基于航迹推算的井下人员定位关键算法研究
发布时间：2019-02-25 20:37
【摘要】：随着国家对矿井安全的重视,煤矿井下人员定位技术也得到了快速发展。研究井下人员定位技术,是为了准确的定位井下人员位置。尤其是发生矿难时,精确的定位信息是快速救援的关键。现阶段井下的人员定位系统多为基于基础设施的定位技术,其需要事先安装系统设施,使用成本较高,而且使用范围有限。本文研究开发了一种基于航迹推算的井下人员定位算法,不需借助外界信息,可实现自主定位,使用方便,成本低廉。该算法主要深入研究了计步算法和航向角解算算法。基于陀螺仪的计步算法容易受到行人身体晃动影响,本文设计了基于加速度计的自适应计步算法,可实现参数的自动设定,精确度高。行人航向角解算引入了无人机姿态解算算法—梯度下降算法和基于Mahony的多轴互补滤波算法。前者利用梯度下降的思想,求得误差函数最小值时对应的旋转四元数,并与陀螺仪计算的旋转四元数相融合,解算出航向角;后者利用加速度计、磁力计的长期有效性和陀螺仪短时间内的精确性,相互补偿来更新姿态角度,解算出行人行走方向。同时本文在多轴互补滤波基础上加以改进,使得算法计算量更小,系统稳定性更高。最后,本文在Matlab环境下对梯度下降算法和改进的多轴互补滤波算法进行仿真分析,同时从算法的估算精度和运算时间两个方面来进行对比,得出改进的互补滤波算法更适合井下个人航向角度解算。
[Abstract]:With the national attention to mine safety, underground personnel positioning technology has also been rapidly developed. The research of underground personnel positioning technology is to locate the location of underground personnel accurately. Especially in the case of mine disaster, accurate location information is the key to rapid rescue. At present, most of the underground personnel positioning systems are infrastructure-based positioning technology, which requires pre-installation of system facilities, high use costs, and limited scope of use. This paper studies and develops a downhole personnel location algorithm based on track reckoning. It can realize autonomous positioning without the help of external information. It is convenient to use and low in cost. This algorithm focuses on the step calculation algorithm and heading angle algorithm. The step calculation algorithm based on gyroscope is easy to be affected by the body sloshing of pedestrians. In this paper, an adaptive step calculation algorithm based on accelerometer is designed, which can realize the automatic setting of parameters with high accuracy. The gradient descent algorithm and the multi-axis complementary filtering algorithm based on Mahony are introduced to solve the heading angle of unmanned aerial vehicle (UAV). In the former, the rotation quaternion corresponding to the minimum error function is obtained by using the idea of gradient descent, and combined with the rotation quaternion calculated by the gyroscope, the heading angle is calculated. The latter uses accelerometers, the long-term effectiveness of magnetometers and the accuracy of gyroscopes in a short period of time to update the attitude angle and calculate the walking direction of pedestrians. At the same time, the algorithm is improved on the basis of multi-axis complementary filtering, so that the computational complexity of the algorithm is smaller and the stability of the system is higher. Finally, in the Matlab environment, the gradient descent algorithm and the improved multi-axis complementary filtering algorithm are simulated and analyzed. At the same time, the estimation accuracy and operation time of the algorithm are compared. It is concluded that the improved complementary filtering algorithm is more suitable for the calculation of downhole individual heading angle.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD65
本文编号：2430516
[Abstract]:With the national attention to mine safety, underground personnel positioning technology has also been rapidly developed. The research of underground personnel positioning technology is to locate the location of underground personnel accurately. Especially in the case of mine disaster, accurate location information is the key to rapid rescue. At present, most of the underground personnel positioning systems are infrastructure-based positioning technology, which requires pre-installation of system facilities, high use costs, and limited scope of use. This paper studies and develops a downhole personnel location algorithm based on track reckoning. It can realize autonomous positioning without the help of external information. It is convenient to use and low in cost. This algorithm focuses on the step calculation algorithm and heading angle algorithm. The step calculation algorithm based on gyroscope is easy to be affected by the body sloshing of pedestrians. In this paper, an adaptive step calculation algorithm based on accelerometer is designed, which can realize the automatic setting of parameters with high accuracy. The gradient descent algorithm and the multi-axis complementary filtering algorithm based on Mahony are introduced to solve the heading angle of unmanned aerial vehicle (UAV). In the former, the rotation quaternion corresponding to the minimum error function is obtained by using the idea of gradient descent, and combined with the rotation quaternion calculated by the gyroscope, the heading angle is calculated. The latter uses accelerometers, the long-term effectiveness of magnetometers and the accuracy of gyroscopes in a short period of time to update the attitude angle and calculate the walking direction of pedestrians. At the same time, the algorithm is improved on the basis of multi-axis complementary filtering, so that the computational complexity of the algorithm is smaller and the stability of the system is higher. Finally, in the Matlab environment, the gradient descent algorithm and the improved multi-axis complementary filtering algorithm are simulated and analyzed. At the same time, the estimation accuracy and operation time of the algorithm are compared. It is concluded that the improved complementary filtering algorithm is more suitable for the calculation of downhole individual heading angle.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD65
【引证文献】
相关会议论文 前1条
1 苏丽娜;董金明;赵琦;;基于加速度传感器的计步器系统[A];全国第二届嵌入式技术联合学术会议论文集[C];2007年
，本文编号：2430516
本文链接：https://www.wllwen.com/kejilunwen/kuangye/2430516.html
