Abstract Objects’ measurements often boil down to the determination of changes due to external factors affecting on their structure. The estimation of changes in a tested object, in addition to proper measuring equipment, requires the use of appropriate measuring methods and experimental data result processing methods. This study presents a statement of results of geometrical measurements of a steel cylinder that constitutes the main structural component of the historical weir Czersko Polskie in Bydgoszcz. In the initial stage, the estimation of reliable changes taking place in the cylinder structure involved the selection of measuring points essential for mapping its geometry. Due to the continuous operation of the weir, the points covered only about one-third of the cylinder area. The set of points allowed us to determine the position of the cylinder axis as well as skews and deformations of the cylinder surface. In the next stage, the use of methods based on artificial neural networks allowed us to predict the changes in the tested object. Artificial neural networks have proved to be useful in determining displacements of building structures, particularly hydro-technical objects. The above-mentioned methods supplement classical measurements that create the opportunity for carrying out additional analyses of changes in a spatial position of such structures. The purpose of the tests is to confirm the suitability of artificial neural networks for predicting displacements of building structures, particularly hydro-technical objects.

中文翻译：

---

经典方法和神经网络在水工物体变形研究中的应用

摘要 物体的测量往往归结为确定由于外部因素影响其结构而引起的变化。对被测物变化的估计，除了需要合适的测量设备外，还需要使用合适的测量方法和实验数据结果处理方法。本研究报告了一个钢圆柱的几何测量结果，该钢圆柱构成了比得哥什历史悠久的 Czersko Polskie 堰的主要结构部件。在初始阶段，对气缸结构中发生的可靠变化的估计涉及选择测量点以绘制其几何形状。由于堰的连续运行，这些点只覆盖了大约三分之一的圆柱体面积。这组点使我们能够确定圆柱轴的位置以及圆柱表面的倾斜和变形。在下一阶段，使用基于人工神经网络的方法使我们能够预测测试对象的变化。人工神经网络已被证明可用于确定建筑结构的位移，尤其是水利技术对象。上述方法补充了经典测量，为对此类结构的空间位置变化进行额外分析创造了机会。测试的目的是确认人工神经网络在预测建筑结构位移方面的适用性，尤其是水利技术对象。使用基于人工神经网络的方法使我们能够预测测试对象的变化。人工神经网络已被证明可用于确定建筑结构的位移，尤其是水利技术对象。上述方法补充了经典测量，为对此类结构的空间位置变化进行额外分析创造了机会。测试的目的是确认人工神经网络在预测建筑结构位移方面的适用性，尤其是水利技术对象。使用基于人工神经网络的方法使我们能够预测测试对象的变化。人工神经网络已被证明可用于确定建筑结构的位移，尤其是水利技术对象。上述方法补充了经典测量，为对此类结构的空间位置变化进行额外分析创造了机会。测试的目的是确认人工神经网络在预测建筑结构位移方面的适用性，尤其是水利技术对象。上述方法补充了经典测量，为对此类结构的空间位置变化进行额外分析创造了机会。测试的目的是确认人工神经网络在预测建筑结构位移方面的适用性，尤其是水利技术对象。上述方法补充了经典测量，为对此类结构的空间位置变化进行额外分析创造了机会。测试的目的是确认人工神经网络在预测建筑结构位移方面的适用性，尤其是水利技术对象。