The Research Of Doppler Radar Signal Processing Algorithm And Software Development

Along with the new generation Doppler weather radar system being built all over the country, the signal processing methods of single Doppler radar have been becoming one of the most important tasks of scholars in the fields of signal processing and weather research. With the view of software algorithm realization, the paper presents a systemic analysis to the signal processing of single Doppler radar, especially deals with the application of wavelet analysis in the field of signal processing for Doppler radar. Our major contributions are as follows:Firstly, the median filter is used for data pretreatment to get rid of interferential signal and improve signal quality. Secondly, a method named "dummy filling" is suggested. And with the hypothesis of the continuity of the wind field, the missing data are complemented. Which makes the resulted radar signal in possession ofthe huge scope character. Searching the boundary of the aliasing region by sectionjudgment technology of computer graphics, we dealiase the aliasing data. Moreover,based on the analysis of the existed wind field retrieval methods of single Doppler radar, we derive a novel algorithm for wind field retrieval from wavelet analysis. It transforms the problem of wind field retrieval into the sparsity of wavelet coefficient, i.e. entropy function minimizing. In order to solve the corresponding non-smooth optimization, we develop an algorithm named MAST. The performance of the algorithm is verified from the theory study and computer simulations. Finally, we apply the "dummy filling" for velocity dealiasing, the algorithm for single Doppler radar wind fields retrieve from wavelet analysis, the technology of downburst exploring and the OOP to the software project of single Doppler radar signal processing.