Filling method based on PSO-GRNN (Particle Swarm Optimization-Generalized Regression Neural Network) for defect wind speed values of wind turbines in wind power plant

Abstract

The invention discloses a filling method based on a PSO-GRNN (Particle Swarm Optimization-Generalized Regression Neural Network) for the defect wind speed values of wind turbines in a wind power plant. The method comprises the following step: performing wind speed data similarity judgments in sequence; determining a parameter set to be optimized of a GRNN filling sub-model and evaluating a fitness function of the parameter set to be optimized; optimizing the parameter to be optimized through a PSO algorithm, and establishing a GRNN filling sub-model; and performing a GRNN combination by the GRNN filling sub-model to generate a final filling result. Specific to the disadvantage of poor stability of a conventional single filling model, measured wind speeds of a plurality of wind turbines having wind speed evolution most similar to the wind turbines subjected to defect wind speed measurement close to a defect sampling point are extracted, and the filling result of each sub-model is combined dynamically through a PSO-GRNN neural network combination filling model. Meanwhile, a determination scheme for a combined GRNN optimal smoothing coefficient is designed, so that the finial filling result accuracy and stability are further improved.

Description

technical field [0001] The invention belongs to the technical field of wind generators, and in particular relates to a method for filling missing wind speed values ​​of wind generators in wind farms based on PSO-GRNN. Background technique [0002] In order to effectively integrate wind energy into the power grid, it is extremely necessary and critical to accurately predict the output of wind farms. Among them, the short-term forecast of 0 to 6 hours is necessary for real-time scheduling of the power grid, ensuring grid frequency, power and voltage balance, etc. The technical parameters of grid security are of great significance. The integrity of the wind speed measured by wind turbines in wind farms is of great significance for the study of wind farm output, as well as for the study of wind turbine layout and the influence of wind turbine turbulence. After analyzing the collected data of many wind farms in my country, after the necessary data quality inspection, the most pr...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the technical problems raised by the above-mentioned background technology, the present invention aims to provide a method for filling the missing wind speed value of the wind turbine in a wind farm based on PSO-GRNN. Aiming at the shortcomings of the poor stability of the previous single filling model

Images