摘要
以陀螺和加速度计为敏感器件的导航参数解算系统称之为惯性导航系统,通过陀螺的输出建立导航坐标系,而加速度计的输出可以算出运动物体的速度和位置。MPU6050整合了三轴陀螺仪,三轴加速度器,磁力传感器或者其他传感器。它可含有两个I2C端口,可以进行数位运动处理,向运动端输出单一形式的数据流。针对运动传感器应用广泛的现状,提出了一种采用MPU6050模块的运动轨迹检测算法设计方案。轨迹检测的原理是加速度经过二重积分后可以得到位移。由于加速度传感器输出存在积累误差,采用了主成分分析(PCA)方法来抑制积累误差。用PCA进行特征提取,对运动状态下的加速度和速度进行修正,消除积累误差并重建轨迹。传感器的硬件平台搭建主控制器为STM32,通过MATLAB软件绘图功能,利用多组三个方向加速度值和三个方向角速度值来做出运动轨迹和解算姿态。实验结果表明:所述方法使得物体运动轨迹检测的精准度有所提高。
Parameter calculation system is called inertial navigation system by using gyro and accelerometer as sensing devices. The navi- gation coordinate system is established through the output of the gyro. The output of the accelerometer can calculate the speed and position of the moving object. MPU6050 integrates three-axis gyroscope, three-axis accelerometer, magnetic sensor or the other sensors. It can contain two I2C ports and perform digital motion processing. A single form of data stream can be output to the moving end. The applica- tion of motion sensor is wide, and the trajectory detection algorithm of a moving object used by MPU6050 module has been designed in this paper. The principle of trajectory detection is that the acceleration can be obtained by double integral. There is an accumulated error in the output of the accelerometer, so an approach for eliminating accumulated error of accelerometer based on PCA has been presented. After feature extraction has been conducted with PCA, the acceleration and velocity are modified. Elimination of accumulated error is carded out and trajectory is reconstructed. Sensor hardware platform to build the main controller is STM32. Through the drawing function of the MATLAB software, the use of multiple sets of three direction acceleration and three direction angular velocity values is to make the motion trajectory and calculate the attitude. The experimental results show that the method can improve the accuracy of the object motion trajectory detection.
出处
《计算机技术与发展》
2017年第5期179-182,共4页
Computer Technology and Development
基金
国家"863"高技术发展计划项目(2014AA01A705)
关键词
加速度传感器
主成分分析
轨迹检测
误差处理
aecelerometer
principal component analysis
trajectory detection
error handling
