The measurement value of the zinc coating weight (mass) of continuous hot-dip galvanizing line often deviates from that of gravimetric method. We employed 17 samples on-line and off-line comparative test, and revealed the deviation is inseparable from the measured value of iron content. In view of influence of the iron content on the phase distribution of the zinc-iron alloy layer, using iron content to make up for the shortcomings of X-ray fluorescence (XRF) fundamental parameter method, the accurate value of zinc coating weight were estimated by dynamic fuzzy neural networks. The results show that the average deviation of the estimated zinc coating weight is reduced from 3.07 g/m² of XRF to 1.16 g/m², with an iron content ranging from 5 to 15 %. This estimator is suitable for continuous production lines with fast speed and high control precision.

连续热浸镀锌线的镀锌重量（质量）的测量值经常与重量法不同。我们使用了17个样品的在线和离线对比测试，发现偏差与铁含量的测量值密不可分。鉴于铁含量对锌铁合金层相分布的影响，利用铁含量弥补了X射线荧光（XRF）基本参数法的不足，估算了镀锌层重量的准确值通过动态模糊神经网络。结果表明，估计的锌涂层重量的平均偏差从XRF的3.07 g / m ²降低到1.16 g / m ²，铁含量为5％至15％。该估算器适用于速度快，控制精度高的连续生产线。