Method for predicting daily generating capacity of grid-connected photovoltaic power station based on factor analysis

Abstract

The invention relates to a method for predicting the daily generating capacity of a grid-connected photovoltaic power station based on factor analysis. The method comprises the steps of S1, obtaining historical meteorological observation values and daily temperature range, articulation index and historical photovoltaic generating capacity data, and conducting standardization processing; S2, conducting factor analysis on standardized data, conducting factor axis rotation by means of the maximum variance method, and extracting two common factors obtained after rotation; S3, establishing a photovoltaic power generating capacity predicting model, conducting multivariate regression analysis with the two extracted common factors as an input variable and the standardized historical photovoltaic power generating capacity data as an output variable, and obtaining a regression coefficient in the model; S4, obtaining two new common factor values calculated through meteorological element predicating data; S5, substituting the two new common factor values and the regression coefficient into the predicting model, and obtaining the predicted value of the daily photovoltaic power generating capacity through inverse standard conversion. Compared with the prior art, the method has the advantages that the algorithm and modeling are easy, and generating capacity predicting accuracy is high.

Description

technical field [0001] The invention relates to a method for predicting the power generation of a power station, in particular to a method for predicting the daily power generation of a grid-connected photovoltaic power station based on factor analysis. Background technique [0002] Solar photovoltaic power generation is a solar energy utilization method with high conversion efficiency, long service life, and can provide a large amount of electricity, and the technological development potential of solar energy is the highest among the six renewable energy sources, which is more than 10,000 times the current global energy demand . In actual development and utilization, solar photovoltaic (PV) technology has developed rapidly in the past ten years. By the end of 2012, the cumulative installed photovoltaic capacity in the world had reached 102 gigawatts (GW), and it has become the world's renewable energy after hydropower and wind power. most important source. my country is r...

AI Technical Summary

Problems solved by technology

However, most of the previous studies focused on the incident total radiation and panel temperature of photovoltaic cells under laboratory conditions, and it is necessary to accurately measure such as the incident total radiation and panel temperature, and the current-voltage (I / V) characteristic curve of photovoltaic cell output. etc., the cost of obtaining these data is high, and even cannot be obtained in practice, and it cannot really represent the actual operating conditions of outdoor photovoltaic power plants; secondly, it is necessary to introduce an additional grid-connected inverter efficiency model or grid-connected inverter output model , used to predict the final AC output of the photovoltaic grid-connected power generation system; at the same time, according to this method, there will be many intermediate links, and each conversion stage may introduce system errors

[0006] In view of the above reasons, traditional photovoltaic power generation prediction methods are difficult to meet the practical, simple and predictive requirements of photovoltaic power plants and power systems.

[0007] Chinese patent 201110119239.7 discloses a method for predicting photovoltaic power generation. The invention filters and classifies the historical power generation data of photovoltaic power plants, and reprocesses the meteorological data to obtain the total radiation of photovoltaic power plants. Then the weather type, temperature, humidity , season, and total radiation are normalized and used as the input of the prediction model, and the 12-hour or 4-hour photovoltaic power generation is predicted by neural network calculations, but this method does not consider the nature of various influencing factors and many factors The multi-collinearity or coupling relationship between them affects the prediction accuracy of the model, the calculation of the neural network is complex, and the prediction time is short

Images