基于位置的服务作为一种战略新兴产业，已经广泛的进入了人们的生活。随着应急救援、公共安全、购物服务等领域对室内高精度定位的需求越来越迫切，室内定位技术的研究已经成为一个重要课题。已有的室内定位技术存在着定位精度低、系统容量小、容易受室内恶劣环境影响等问题，从而未能推广普及。尤其是在煤矿井下等复杂的环境下，由于不能获取准确的定位信息，救援工作难以开展。

在煤矿井下精确的人员、货物位置信息对于保障煤矿安全生产和矿工人身安全具有重要意义。本文基于超宽带（UWB，Ultra Wide Band）通信的定位技术，以煤矿井下作为实际应用场景，开展有关于室内定位关键技术的研究。本文围绕着基于时间测距的两种定位方式展开研究工作，它们分别是到达时间（TOA，time of arrival）和到达时间差（TDOA，time difference of arrival）。本文的主要研究内容如下：

（1）针对经典的最小二乘算法在非视距（NLOS，non-line-of-sight）环境下定位性能急剧下降的局限，在TOA定位中提出了抑制NLOS误差的改进算法，有效的减小了人体阻挡等NLOS传播的影响，使得在恶劣的NLOS环境达到和视距（LOS，line-of-sight）环境一样的定位精度。

（2）在TDOA定位中提出了基于卡尔曼滤波的高精度时间同步方法，实现了纳秒级别的时间同步精度，并对NLOS误差和时间同步频率具有强鲁棒性，使得TDOA高精度实时定位成为可能。

（3）比较了TOA和TDOA两种定位方式的优劣，结合煤矿井下的环境特点，选择TDOA定位方式，设计了一套由卡片、基站和上位机组成的矿井精确定位系统。

本文完成了非视距误差抑制算法和时间同步算法等关键技术的研究，并基于TDOA定位方式实现了矿井精确定位系统的设计。论文所做的工作为精确定位系统在煤矿井下的实施奠定了坚实的基础。

Recently, Location Based Services (LBS) technologies have been integrated into our daily life as a new strategic industry. Research on indoor localization has become a hot topic with increasing demands for precise localization in the fields of emergency rescue, public security and shopping services. The indoor localization technologies that already exist are easy to be interfered by harsh indoor environments and have low precision and small system capacities. As a result, the spread of these techniques didn’t realize. Especially in coal mines, the rescue work becomes difficult due to the inaccurate location information.

Precise location information of miners and cargos in coal mines is of great significance for the safety of the miners and the safety in production. In the paper, we focus our research on the key techniques of indoor localization for mine application scenarios, which is based on Ultra Wide Band (UWB) technology. The main research contents in this paper include two classical ranging methods in time-based localization: time of arrival (TOA) technique and time difference of arrival (TDOA) technique. The details are as follows:

Firstly, a novel non-line-of-sight (NLOS) identification and mitigation method in TOA location is proposed, which solves the problem that the performance of classical least squares (LS) algorithm severely degrades in NLOS environments. The modified algorithm can attenuate NLOS errors caused by human bodies so that high localization resolution will be realized in NLOS environments.

Secondly, we introduce a clock synchronization method based on the Kalman filter in TDOA location, which can provide a nanosecond level synchronization resolution and has good robustness to NLOS errors and clock synchronization period. Experiments show that the clock synchronization is quite suitable for Real Time Locating Systems (RTLS).

Finally, we compare TOA scheme with TDOA scheme and choose TDOA method for the design of moving target precise positioning system in coal mines, which is composed of tags, anchors and an upper computer.

In this paper, key techniques such as NLOS error attenuation and clock synchronization have been studied and a moving target precise positioning system in coal mines is designed based on TDOA method. All the work in this paper lays the foundation for the implementation of precise positioning system in coal mines.