一种基于图像的燃气轮机叶型参数测量方法

An Image-Based Turbine Blade Parameter Inspection Method

摘要: 叶型参数的准确测量是实现叶片性能诊断和逆向建模的关键步骤。为了提高叶型参数的测量效率，提出了一种基于图像的参数提取方法。首先，采用骨架提取算法对叶片中弧线进行定位；然后，通过拟合计算获得中弧线函数及叶型厚度分布；最后，采用优化算法提高叶型参数辨识精度。实验结果表明，所提出的基于图像的叶型测量方法具有较高的准确性，测量相对误差小于1.5%，且应用方便、灵活，为叶片几何参数的快速、准确测量提供了新的解决方法。

Abstract: Accurate inspection of the feature parameters of the blade is a pivotal step to conduct the performance diagnosis and reverse modeling. To improve the inspection efficiency, an image-based parameter inspection method was proposed. First, the skeleton extraction algorithm was used to locate the middle arc of the blade. Then, the middle arc function and the blade thickness distribution function were obtained by the 3-order polynomial curve fitting algorithms.Finally, an optimization algorithm was derived to improve the inspection accuracy. Experimental results show that the proposed image-based feature parameter inspection method has high accuracy, and the relative measurement error is less than 1.5%. It provides a new solution for the rapid and accurate measurement of blade geometric parameters.