Wind power prediction method based on improved depth extreme learning machine

Abstract

The invention discloses a wind power prediction method based on an improved depth extreme learning machine, and the method comprises the steps: S1, obtaining wind power data, and constructing a DELM model of the depth extreme learning machine; S2, carrying out normalization processing on the collected wind power data, and initializing DELM model parameters; S3, introducing a sobol sequence to initialize a population in a standard grey wolf algorithm, and introducing a Cauchy variation strategy in local search; S4, sending the preprocessed training sample into a DELM model, taking a composite weighted objective function value obtained through a DELM training test as a fitness value of a grey wolf algorithm improved by a mixed composite weighted objective function, and performing training optimization on parameters by using the improved grey wolf algorithm to obtain an optimal node number and an optimal regularization parameter C of each hidden layer of the DELM; and S5, sending a test sample of the wind power data into the DELM model containing the optimal parameters to obtain a wind power prediction output value. According to the invention, the wind power prediction is closer to the actual value, and the accuracy of wind power prediction is improved.

Description

technical field [0001] The invention belongs to the technical field of wind power prediction, and in particular relates to a wind power prediction method based on an improved deep extreme learning machine. Background technique [0002] With the sharp reduction of non-renewable resources, the research on renewable resources has received extensive attention. Wind power, as one of the renewable resources, has made the application technology of wind power more mature and more efficient through years of continuous and in-depth research. However, due to the fluctuating and intermittent characteristics of wind power, with the continuous expansion of the grid-connected wind power system, this will bring great challenges to the safe and stable operation of the power system. Doing a good job in wind power forecasting can effectively improve the ability of the power system to operate safely and stably, which is of great significance for the use of wind power generation. [0003] At p...

AI Technical Summary

Problems solved by technology

The Kalman filter method is suitable for real-time online prediction, but it needs to use the statistical characteristics of known noise, which makes its practicability poor; although the neural network method has good learning ability, but the convergence speed is slow, and it is prone to overfitting or falling into local The optimal solution and other shortcomings are removed; the continuous method is simple to operate, but it is only suitable for ultra-short-term forecasting, and its application range is narrow

Images