一种适用于跑步状态的惯性/零速/GPS室内外无缝组合导航定位方法

目前，行人导航定位技术已经深入社会的众多领域，受到诸多学者的广泛关注。针对行人跑步状态，研究了一种惯性/零速/GPS室内外无缝组合导航定位方法。首先提出了可靠的、适用于行人跑步零速检测的方法，有效提高了在行人跑步状态下的零速检测的准确性。针对GPS信号容易受到高楼、高架等环境的干扰及在室内容易完全丢失的特点，提出了基于BP神经网络的GPS可用信号筛选方法，提高了GPS信息的可靠性与精准性。在此基础上，研究了基于可变量测的Kalman滤波器，实现了惯性/零速/GPS信息的有效融合，显著提高了在行人跑步状态下的导航定位精度。试验结果表明，所提出的这种适用于跑步状态的惯性/零速/GPS室内外无缝组合导航定位方法的平均定位误差可减小到行人跑步总里程的1%以内。

A Method on Indoor-outdoor Seamless Integrated Navigation and Positioning Technology Based on INS/ZUPT/GPS for Running State

Nowadays, the pedestrian navigation and positioning technology has been deeply involved in many fields of society and has attracted wide attention from many scholars. In this paper, a method on indoor-outdoor seamless integrated navigation and positioning technology based on INS/ZUPT/GPS for human running state is studied. Firstly, the zero velocity detection method adopted in this paper is reliable and applicable to effectively improve the accuracy of zero velocity detection for human running state. Considering that the GPS signal is susceptible to environmental disturbance such as buildings, elevated, and even indoor complete missing, the GPS usable signal screening method based on BP neural network is proposed to improve the reliability and precision of GPS. In order to improve the accuracy of pedestrian navigation and positioning remarkably, an integrated navigation system which fuses inertial navigation, zero velocity information and GPS based on Kalman filter is studied. The experiment results show that the average positioning error of pedestrian running indoor-outdoor seamless integrated navigation and positioning technology based on INS/ZUPT/GPS can be reduced to less than 1% of the total mileage.