Research On Ultra-broadband Multi-frequency Point Solar Radio Image Processing Algorithm

Since the birth of solar radio astronomy,a variety of solar observation equipment has been developed with the efforts of astronomers from all over the world.China’s scientists have made great contributions in the field of solar radio observation,from Huairou Observatory,Miyun Observatory,to the establishment of Ming’antu Observatory,improving the level of solar radio observation in China,and making our country squeeze into the world’s leading radio observation technology countries.China’s ultra-wide spectrum radioheliograph(MUSER)is located in Ming’antu.Due to the limitations of comprehensive aperture imaging principle,array composition and cost,MUSER’s images contain many false components,and specific image processing algorithms are needed to reconstruct them later.There is no international processing algorithm for solar radio images with ultra-broadband multispectral spectrum.This thesis is based on this background to carry out research.Main work is as follows:1.This thesis discusses the development of solar radio astronomy and related solar radio observation instruments at home and abroad.The parameter composition,imaging principle and array design of MUSER are emphasized.This thesis introduces the current situation of solar radio image processing algorithms at home and abroad.The idea of the H(?)gbom CLEAN algorithm is described in detail,and the multi-scale CLEAN and the maximum entropy algorithm are briefly introduced.2.Using the weighted function to compensate for the UV coverage of MUSER,mainly in the form of natural weighting and uniform weighting.Three solar radio images from Japan’s Nobian Mountain were selected as real radio power sources and simulated different frequency channels to observe the sun.And with the weighted function of the UV cover to get different dirty beams,and then use dirty beams and wild mountains to get three suns to do convolution.Such different frequency channels,different weighty methods,and the combination of different solar images produce multiple sets of dirty maps.The cleaning effect of the H(?)gbom CLEAN algorithm in the above-mentioned groups of dirty diagrams is then studied.In this way,the simulation of the dirty graph devolution under different frequency bands and under different weighting methods is realized.Through the above research,this paper has realized the research on the effect of solar radio image reconstruction on the ultra-wideband multi-spectrum.And some preliminary conclusions have been obtained.After natural weighting of UV coverage,when the H(?)gbom CLEAN algorithm is used for cleaning treatment in the later stage,the overall restoration effect is good in the frequency range of 1GHz-8GHz,but the effect is very poor in the frequency range greater than 9GHz;UV coverage of uniformly weighted processing,when the same algorithm is used to process dirty images,the overall result is not as good as the processing effect of natural weighting.The performance is acceptable in the range of 3GHz-8GHz,and the recovery result is very poor in the frequency range of less than 3GHz and greater than 9GHz.This result provides a good reference for the image processing of the Mingantu radio heliograph.