Research On Inversion Algorithm Of Thunder Cloud Charge Based On Multi-station Atmospheric Electric Field

The observation of the electrostatic field in ground is an important part of the atmospheric electricity, the detection of electric field in the ground becomes one of the important means of thunderstorms deep research. By analyzing observation data of thunderstorm electric field in ground can get the fractal characteristics of lightning so that we can strengthen the understanding of physical properties of lightning. Multi-station network can get the evolution and distribution information parameters of thunder cloud strong charge center by inversion. The inversion problem of thunder cloud charge structure is a complex nonlinear inversion process, the calculation results are influenced by many factors, the model has too many assumptions and estimates, at the same time, the performance of the algorithm is directly related to the precision of inversion.In this paper, inording to improve the fit goodness of inversion, hunder problems involved in the inversion of cloud charge structure, we combined the particle swarm method and Newton method through the combined mosaic hybrid structure and controlled the hybrid timing by constructing the mixed probability function, and a thunderstorm cloud equivalent charge hybrid inversion algorithm is proposed. The validity of the charge distribution pattern, the effectiveness of the proposed site distribution were discussed based on our proposed hybrid inversion algorithm. The main research contents include:Hybrid inversion algorithm of thunder cloud equivalent electric charge based on multi-station atmospheric electric field. Pair algorithm was chosen by studying the search feature of particle swarm method and Newton method’s. Global search ability of particle swarm is improved by improving particle swarm optimization (PSO) binding and mint of inertia weight. We studied the hybrid structure form, select embedded hybrid structure, put the Newton’s method into the particle swarm method through the establishment of the probability density function. A Hybrid Inversion Algorithm is proposed which can improve the calculation accuracy and ensure efficiency calculation.Analysis of influence of charge distribution pattern on the inversion result. Based on the distribution patterns of point charge, we discussed effective distance between sites. Then we talk about the effectiveness of point charge mode in the retrieval of a thunderstorm by analyzing of inversion results of different samples when the site distribution is in different forms. Established disc distribution patterns and analyzed the results compared with the point charge mode, we found that disc distribution patterns can effectively improve the overall fit goodness of inversion, difference of specific parameters can reflect the degree of the effectiveness of the model in different aspects.