Research And Application Of Evolutionary Algorithms In Radar Data Inversion

The constrained optimization problems are widespread but difficulty to solve in practical engineering and scientific research. At present there is not yet generally effective method for such problems. Lots of the existing algorithm is based on the gradient optimization can’t do it, for its target function and constraints discontinuous differentiable, and such algorithms usually can only get local optimal solution.Evolutionary algorithms are a kind of simulation of the natural evolutionary process of global intelligent optimization algorithms. Because it’s not limited to the objective function and constraints, it is very suitable for solving a variety of constrained optimization problems. In recent years the use of evolutionary algorithms for solving constrained optimization problems are the concern of many scholars at home and abroad, and have each put forward an improved evolutionary algorithm, to obtain better results.In this dissertation, the paper focuses on the above aspects, in-depth study. The main content of the paper is as follows:Firstly, phased array weather radar system and the format of radar data are researched. Under the counseling and advice of meteorological experts, the nature of the phased array radar data inversion is understood. The radar data inversion model is proposed. The model is a typical constrained optimization problem, with lots of constraints.Secondly, the constrained optimization evolutionary algorithm applied to the phased array radar data inversion problem. Phased array radar data inversion algorithm base on evolution strategy with stochastic ranking is designed. Verify the effectiveness of the algorithm through simulation optimization problems, and then the actual data the algorithm anti-higher performance accuracy of the data, the final results of the experiment are analyzed and discussed;Evolution strategy although can effectively solve the radar data inversion problems, but its calculation process is very time-consuming. According to the original differential evolution algorithm, Phased array radar data inversion algorithm base on fast differential evolution is designed. The algorithm is applied to radar data more effectively. The experiment results are analyzed and discussed, the constraint conditions parameter is researched.Finally, Phased array radar data inversion system is designed and realized. The system can be input and entered the standard radar data, anti-performed more accurate data, the field of meteorology experts can study the effect of inversion to facilitate the design algorithm.This method improves the accuracy of the radar data by pure software optimization methods, without adding additional hardware investment. It does not increase the cost of the phased array weather radar system at the same time with the increasing of the accuracy. Improve the accuracy of the phased array weather radar observations to improve the fine structure of the severe weather, severe weather emergency observation capabilities and to improve the prediction accuracy of weather forecasting and climate models, and so is of great significance.