煤矿井下基于新息理论的AKF超宽带定位算法

AKF Ultra⁃wideband location algorithm based on innovation theory in underground coal mine

针对传统煤矿井下定位精度低的问题,提出了一种基于新息理论的AKF超宽带定位算法.首先建立基于到达时间(TOA)的超宽带定位模型,其次利用极大似然估计准则,将新息方差引入到卡尔曼滤波中参与计算,对系统噪声进行实时的估计和调整,最终实现了系统的定位.实验结果表明,在选定的坐标系中,X轴方向最大误差为14.26 cm,Y轴方向最大误差为14.01 cm,均方根误差为12.68 cm.本文所提出的算法比传统UWB定位算法得到的定位误差更小,有效的提高了定位精度,能够更好的满足煤矿井下定位精度的需求.

Aiming at the problem of low positioning accuracy in traditional coal mine,we proposed an adaptive Kalman filtering UWB positioning algorithm based on innovation theory.Firstly,a UWB location model based on TOA(time of arrival)was established.Sec⁃ondly,the innovation variance was introduced into the Kalman filter to calculate the system noise using the maximum likelihood estima⁃tion criterion,then the system noise was estimated and adjusted in real time.Finally,the positioning of system was realized.Experi⁃mental results show that the maximum error of X axis direction is 14.26 cm,the maximum error of Y axis direction is 14.01 cm,and the root mean square error is 12.68 cm in the selected coordinate system.Compared with the traditional UWB positioning algorithm,the proposed algorithm has lower positioning error,effectively improves the positioning accuracy,and can meet the positioning accuracy requirements in underground coal mine.