Solar 10.7 cm radio flow forecasting method based on BP neural network

Abstract

The invention belongs to the technical field of space weather prediction, and particularly relates to a solar 10.7 cm radio flow forecasting method based on a BP neural network, which comprises the following steps: acquiring observation value data of an F10.7 day and preprocessing the observation value data; converting the preprocessed data into supervised learning sample data capable of being trained by a BP neural network, and performing training set and test set division on the sample data; establishing a BP neural network according to the sample data; using the F10.7 training set to train a BP neural network; forecasting the trained BP neural network by using an input part of an F10.7 test set to obtain a forecast value, and performing error analysis on the forecast value and an output part of the test set to obtain forecast accuracy of the BP neural network; and inputting corresponding F10.7 historical data into the trained BP neural network to obtain prediction data. According to the method, the forecasting model of any period can be conveniently built, and the medium and short term forecasting precision of F10.7 can be effectively improved.

Description

Technical field [0001] The present invention belongs to the field of space weather prediction, and is specifically involved in a BP neural network based on the sun 10.7 cm radion flow forecasting method. Background technique [0002] The sun's F10.7 parameters refer to the radiation flow value of 10.7 cm wavelength, unit 10 -22 W / M 2 / Hz. In the space weather field, this value is defined as the sun 10.7 cm radiation radiation index, ie F10.7. The system's F10.7 observation record begins in 1947, located in the south of Ontario, Ontario, Canada, and later, due to the deterioration of observing the environment, a new facility was built in Ottawa's 250 kilometers in West York in 1962 and replaced the old Observation point. In 1991, the observation point was again moved to Penscon and served as a measurement task. F10.7 observed task is performed by a radio telescope installed on the ground, and the technology is very mature, and data can be precisely acquired under almost all met...

AI Technical Summary

Problems solved by technology

The method of constructing polynomials needs to find the mathematical polynomial expression of F10.7 in a specific forecast period. This method is also a medium-to-long-term trend forecast method, and there are shortcomings in the flexibility of the forecast period and the accuracy of short- and medium-term forecasts

Images