基于Jerk约束的加速度校正算法

Acceleration Correction Algorithm Based on Jerk Constraint

在惯性传感器的姿态估计算法中,陀螺仪在长时间的积分过程中易产生误差积累,而加速度计常被用于辅助校正姿态。然而,加速度计易受干扰而导致测量值不只包含有用的加速度成分。为了求解目标加速度,提出一种基于Jerk(加加速度)约束的无积分误差积累的方法。首先,直接利用陀螺仪和加速度计的输出来提取系统的Jerk信息。其次,在无积分误差积累的Jerk信息约束下,构建加速度与Jerk的关系模型,以此来获得目标加速度。实验表明,所提出的算法能获得较准确的加速度。

In the attitude estimation algorithm of the inertial sensor, the gyroscope is prone to error accumulation during the long-term integration process, and the accelerometer is often used to assist in correcting the attitude.However, the accelerometer is susceptible to interference and the measured value does not only contain useful acceleration components.To solve the target acceleration, a method based on Jerk constraint without integration error accumulation was proposed.First, the output of the gyroscope and accelerometer was used to extract the Jerk information of the system.Secondly, under the constraint of Jerk information without integration error accumulation, the relationship model between acceleration and Jerk was constructed to obtain target acceleration.Experiments show that the proposed algorithm can obtain more accurate acceleration.