The working modes of star sensors generally include four modes, namely self check mode, all sky sphere recognition mode, tracking mode, and imaging working mode. Among them, all sky sphere recognition mode and tracking mode are the two main working modes of star sensors in orbit operation. When in the global recognition mode, the star sensor matches and recognizes the observed stars in the captured star map within the global range to obtain the initial pose of the star sensor; In tracking mode, the star sensor matches the observed star in the local sky region based on the prior pose information of the previous frame or frames, achieving the tracking of stars in the field of view. The three-axis attitude data at this time is calculated from the coordinate system of the star sensor where the successfully tracked navigation star is located and its corresponding celestial coordinate system vector. When the star sensor is first turned on or restarted, it first enters the all sky recognition mode to capture the initial pose without prior attitude conditions. After the initial pose data is successfully obtained, the star sensor switches to tracking mode. In some cases, if star tracking fails in the field of view and pose information is lost, the star sensor will once again switch to all celestial mode. The two working modes switch between each other, as shown in Figure 2.5.
The relationship between two main operation modes:
The data processing link of the star sensor is shown in Figure 2.6, where star point positioning, star map recognition, and attitude calculation are the three key technical points that affect the performance of the star sensor.
The data processing link of the star sensor is shown in Figure 2.6, where star point positioning, star map recognition, and attitude calculation are the three key technical points that affect the performance of the star sensor.
(1) Star point extraction
Star point extraction is the process of calculating the coordinates of the star sensor image coordinate system where the target star point imaging position is located from the captured star map. The measurement errors in this process directly affect the success rate of subsequent star map recognition and attitude calculation accuracy. In star point extraction, subpixel interpolation is usually used to locate the star point position, in order to overcome the positioning accuracy limitations caused by the pixel size of the detector. The so-called subpixel interpolation technique refers to the scattered distribution of starlight energy on multiple detector pixels, and the use of star signal received by multiple pixels for interpolation calculation to achieve sub pixel level star positioning accuracy. At present, subpixel interpolation technology has been widely applied in star sensor engineering and is also one of the key research contents of this article.
(2) Star Map Recognition
According to the working principle of star sensors, identifying and matching star maps captured by image sensors is a key link for star sensors to achieve attitude measurement functions. The correctness of the recognition results directly affects the subsequent calculation of attitude data. In addition, the robustness and speed of star map recognition also affect the reliability and frequency of star sensors.
The essence of star map recognition is to find the navigation star corresponding to the observed star in the celestial coordinate system in the captured star map. Figure 2.7 shows a schematic diagram of star map recognition. Star sensors usually work in both the all sky mode and tracking mode states. Due to the lack of prior attitude data, the former requires star map matching on the captured star map within the entire sky sphere, and requires a high recognition rate; The latter performs star map recognition on the premise that prior pose information is known, and the matching range is reduced from the entire celestial sphere to a local celestial region, thus reducing the difficulty of recognition relatively much. In addition, due to factors such as star point position, magnitude measurement errors, and pseudo star interference in practical applications, there is a deviation between the captured star map and the standard star catalog database, resulting in failed or erroneous star map recognition, and the star sensor cannot output correct attitude data. Therefore, the autonomous star map recognition algorithm in all sky sphere mode is a key technology in the field of star map recognition, and it is also the focus of this article’s research.
(3) Pose calculation
The purpose of attitude calculation is to calculate the attitude conversion matrix, whose accuracy directly determines the three-axis attitude output accuracy of the star sensor, and is also one of the key technologies in the data processing link. Generally speaking, attitude solving algorithms require at least two correctly matched observation stars in the star map. For example, the TRIAD algorithm [175176] only uses the information of two observation stars for attitude solving, while the Quaternion Estimation algorithm [177] (QUEST) requires three or more observation stars to solve the attitude overdetermination problem, but its solution accuracy is better than the TRIAD algorithm. At present, attitude calculation methods are relatively mature and generally do not introduce errors.
Data process of the star sensor
Star identification
Send us a message,we will answer your email shortly!
TY-Space Technology (Beijing) Ltd. is professional focusing on advanced attitude optical sensors, especially Star Trackers for space industry.
