Wind power prediction method based on historical meteorological data and stochastic simulation

Abstract

The invention relates to a wind power prediction method based on historical meteorological data and stochastic simulation. The method comprises the following steps of a pattern library training step based on the historical meteorological data: constructing training images according to the historical meteorological data, and capturing data events of the training images to establish a training imagemodel library; a wind power prediction step based on the stochastic simulation: establishing a wind power prediction region, and using the training image model library for traversing unknown nodes inthe wind power prediction region to obtain a wind power prediction result. Compared with the prior art, the method has the advantages of high prediction accuracy, small calculation amount, wide application in the field of wind power generation and the like.

Description

technical field [0001] The invention relates to the field of wind power generation, in particular to a wind power prediction method based on historical meteorological data and random simulation. Background technique [0002] Wind power generation is one of the most potential new energy generation methods in the world today. Large-scale development and utilization of wind power has become an effective measure for countries all over the world to solve energy and environmental problems and ensure sustainable development of the national economy. Since the natural wind that can generate electricity has strong uncertainties, wind power generation also has strong uncertainties. The uncertainty of wind power generation has brought many adverse effects to the power system, such as: the impact on power quality, the impact on system stability, the impact on power generation scheduling, and the impact on environmental protection. Among them, most of the impacts will increase with the ...

AI Technical Summary

Problems solved by technology

[0005] 1) The continuation method, which is the simplest prediction method, uses the sliding average of the nearest points as the wind power prediction value of the next point, but this method is relatively simple and the prediction accuracy is insufficient;

[0006] 2) Time series method, which regards the wind power at equal intervals in time sequence as a random sequence, and establishes a linear relationship model between the subsequent data and the previous data on this basis, such as: autoregressive, moving average, For models such as autoregressive moving average and differential autoregressive moving average, the time series method requires less modeling information, and the operation is simple, convenient and fast. However, for low-order models, the prediction accuracy is often not high enough, while the parameter estimation of high-order models is difficult more difficult;

[0008] 4) Artificial neural network method. Artificial neural network is a bionic method that imitates the structure and function of the human brain. This method is essentially a non-parametric method. It has the characteristics of parallel processing and self-organizing learning. It has strong Nonlinear mapping capability is currently a research hotspot in short-term wind speed and wind power prediction, but this method requires a long training time and a large amount of training data

Images