There are many pure phase methods for designing acoustical holography structures in the known literature. Different methods may have different characteristics and lead to different results, but few researchers put these methods together for comparison. So we have no idea which method should be taken when dealing with a specific optimization case. This paper investigated the characteristics of three different PM methods, one is the direct search method taking suitable objective function and another two are the conventional IASA methods but with different restricts in image plane, one called ISREL relaxes the whole background pressure constraints and one called ISRES restricts the whole background pressure to zero. We find that DS is a balanced method in terms of various metrics evaluating target pattern and background noise. Finally, the ability of particle manipulation of DS was verified experimentally by designing a holography structure projecting a circle with a non-uniform phase gradient. In a word, it provides a specific comparison between the two mainstream methods DS and IASA by two optimization cases. On one hand, it gives us a clear view of the characteristics of these methods in designing acoustic holography; On the other hand, we know that DS may be a preferred algoritm for most optimization cases owing to its balanced characteristic regardless of time costs.

在已知文献中，有许多用于设计声全息结构的纯相方法。不同的方法可能具有不同的特性并导致不同的结果，但是很少有研究人员将这些方法放在一起进行比较。因此，我们不知道在处理特定的优化情况时应采用哪种方法。本文研究了三种不同的PM方法的特点，一种是具有适当目标函数的直接搜索方法，另一种是常规的IASA方法，但在图像平面上存在不同的限制，一种称为ISREL放宽了整个背景压力约束，一种称为ISRES将整个背景压力限制为零。我们发现，就评估目标模式和背景噪声的各种指标而言，DS是一种平衡的方法。最后，通过设计全息结构投射出具有不均匀相梯度的圆形的全息图，实验验证了DS的粒子操纵能力。简而言之，它通过两种优化情况对两种主流方法DS和IASA进行了具体比较。一方面，它使我们清楚地了解了这些方法在设计声学全息术中的特点；另一方面，我们知道，由于DS的均衡特性，无论时间成本如何，DS可能是大多数优化情况下的首选算法。它使我们清楚地了解了设计声全息技术时这些方法的特点；另一方面，我们知道，由于DS的均衡特性，无论时间成本如何，DS可能是大多数优化情况下的首选算法。它使我们清楚地了解了设计声全息技术时这些方法的特点；另一方面，我们知道，由于DS的均衡特性，无论时间成本如何，DS可能是大多数优化情况下的首选算法。