CAJ | 학술논문

针对标准ERTS平滑算法在位置和姿态估计中计算复杂、效率低、精度不高等问题，提出了利用奇异值分解法改进ERTS平滑算法优化位置和姿态数据的新方法。对系统采集到的位置和姿态信息进行前向扩展卡尔曼滤波，降低系统噪声的初步影响；对滤波后的均方误差阵进行奇异值分解，并降低后向递推增益和预测值计算量，提高了预测精度，有效增强了系统的抗干扰性和稳定性。Turtlebot移动机器人平台的试验效果证明该算法在位置和姿态估计中的高效性和稳定性。
침대표준ERTS평활산법재위치화자태고계중계산복잡、효솔저、정도불고등문제，제출료이용기이치분해법개진ERTS평활산법우화위치화자태수거적신방법。대계통채집도적위치화자태신식진행전향확전잡이만려파，강저계통조성적초보영향；대려파후적균방오차진진행기이치분해，병강저후향체추증익화예측치계산량，제고료예측정도，유효증강료계통적항간우성화은정성。Turtlebot이동궤기인평태적시험효과증명해산법재위치화자태고계중적고효성화은정성。
To overcome the disadvantages of standard ERTS smoothing algorithm in position and attitude estimation, e. g. , complexity, low efficiency, and poor precision, etc. , the new improved ERTS smoothing algorithm by adopting singularity valve decomposition is proposed for optimizing position and attitude data. After forward extended Karman filtering for the information of position and attitude collected in the system, the initial impact ofthe system noise is reduced; the singularity value decomposition is conducted for the MSE matrix after filtering, thus the backward recursion gain and the calculated amount of the predicted valueare decreased, and the prediction accuracy is improved;as well as the anti-interference and stability of the system are effectively strengthened. The experimental result on Turtlebot mobile robot platform verifies the high effectiveness and stability of this algorithm in position and attitude estimation.