基于小生境遗传算法的雷达组网误差配准

多传感器信息融合须进行误差配准。传统的误差配准技术采用RTQC、最小二乘法或极大似然估计法，将非线性方程进行线性化，而线性化过程会引入误差。给出了一种基于小生境遗传算法的误差配准算法，该方法在采用基于ECEF坐标系的误差配准技术的基础上，克服了将非线性方程线性化带来的误差，并在传统遗传算法的基础上引入小生境技术，提高了遗传算法全局寻优能力、收敛速度以及系统误差估计结果的精度。最后，将该方法与基于ECEF坐标系的最小二乘法及传统遗传算法进行了比较，仿真实验结果验证了算法的有效性。

Registration Algorithm of Radar Networking Based on Niche Genetic Algorithm

Registration is a prerequisite process for the information fusion in multisensor system. The traditional error estimation methods, including RTQC, least square method, and maximum likelihood estimation method, involve the linearization of nonlinear equations, which introduces errors to the registration. This paper proposes an error estimation method based on the niche genetic algorithm. In this work, we formulate the multisensor registration problem in the earth-centered earth-fixed (ECEF) coordinate and overcome the disadvantage of linearization. Then, with introduction of niche technology, the global optimization ability and convergence speed of genetic algorithm are improved, and the precision of error estimation is increased. Simulated data are used to verify the effectiveness of the proposed algorithm. Comparisons are made between the least squares (ECEF-LS) registration algorithm and the registration algorithm based on traditional genetic algorithm.