Better understanding of the loss mechanisms and higher confidence in material data for piezoelectric materials are very important for transducer manufacturers. In this paper, a method is described to characterize these loss mechanisms using a global optimization algorithm, a 1-D equivalent circuit, and a simple experimental measurement. Two different cost functions were used in the optimization algorithm, one based on impedance and the other based on admittance. The sensitivities of these cost functions to the parameters being characterized are shown to be nonuniform. The results are compared to the results obtained from a local optimization method to show the advantage of using a global optimization algorithm. A way to quantify the uncertainty of the results is introduced by looking at the difference between the results obtained from the two different cost functions. It is shown here that the more ambient noise there is in the data, the wider the gap between the results found by the two cost functions. However, although the results from the individual cost functions diverge from the correct values with increasing noise, the average of the two cost function results remains within $5\%$ of the original value. Hence, even with noise in the measured data, the use of two cost functions can yield accurate material parameters.

中文翻译：

---

通过全局优化算法表征压电材料参数

对于换能器制造商而言，更好地了解损耗机制并提高对压电材料材料数据的信心非常重要。在本文中，描述了一种使用全局优化算法、一维等效电路和简单的实验测量来表征这些损耗机制的方法。优化算法中使用了两种不同的成本函数，一种基于阻抗，另一种基于导纳。这些成本函数对被表征的参数的敏感性被证明是不均匀的。将结果与从局部优化方法获得的结果进行比较，以显示使用全局优化算法的优势。通过查看从两个不同成本函数获得的结果之间的差异，引入了一种量化结果不确定性的方法。此处显示数据中的环境噪声越多，两个成本函数发现的结果之间的差距越大。然而，尽管单个成本函数的结果随着噪声的增加而偏离正确值，但两个成本函数结果的平均值仍保持在原始值的 $5\%$ 以内。因此，即使测量数据中有噪声，使用两个成本函数也可以产生准确的材料参数。尽管单个成本函数的结果随着噪声的增加而偏离正确值，但两个成本函数结果的平均值仍保持在原始值的 $5\%$ 以内。因此，即使测量数据中有噪声，使用两个成本函数也可以产生准确的材料参数。尽管单个成本函数的结果随着噪声的增加而偏离正确值，但两个成本函数结果的平均值仍保持在原始值的 $5\%$ 以内。因此，即使测量数据中有噪声，使用两个成本函数也可以产生准确的材料参数。