超声回波的参数矢量含有被测目标丰富的信息,如反射体的几何形状、大小、方位以及传输通道的系统特性信息,因此准确的估计回波模式对于确定目标和传输通道的特性是十分必要的。针对以往参数估计都是基于盲信息搜索的,没有充分利用超声回波的先验知识,计算量大,优化容易陷入局部极小点。针对上述问题,提出一种基于初始参数信息的能级遗传算法来实现超声回波参数的估计,最后应用到应力检测,结果是令人满意的。
The parameter set of ultrasonic echo has not only information of detected target or crack such as geometrical shape, size and orientation, but also has information of the system response characteristic of propagation path. Therefore accurate echo estimation is most important to acquire the characteristic of target and propagation path. Traditional parameter estimation methods are all based on blind information search, ignoring prior information, so the results have large calculations and trap in local extremum point. In this paper a new method of energy level-genetic optimization based on prior parameter information to accomplish echo parameter estimation is proposed, and a satisfied result is shown when it is applied to stress test.