The research situation and progress of starlight navigation are analyzed. The current situation of many kinds of typical star sensor overseas are collected and presented. At last, the current problem and development trends of starlight navigation are introduced.

Space optical sensor (SOS) as a kind of measurement instrument for spacecraft guidance, navigation and control (GNC) and space surveillance measurements plays an important part in space mission. For roundly understanding of the SOS in this paper a set of outline of the SOS is explained such as the connotation and extension of the concept, classification method, performance trait, design target and method, application, and etc.The technical progress and application of the listed below SOS are summarized. At the end of the paper the development trend of SOS is concluded.

Attitude determination of star tracker is executed by observing star light, which is rather weak, the optics and the imaging array are not applicative to the sun and the moon, the two strong light sources in the space, so a geometrical analytic method is put forward for star tracker to evade strong light sources. First, the sun and moon direction vectors at the observing time are calculated according to the astronomical equations and converted to the local geographic coordinate system. Then, the geometrical relationships among the direction vectors of the sun, the moon the geocenter and the conical FOV of star tracker are divided into 3 classes totally, combined with the exclusive angle parameters of the sunlight and of the moonlight designed for the baffle, the vectorial equations are built for each case to solve the command boresight orientation to direct the following attitude maneuver. Finally, the algorithm verification is carried out using STK (Satellite Tool Kit). The simulation environment is set up by selecting a spell of time and settling the boresight orientation of the star tracker, to conduct a scenario that has more chances for the Sun and the Moon lights interfering into the star tracker FOV or the baffle. After many times of tests, the commanded orientation of the boresight by the algorithm always succeeds in evading the sun, the moon and the horizon plane. This method can conduct protecting of the optics of the star tracker autonomously, selection of star observation window, and can carry out the valid starlight observation by the evasion of the Sun and the moon through attitude maneuver.

In this paper, an analysis of abnormal work mechanism for star trackers while passing through the south atlantic anomaly (SAA) space zone is put forward. And, based on the correlated simulating model, two algorithms of resisting highenergy particle radiation interference are both proposed, namely, real stars detection and alignment between frames. The simulation shows that for star trackers with improved algorithms, the computational cost time and the successful identification rate are greatly higher than those with normal algorithms in capture mode and in track mode, being competent with relatively severe space radiation environment.

In order to further improve star positioning precision of star sensor, the pixel error source and formation mechanism are analyzed.Based on multiframe correlation filtering, a correction method of nonuniformity pixel error applied to star sensor is proposed. Star centroiding field and local correction field are determined according to the position of star. The information of currentframe star map and formerframe local correction filed are utilized to predict the currentframe local correction field. Corrected star centroiding filed can be obtained by difference method. Experimental results show that nonuniformity noise can be predicted and positioning accuracy is improved better by proposed method, which is also with good robustness.

Boresight axis thermal stability indicates the star sensor’s capability to anti thermal change. Ground experiment and numerical simulation are the main approaches to analyze and verify this key performance. The simulation method is the primary one that can be used during the whole design period. An integrated method named thermalstructuraloptical analysis approach is developed to analyze the thermal stability of star sensor. This approach includes thermal, structural, optical analysis and thermal drift calculation. A high accuracy star sensor’s thermal stability is analyzed as an example. The results show the efficiency of this integrated method.

A technical analysis of daytime star detection is given, which is applied to celestial navigation. Firstly, based on the analysis of atmospheric transmittance and correlated sky background scattering as well as the spectral distribution, the detection band from 900~1700nm is ascertained. Secondly, the relationships between the single field of view (FOV) with a stationary optical axis and detection sensitivity and daytime star capture probability are calculated. Finally, the programrelated details for the daytime star detection, the detector selection, integral time, the design of catalogue of stars and multiframe algorithm strategy are given. The experiment results accord with the analysis.

The measurement of the pose of noncooperative targets is one of the very popular research directions in the field of space technology. The sensors include laser sensors and binocular measurement sensors. The precision of threedimension sensors such as lidar is sensitive to the reflectivity of the target surface. The precision of the binocular measurement sensors is limited by stray light and the length of base line. According to the shortage of lidar and the binocular measurement sensors, a computing method of the pose of noncooperative targets by fusing the 2D image and 3D point cloud is presented, in which the advantage of lidar and the binocular measurement sensors are utilized. The feasibility of the method is validated by simulating and experimentation.

A stereo vision based pose measurement method is proposed for noncooperative space targets. Laser radar can provide prior information such as distance and position for binocular vision system. Based on the information, the search area can be narrowed for the image pair because the disparity is determined for a certain distance. The compute complexity can be reduced so that measurement efficiency is improved. Meanwhile, algorithm is more robust because the texture complexity caused by stray light is lower at local area than global image. A closedloop experiment is conducted to validate the performance of proposed method.

In order to solve the problem of high precision distance and angle measurement of lidar with large measurement range, a twostep calibration method of lidar sensor based on calibration field is proposed in this paper. Based on the analysis of the angle measurement error and the distance measurement error of lidar, an error correction model is proposed in this paper. This model separates the distance correction from the angle correction. First, the calibration field is used to complete the angle calibration, and the external parameters are also calculated in this step. Second, the baseline field is used to complete the distance calibration. This method reduces the correlation between the distance calibration parameters and the angle calibration parameters, and guarantees the calibration precision and accuracy. The experimental results show that the method is reasonable and effective, and is able to complete the high precision calibration of lidar.

An adaptive centroid extraction algorithm module is described. This module includes three sub modules, which are adaptive integration time adjustment, smoothness processing of multi wave crests, and extraction of centroid based on gray weighted. This module is realized on FPGA with small delay and little resource occupation. Stable extraction of centroid in the detector imaging system is beneficial to improve the dimensional precision of laser ranging.

Star tracker (ST) is a sensor with high accuracy, which is wildly applied to measure the attitude of spacecraft. The accuracy of a ST can be easily influenced by the outer environment, especially by the temperature environment. Since the heat stability is important for the accuracy of ST, an optimal design method for the baffle based on Bipod structure is presented to improve the heat stability. Through FEA simulation, this design is proved to be effective in intensity and rigidity, and greatly improved the heat stability, and enhances the stability of optical axis directing accuracy while the ST is working in orbit.

The method of space objects recognition with star pattern algorithm is discussed. An allsky map recognition algorithm is carried out based on the research of triangle algorithm and tetrahedral algorithm. Firstly, the algorithm optimizes the method of creating angular distance table in order to save vacuum. Then, the K vector method is used to increase the speed of anglepair matching in star pattern. Tetrahedron algorithm is introduced when the results of triangle recognition are not unique. The projection validation is increased after the triangle recognition in order to overcome error match. Simulation is carried out. The result shows that the algorithm proposed in this paper is effective, which can not only improve the speed of star pattern recognition, but also increase the rate of correct recognition. Thus, the algorithm has great practical value.