论文标题
部分可观测时空混沌系统的无模型预测
A method to estimate the size of particles using the open source software OpenPTV
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
A method to obtain particle sizes from images that are used for particle tracking velocimetry is proposed. This is an open source method, developed to use together with the open source software OpenPTV. First, the analysis of different factors that affect the estimation of the particle size is made. Then, a transformation is proposed to estimate the sizes from information obtained in OpenPTV. The method requires an extra calibration with a flat target containing circles of different sizes, but if the calibration is successful, the sizes of particles can be estimated reliable within certain limits. For example, for solid of non-translucent particles, sizes can be estimated with an error of maximum 10 %, as long as the particle diameter correspond to at least three pixels width in the obtained images. For translucent particles some extra assumptions will be needed in the transformation.
