovoltaic generations.

Keywords

• Power estimation
• Low-voltage
• Distributed photovoltaic generation
• Similarity aggregation
• L2 norm
• K-means