一种基于遗传算法的自适应时延估计

An Adaptive Time Delay Estimator Based on Genetic Algorithm

基于四阶累积量的约束自适应时延估计（FOC—ETDE）算法具有优良的时延估计性能，但它需要较好的时延估计初值。为了克服此缺点，文章引入遗传算法进行时延估计的寻优，无需时延的先验信息，在低信噪比的情况下可以准确地直接估计非整数倍采样间隔的时延。计算机仿真实验验证了新方法的有效性。

The adaptive algorithm for explicit time delay estimation based on fourth-order cumulants （FOC-ETDE） has a good performance of time delay estimation, but it must have a good initial value. To overcome this disadvantage, genetic algorithm is employed for optimizing search process of time delay estimation. Without the initial information of time delay, this algorithm can achieve time delay which is the nonintegral sample period accurately and directly at a low signal-to-noise ratio environment. The simulation examples are presented to demonstrate the effectiveness of this new method.