论文部分内容阅读
基于四元数微分方程,采用四阶龙格-库塔法处理输出数据,建立了递推关系式,并以四元数作为系统的状态矢量,构成一种四维Kalman滤波器。最后,在DM3730 Cortex-A8处理器平台上对算法进行了实现和实验验证。实验结果表明,经过降维处理后的算法,显著降低了算法的实现复杂度和计算量,滤波效果及实时性能均达到了预期的目标,具有较强的实用性。
Based on the quaternion differential equation, the fourth order Runge-Kutta method is used to process the output data, and the recursive relation is established. The quaternion is used as the state vector of the system to form a four-dimensional Kalman filter. Finally, the algorithm is implemented and experimentally verified on the DM3730 Cortex-A8 processor platform. The experimental results show that the algorithm after dimensionality reduction significantly reduces the complexity and computational complexity of the algorithm, and achieves the expected goal of filtering and real-time performance, which is of strong practicality.