VR机舱人机交互姿态追踪器的算法设计

Algorithm Design of VR Human Machine Interactive Attitude Tracker

以现有的虚拟现实三维机舱训练系统为基础的实景仿真,在单一的加速度计测量倾角的研究基础上增加陀螺仪角速率测量传感器,完成陀螺仪与加速度计数据融合进行姿态识别的追踪器设计。分析了基于单一的加速度计测量倾角的不足和存在的误差,引入角速率测量的传感器平滑单一加速度计的数据,参考卡尔曼滤波的思想,设计固定权值的滤波数据融合算法。设计通过前一时刻估算状态和当前角速率测量数据推导得出当前时刻的估算状态,选择合适的权值改变加速度计与角速率测量数据的信赖程度,分析算法的适用性和不足,给出解决方案。

In light of the existing simulation system of virtual reality 3 D cabin training system,based on the research of single accelerometer measurement inclination angle,the gyroscope angular rate measurement sensor was added to complete the track gesture recognition design of the gyroscope and accelerometer data fusion. The shortcomings and errors of measuring the inclination with a single accelerometer was analyzed,the use of angular velocity can smooth single accelerometer data. The Kalman filter idea was referenced to design a fixed weight of the filter data fusion algorithm. The design could estimate the current state and the current angular rate measurement data from that of the previous moment,and select the appropriate weight to change the reliability of the accelerometer and angular rate measurement data. The applicability and deficiency of the algorithm were analyzed and the solution was given.