摘要
针对利用惯性传感器解算姿态易受单种六轴传感器噪声特性和漂移特性影响的问题,提出了一种基于多源IMU组合与粒子滤波优化的互补滤波姿态融合算法。该算法利用多源数据与粒子滤波方法,实现了对姿态解算精度的提升。基于BMI088、ICM20689两种六轴传感器与TMS320F28379S微处理器构建数据采集平台,利用高精度三轴转台驱动数据采集平台进行双轴摇摆实验。基于摇摆实验所得陀螺仪数据和加速度计数据进行改进型互补滤波融合以计算摇摆实验中的姿态角。基于均方误差(MSE)理论利用转台记录得到的三轴实际姿态角曲线对解算所得姿态角曲线做迟滞计算和均方误差计算。结果表明:在设计的改进型姿态解算算法下,横滚角均方误差为0.1596(°)^(2),俯仰角均方误差为0.1498(°)^(2),有效提高了互补滤波的解算精度。
A complementary filtering attitude fusion algorithm based on multi-source IMU combination and particle filter optimization was proposed to address the issue of using inertial sensors to calculate attitude,which was easily affected by the noise and drift characteristics of a single six axis sensor.This algorithm utilized multi-source data and particle filtering methods to improve the accuracy of attitude calculation.A data acquisition platform was constructed based on two types of six axis sensors,BMI088 and ICM20689,and a TMS320F28379S microprocessor.A high-precision three-axis turntable was used to drive the data acquisition platform for biaxial swing experiments.Based on the gyroscope data and accelerometer data obtained from the swing experiment,an improved complementary filtering fusion was performed to calculate the attitude angle in the swing experiment.Based on the mean square error(MSE)theory,hysteresis calculation and mean square error calculation were performed on the calculated attitude angle curve using the three-axis actual attitude angle curve recorded by the turntable.The results show that under the designed improved attitude solving algorithm,the mean square error of roll angle is 0.1596(°)^(2),and the mean square error of pitch angle is 0.1498(°)^(2),effectively improving the solving accuracy of complementary filtering.
作者
宗意凯
苏淑靖
高瑜宏
ZONG Yikai;SU Shujing;GAO Yuhong(North University of China,State Key Laboratory of Dynamic Measurement Technology,Taiyuan 030051,China)
出处
《仪表技术与传感器》
CSCD
北大核心
2023年第8期88-95,共8页
Instrument Technique and Sensor
基金
国家自然科学基金项目(51875534,62075199)。
关键词
MEMS
多源IMU
姿态解算
互补滤波
粒子滤波
MEMS
multi-ssource IMU
attitude solution
complementary filtering
particle filter