Multi-sensor-assisted measurement and calibration of alignment deviations in spatial reconstruction of ultra-large-aperture optical systems.

Abstract

To meet the measurement requirements for the precise assembly of support trusses during the spatial reconstruction of ultra-large-aperture optical systems, this paper presents a multi-sensor-assisted alignment deviation measurement system and a suitable global calibration method. By integrating multi-source data from dual visual cameras, a biaxial inclinometer, and laser rangefinders, the system represents a unified measurement network, thereby overcoming the limitations of monocular vision systems in scenarios with sparse targets, restricted fields of view, and environmental disturbances. The paper describes the modeling of the measurement system and the calibration of the sensors. By defining coordinate frameworks and leveraging the respective transformation relationships, a measurement model for optimal truss alignment is developed. The systematic calibration approach can be applied in cases in which the system has unknown parameters, including camera focal lengths, laser ranging data, relative poses of dual cameras, and the relationship between the cameras and the alignment coordinate frame. Subsequently, the calibrated system parameters are integrated into the measurement model to quantify truss-alignment deviations. Experimental measurements confirm both the effectiveness of the developed multi-sensor measurement framework and the accuracy of the calibration parameters. Therefore, this study provides a feasible measurement and calibration solution for truss assembly in the spatial reconstruction of extremely large-aperture optical systems.