Research On The Radio Heliograph Array Design

This dissertation is the synchronous academic achievement and memorandum ofthe National Great Scientific Project "Cm-dm Band high resolution spectrum radioheliograph"led by professor Yan Yihua, the chief scientist of National AstronomicalObservation Center, CAS. Radio heliograph is a large scale antenna array to image theradio sun, its successful reseach and construction will supply the internationalobservation gap of solar imaging in dm band. It has not only the title of pioneer inhomeland, but also has the strong international competitive power.Guided by the heliograph working principle, the dissertation employed themulti-subject co-operational research methods synthetically such as manufacture systemengineer, virtual design method, computer simulation, topology optimization design,error analysis, numerical analysis and image reconstruction, and conducted the researchof cell antenna design, the performance-determining factors of antenna array, antennaarray topology optimization design, array calibration and image reconstructionsimulations.The main creative results are as follows:1.In the part of cell antenna structure design, according to the performancedemand of radio heliograph, confirmed the performance specification of cell antenna.Determined the pedestal style of horse-hoofed structure. Proposed the general structurescheme of closed worm wheel plus spiral handspike. Designed the cell antennamechanical structure using the virtual design method. Conducted the finite elementmechanical analysis and mode analysis of reflector whose structure of two-lamella wasfirst used. Analysised the pointing error caused by structure error and axis system error.2.In the part of array topology optimization design, the antenna array of our radioheliograph developed tended to have the spiral profile resulted in the limited budgetwhile the Boone's Pressure Algorithm could only used for irregularly arrayconfiguration. The Accelerated Pressure Algorithm (APA) proposed not only could getthe regular array, but also had the higher computing efficiency. The synthesis beam ofoptimized spiral array achieved by the APA has the resolution of 20.5 arcseconds, thefield-of-view maximum side-lobe of 0.15 when working in the frequency of 1GHz andnatural weighting, which is superior obviously to that of standard spiral array3.In the part of array calibration and error correction, aim at the limitation of leastsquare model to deal with observed data with unexpected values, a new model ofmaximum probability density function to fit the complex visibility was proposed on theassumption that the error had a gauss distribution. It could conclude that the least squaremodel was only one especial case of the proposed model. The fit model of visibilityfunction was commonly solved by iterative method of Matrix Inversion. For our radioheliograph, this method had the limitation of unstable stability as the result of thelarge-scale constraint functions. Therefore, a grafted genetic algorithm was presented tosolve the gain during the self-calibration process, which featured the rapid speed ofevolution and its strong ability to avoid premature convergence. Simulation indicatedthat the dirty map obtained by grafted genetic algorithm based on the new proposed fitmodel was excellent than the one obtained by iterative method of Matrix Inversionbased on the least square model.4.In the part of heliograph imaging simulation, clean algorithm has the limitationof iterative dithering and convergence instability when applied to extend sourcesprocessing. Inspired by the ability of Maximum Entropy Method to proceed the extendsources, the paper proposed a clean algorithm with smooth constraint by modifying thedirty beam. Finally the simulation was conducted of radio heliography solar image atlow band. Results show superiority of the improved algorithm in both image quality andresolution. For images reconstructed at high band, the paper inherited and developed theSkilling Maximum Entropy Method (SMEM). An Improved Skilling Maximum EntropyMethod (ISMEM) was presented through the innovation in search directions of SMEM.Both the two directions, one came from the large amount of simulation experiments andthe other reasoned out from Newton-Raphson algorithm, combines the speed withexactness harmoniously as the result of error adjustment.