【作者】 张鑫；

【导师】 来智勇；

【作者基本信息】 西北农林科技大学 ， 计算机系统结构， 2015， 硕士

【摘要】 本文旨在研发肉牛产肉量活体预测装置中的位置姿态测量模块,测定预测装置的位置与姿态,为构造牛体三维模型、进而预测活牛产肉量做准备。该模块依据惯性导航原理,采用ZigBee无线传输技术,通过采集的加速度和角速度数据计算预测装置的位置与姿态,主要工作如下:(1)设计实现位置姿态测量模块。依据肉牛产肉量预测装置的实际需求,从方便使用、低成本出发进行模块设计,采用MEMS加速度计和陀螺及ZigBee模块等构建该模块。ZigBee模块选用CC2530控制加速度计和陀螺进行数据的采集,并完成数据的无线收发,将数据传送到PC机上进行处理。(2)建立加速度计和陀螺的误差模型。分析加速度计和陀螺的误差成分,构建以零偏、刻度因子、交叉耦合项和随机误差为主的、含有24个待定参数的加速度计误差模型和陀螺误差模型。分别使用12位置法和角位置法,对加速度计误差模型和陀螺误差模型中的参数进行标定,导出误差模型中的待定参数,得到误差模型的数学表达式。采用Kalman滤波算法对随机误差进行处理,降低加速度计和陀螺的随机误差。实验表明,加速度计的测量误差方差相对于使用误差模型前的降低了84.78%;陀螺的测量误差相对于使用误差模型前的降低了88.25%。(3)解算载体的姿态和位置。根据采集到的加速度和角速度,采用四元数法先对姿态进行解算。使用四阶龙格库塔算法对四元数进行更新,以解算出的姿态信息为测量值,以采集的加速度数据为观测值,使用UKF滤波算法进行数据融合,降低姿态解算过程中由陀螺漂移引起的误差,300秒静态姿态解算偏差在-0.2o~+0.2o以内,动态姿态解算偏差在-2o~+2o以内。推导位置姿态测量模块适用的比力方程,导出速度和位置的解算公式,位置解算长度为1m时,X轴最大误差为0.0216m,Y轴最大误差为0.0959m,Z轴最大误差为-0.0289m。更多还原

【Abstract】 This research aims to develop a module used for measuring the position and rotation attitude of the living body predictive device for beef meat yield, which can make preparation for building three-dimensional beef body model and predicting living beef meat yield. According to the principle of inertial navigation, this module calculates the predictive device’s position and rotation attitude using the acceleration velocity and angular velocity data collected by ZigBee wireless data transmission technique. The main work of this paper is as follows:(1) Designing and implementation of the position and attitude measurement module. For the purpose of convenience and low cost, MEMS accelerometer, gyroscope and ZigBee are used to compose this module. According to the real requirements for the beef meat yield prediction device, this paper designs the hardware and software for the position and attitude measurement module. The acceleration velocity data and angular velocity data are collected by the ZigBee module, which controls the accelerometer and gyroscope with its main chip CC2530. The collected is then sent to PC to be processed.(2) Building of error model for accelerometer and gyroscope. By analyzing the error components of accelerometer and gyroscope, this study builds the error model for accelerometer and gyroscope, which contains 24 undetermined parameters with the main parameters being zero bias, scale factor, cross coupling and random error. The12-position method and the angular position method are respectively used to calibrate the parameters in the error model for accelerometer and the error model for gyroscope. Then devided the undetermined parameters of error models and worked out the matematical expressions of error models. In above process, the Kalman filtering is also used to deal with random error so as to reduce the random error of accelerometer and gyroscope. The experimental results of the testing for error models show that, compared with the variance of the accelerometer without error model, the variance of the accelerometer with error model reduces by 84.78%; compared with the measuring error of the gyroscope without error model, the measuring error of the gyroscope with error model reduces by 88.25%.(3) The calculation of carriers’ attitude and position. In this part, quaternions are adopted for calculating the carriers’ attitude with using collected acceleration velocity and angular velocity data and 4th order Runge-Kutta algorithm is used to update quaternions. The author takes carriers’ attitude information calculated by angular velocity as measuring data, takes the accelerometer data as observing data and fuses these two kinds of data with UKF filtering algorithm. The error caused by gyroscope drift in the process of attitude calculation is reduced and the precision of attitude calculation is improved because of using UKF filtering algorithm. The results of the attitude calculation experiments show that the static deviation in 300 seconds is within-0.2o~+0.2o and the dynamic deviation is within-2o~+2o. This study also derives the force ratio equation suitable for the position and attitude measurement module and works out the calculation method for speed and attitude. The maximum errors of position calculation along X- axis, Y- axis and Z- axis are 0.0216 m, 0.0959 m and-0.0289 m respectively when the position and attitude measurement module is carried out along a 1-meter long segment.更多还原