Verification of a Real-time Attitude Determination Algorithm Through Development of 48 Channel GPS Attitude Receiver Hardware

J. Jang and C. Kee

Abstract:	This paper verifies a real-time attitude determination algorithm through GPS attitude receiver hardware development. The receiver hardware is based on Zarlink GP2000 series chipsets and S3C2410 microprocessor with ARM920T core, which can support 48 channels to calculate an onboard attitude solution of three degree of freedom. To estimate attitude in real-time, precise relative positions of GPS antenna array have to be determined with GPS L1 carrier phase as fast as possible. Therefore uncertainty of the measurement should be resolved by using the searching technique. However the calculation load based on conventional algorithm is too heavy to perform using RISC microprocessor. Therefore in this paper, cycle ambiguities of each base vectors are resolved using SNUGLAD of which design focus is to reduce the computational load and to allow the receiver to estimate a real-time algorithm which is focused on computational efficiency on multi-base vector system [5], of which number of float point calculation is approximately 15~20 % of LSAST (Least Squares Ambiguity Search Technique, [6]) in the case of the two base vector simulation test. This paper presents experimental verifications with GPS attitude receiver of which firmware is implemented using the new algorithm and includes short introduction about the hardware development. The receiver developed for the experiment is composed of four GP2000 series down converters and correlators [7, 8]. So, 48 channels which support four antennas can be used to estimate 3DOF attitude of vehicle. The experimental result shows that each GP2000 chipset series are successfully synchronized. The receiver could output 10 Hz onboard solutions in real-time with the time synchronized measurements. During the experiment, cycle ambiguities are determined as solution by the stochastical test (the F-distribution test) at the rate of 99~99.7% and attitude errors were approximately 0.2~0.8 deg in one sigma. To keep the cycle ambiguities resolved for a precise solution, cycle slip should be detected and repaired. This reliability issue is important because it is difficult for the rate of attitude determination to reach 100% rates using single epoch data in real-time application [9, 10]. We defined the extended Kalman filter [11] for the cycle slip detection about multi-antennas. For the measurement update, we used carrier phase measurements of each channels. MDE (Minimum Detectable Error) values are calculated for a given probability [12, 13], of which values were about 0.11 m during the real-time static and dynamic experiment. Postprocessing test shows that the filter can detect and repair the slip with the size of one cycle (0.19 cm).
Published in:	Proceedings of the 2006 National Technical Meeting of The Institute of Navigation January 18 - 20, 2006 Hyatt Regency Hotel Monterey, CA
Pages:	236 - 243
Cite this article:	Jang, J., Kee, C., "Verification of a Real-time Attitude Determination Algorithm Through Development of 48 Channel GPS Attitude Receiver Hardware," Proceedings of the 2006 National Technical Meeting of The Institute of Navigation, Monterey, CA, January 2006, pp. 236-243.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In