TY - JOUR
T1 - Improving estimation of satellite attitude when it changes rapidly
AU - Liu, Yingying
AU - Zhou, Jun
PY - 2008/4
Y1 - 2008/4
N2 - Typical algorithms for estimating satellite attitude employ fixed step-length. For rapidly changing attitude about 5 × 10-3 (°/s) or faster, fixed step-length algorithms provide unsatisfactory precision. We now present an estimation algorithm that can provide satisfactory precision. In the full paper, we explain our estimation algorithm in some detail; in this abstract, we just add some pertinent remarks to listing the three topics of explanation. The first topic is: The attitude sensor measurement model. In this topic, we discuss the gyro measurement model and star sensor measurement model. The second topic is: The working principles of attitude estimator. Its three subtopics are: State vector and state equations (subtopic 2.1), star sensor measurement equations and their linearization (subtopic 2.2) and the model of satellite attitude estimator (subtopic 2.3). The third topic is, the estimation algorithm considering multi-output frequencies. Its three subtopics are: Forecast calculation (subtopic 3.1), updating calculation (subtopic 3.2), and updating the attitude quaternion and gyro drift (subtopic 3.3). Finally we simulate the attitude estimation algorithm, and the simulation results, shown in Figs. 1 and 2 in the full paper, indicate preliminarily that: (1) The error variance of estimated attitude is 1.20 × 10-4 (°) when the angular velocity of a satellite is 1 × 10-4 (°/s), the same as that of the typical fixed step-length algorithm; (2) The error variance of estimated attitude is 1.51 × 10-4 (°) when the angular velocity a satellite is 5 × 10-3 (°/s) but that of fixed step-length algorithm is 1.10 × 10-3 (°).
AB - Typical algorithms for estimating satellite attitude employ fixed step-length. For rapidly changing attitude about 5 × 10-3 (°/s) or faster, fixed step-length algorithms provide unsatisfactory precision. We now present an estimation algorithm that can provide satisfactory precision. In the full paper, we explain our estimation algorithm in some detail; in this abstract, we just add some pertinent remarks to listing the three topics of explanation. The first topic is: The attitude sensor measurement model. In this topic, we discuss the gyro measurement model and star sensor measurement model. The second topic is: The working principles of attitude estimator. Its three subtopics are: State vector and state equations (subtopic 2.1), star sensor measurement equations and their linearization (subtopic 2.2) and the model of satellite attitude estimator (subtopic 2.3). The third topic is, the estimation algorithm considering multi-output frequencies. Its three subtopics are: Forecast calculation (subtopic 3.1), updating calculation (subtopic 3.2), and updating the attitude quaternion and gyro drift (subtopic 3.3). Finally we simulate the attitude estimation algorithm, and the simulation results, shown in Figs. 1 and 2 in the full paper, indicate preliminarily that: (1) The error variance of estimated attitude is 1.20 × 10-4 (°) when the angular velocity of a satellite is 1 × 10-4 (°/s), the same as that of the typical fixed step-length algorithm; (2) The error variance of estimated attitude is 1.51 × 10-4 (°) when the angular velocity a satellite is 5 × 10-3 (°/s) but that of fixed step-length algorithm is 1.10 × 10-3 (°).
KW - Attitude estimation
KW - Computer simulation
KW - Gyro
KW - Multi-output frequency
KW - Satellites
KW - Star sensor
UR - http://www.scopus.com/inward/record.url?scp=44849114904&partnerID=8YFLogxK
M3 - 文章
AN - SCOPUS:44849114904
SN - 1000-2758
VL - 26
SP - 189
EP - 193
JO - Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University
JF - Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University
IS - 2
ER -