Abstract A lot of methods for pyroelectric coefficient measurement have been developed in literatures which, however, are not accurate enough, as a result, there is no standard device for pyroelectric coefficient measurement hitherto. In this work, a dynamic method has been reported and successfully applied back propagation neural network (BPNN) to the control of thermo electric cooler (TEC). It averts poor calculative ability of monolithic computing machine by taking advantage of powerful calculative ability of Matlab on PC and therefore realizes sinusoidal-wave and triangular-wave temperature control with high accuracy. Practical p measurement results for classic ferroelectric materials such as polyvinylidene fluoride (PVDF), lithium tantalate (LT) and lead zirconium titanate (PZT) samples revealed that relative measurement error with sinusoidal-wave temperature control was less than 10%, implying its good accuracy and practical ability.

中文翻译：

---

一种确定铁电材料热电系数的精确动力学方法

摘要 文献中发展了大量的热电系数测量方法，但都不够准确，导致目前还没有标准的热电系数测量装置。在这项工作中，已经报道了一种动态方法，并成功地将反向传播神经网络 (BPNN) 应用于热电冷却器 (TEC) 的控制。它利用Matlab在PC上强大的计算能力，避免了单片机计算能力差的问题，从而实现高精度的正弦波和三角波温度控制。聚偏二氟乙烯 (PVDF) 等经典铁电材料的实际 p 测量结果，