Real-Time Performance Analysis of the UniNa APS-Based Micro Sun Sensor

Abstract

This paper describes the research activity conducted at the University of Naples for the development of a micro-digital sun sensor in view of its flight validation by the end of 2011. The sensor project is financed by the Italian Space Agency in the framework of a program for in-flight validation of new technologies for a future Italian microsatellite standard. The sensor exploits a CMOS photodetector and a multiple-aperture mask to improve precision in the sun-line determination. This particular design of the optical head limits both sensor field of view and measurement updating frequency, due to the need of processing multiple sun spots on the focal plane. Therefore, following a short description of the sensor concept, and prototype, special emphasis is given to the description of the multi-spot processing algorithms, and of the calibration scheme adopted to widen the field of view and guarantee sensor operation in real-time. Specifically, a neural-network calibration scheme is developed which exploits a variable number of sun spots and the average centroid position on the focal plane to output an accurate estimate of the sun line orientation. The effectiveness of the proposed calibration scheme is verified by means of extensive test campaigns carried out with the sensor prototype and the laboratory test facility. Experimental results show that the sensor prototype can provide average accuracy and precision of about 0.06° and 0.001°, respectively, over a field of view of about 80°x70°, with a minimum updating frequency of 1 HZ when 100 sun spots are processed.