Radar Signal Tracking Filter Design Based On Evolutionary Algorithm

With the gradual development of digital circuit technology, the overall performance of radar has been improved substantially. The early radar could simply show the target’s information on the screen. In order to know the target’s state of movement, we have to plot and calculate manually to determine the target’s information such as direction and speed of its movement. And with today’s radar system, we can perform operations like filtering, tracking and forecasting on the target. Among these mentioned above, tracking filter is an important part of the radar system.As the target maneuver performance becomes more and more strong, the filter needs corresponding filter performance with adapted precision and better stability. Evolutionary algorithm is a self-adapting global optimization search algorithm which came into being by simulating inheritance and evolution procedure of biology in natural environment. It has a very strong nature of intelligence and robustness. Therefore, it has been applied to optimization areas more frequently. And the designing problems of radar target tracking filter actually all boils down to optimization problem.In this paper the author uses evolutionary algorithm to optimize the radar signal tracking filter design and realize the design of the filter through the Matlab software simulation based on the research of radar signal tracking filter. This paper’s innovation points are as follows:First, the paper combines the optimum solution search characteristics of evolutionary algorithm with the solving of radar signal processing and proposes a new algorithm for the design of the signal filter based on the evolutionary algorithm. In accordance with the implementations of evolutionary algorithm, the algorithm completes the solution for the filter design by evolution operations such as initialization, fitness calculation, selection and crossover variation, etc. It has very high solving precision and can overcome the disadvantage of traditional mathematical algorithm which has great calculating error.Second, the paper analysis of the Kalman filter model and basic algorithm based on the establishment of the state equation and measurement equation on line tracking filter. And proposes the new algorithm of evolutionary algorithms based on the designed ideologically of Kalman filter.It design of linear filters from a two-dimensional and three-dimensional perspective, and through experimental simulation to comparison and analysis of its advantages with Kalman filter algorithm.Third, the paper proposes the tracking filter algorithm of evolutionary algorithms based on nonlinear tracking filter commonly used models and on the principle of nonlinear filtering algorithm, which can’t limited by model and noise. It improved and optimized on the basis of particle algorithm, so that the crossover probability and mutation probability according to the actual situation of the target in the evolution process. The different stages of evolution to take different parameters, randomly adjust the parameter values. Comparative analysis of experimental simulation and particle algorithm can be drawn from the merits of experimental simulation and particle algorithm can be drawn from the merits of the particle performance of the algorithm adaptive tracking filter based on evolutionary algorithms.