【摘要】 随着我国航空业的蓬勃发展，航空发动机技术取得了长足的进步。作为航空发动机的关键组件，航空发动机叶片的研制关乎全局。对其温度进行实时、精确的测量监控有助于推动航空发动机叶片相关技术的突破。本文提出了一种基于红外测温技术的航空发动机叶片温度校准算法和温度场三维重构策略。该工作利用多层感知机网络的函数拟合能力，建立了对目标点精确温度与其周围温度分布的函数关系，并将红外测温的误差控制在1.24℃。在此基础上，采用投影法和三维点云法向量估计的创新策略，实现了二维红外图像到三维空间位置映射。该方法能从单视角红外图像得到温度场三维重构，重构过程简单，摆脱了对多视角图像的依赖。更多还原

【Abstract】 With the vigorous development of the aviation industry in China, significant progress has been made in aeroengine technology. As key components of aeroengines, the development of aeroengine blades is crucial. Real-time and accurate monitoring of the blade temperature will help promote breakthroughs in related technologies for aeroengine blades. This study proposes a temperature calibration algorithm and three-dimensional reconstruction strategy for the temperature field based on infrared temperature measurements for aeroengine blade temperature monitoring. Based on the function fitting ability of the multilayer perceptron network, the functional relationship between the precise temperature of the target point and its surrounding temperature distribution is established, and the error of infrared temperature measurement is controlled within 1.24 °C. On this basis, through an innovative projection method and 3D point cloud normal vector estimation, the position mapping from 2D infrared image to 3D space has been achieved. We achieved a three-dimensional temperature field reconstruction via single-view infrared images through a simple process without dependence on multiview images.更多还原