An On-Orbit Measurement Method for Visual Axis Pointing of Remote Sensing Satellite Using the FBG Sensor Network and Machine Learning-Based Cascade Algorithm

On-orbit monitoring of visual axis pointing has always been a challenge in the field of space remote sensing, but it is crucial for correcting the remote sensing data, improving the positioning accuracy, and optimizing the design of key structures, such as a bracket structure of an optical remote sensing camera. In view of that, this article proposes a 3-D measurement method of visual axis pointing based on a fiber Bragg grating (FBG) sensor network and a data processing method that combines the support vector machine regression (SVR), the error backpropagation neural network (BPNN), the inverse finite element (iFEM), and the centroid-angle–normal vector (CANV) algorithm. A carbon fiber-reinforced plastic composite laminate simulating the support surface of a camera installation platform is used as an experimental object. The topological layout of an FBG sensor network is designed using the iFEM algorithm, and an experimental system is constructed. A residual-structured SVR-BPNN cascade model is developed and trained to process the FBG sensing data. The high-precision measurement of the full-field thermal-induced strain and 3-D displacement is realized. On this basis, the visual axis pointing is obtained by the CANV algorithm, and the accuracy is verified. The results show that the proposed method can effectively measure the change in visual axis pointing caused by thermal deformations of a camera installation platform. The proposed method can be used in the field of on-orbit monitoring of visual axis pointing of high-resolution remote sensing satellites.