Identifying the stars of field of view(FOV) is a difficult stage in determining the attitude through star tracker. Stars identification algorithms identify the stars of FOV, creating patterns from the stars and comparing them with the information saved in the memory. This research presents a new algorithm composed of pyramid, liebe and planer triangle algorithms. It has at least 4 stars as the stars of FOV and forms a triangle pattern. This algorithm is simulated with 16000 stars in MATLAB. Its accuracy, needed memory and searching time are studied and compared with those of pyramid, Liebe and triangular methods. Two errors, centroiding and in fake stars, are considered in the simulation for more approaching towards real status. According to the obtained results, the new algorithm can well identify the FOV among 140 stars in which only four stars are real and the rests are false. Its searching time is 8% lower, comparing to pyramid method, despite their similar accuracies. It compares the areas and polar moments of the four triangles with the information saved in the memory to meet a unique response. Besides its robustness and proper performance, it includes the false stars as well.

Star Tracker Attitude Determination Pyramid Algorithm Star Catalogue Triangle Algorithm Liebe Algorithm Centroiding Error.