GPS attitude measurement with baseline constrained optimization algorithm for unpiloted car

Abstract

Unpiloted technology requires attitude measurement from global positioning system to adjust the moving direction of car. However, the measurement result from the satellites suffer from fluctuation, which reduces the control ability of system. Thus, we propose our novel algorithm. First, it builds double difference observation equation based the measurements from two satellites. Second, the least squares algorithm is utilized to calculate double difference integer ambiguity. Third, it constructs the length of baseline to calculate baseline vector, which improves the solution success rate of integer ambiguity and reduces frequency of mutation occurs in attitude angle solving process. Finally, baseline vector is employed to solve the attitude angle of unpiloted car. The extensive experimental results demonstrate that the success rate of the attitude reach 95% within 2° of error under the constraint of the baseline, and the error of the attitude is less than 1° under the dynamic condition.