Fan short-term generated power prediction method

Abstract

The invention discloses a fan short-term generated power prediction method. The fan short-term generated power prediction method comprises the following steps: firstly, reasonably pre-selecting parameters which have close influence on wind power output, denoising pre-selected original input data by adopting a complete set empirical mode decomposition algorithm and a correlation coefficient method,removing noise signals of which the correlation coefficient is lower than 0.1, carrying out refitting, and carrying out data reconstruction on denoised data to sequentially obtain previous moment data; adopting a decision tree method to perform feature selection on the input data, and eliminating data with importance lower than 0.9 to obtain final input data; and finally, giving a predicted valueby using a least square support vector machine algorithm. The fan short-term generated power prediction method provided by the invention combines a data denoising method and a data-driven modeling method to perform wind power prediction, and belongs to the field of clean energy power generation prediction. The fan short-term generated power prediction method provided by the invention has the advantages of simple steps and low calculation cost, and is suitable for actual prediction application.

Description

technical field [0001] The invention relates to a method for predicting short-term generating power of wind turbines, in particular to a method for predicting generating power of wind turbines based on a complete set of empirical mode decomposition algorithm, decision tree algorithm and least squares support vector machine algorithm, belonging to the field of clean energy power generation prediction . Background technique [0002] Affected by environmental pollution and the depletion of fossil energy, wind power has been widely used as an efficient and clean energy. However, wind power generation has obvious chaotic characteristics and volatility, and it will have a serious impact on the power grid after being connected to the grid. Predicting the power generated by wind turbines is helpful for early warning of fluctuations in wind power generation, which is conducive to the safe operation of the power grid. [0003] For a series of problems caused by the randomness and vo...

AI Technical Summary

Problems solved by technology

[0004] However, the existing wind energy forecasting methods still have the following deficiencies: 1) The existing wind energy forecasting methods rely on the meteorological information of wind farms; 2) The computational cost of the existing forecasting methods is high; 3) There is still much room for improvement in the forecasting accuracy

Images