基于位置指纹的煤矿井下定位算法研究

发布时间：2018-12-15 20:14

【摘要】：安全是煤矿生产的第一要务,但煤矿的特殊工作环境导致安全隐患依然存在。煤矿井下人员定位技术能对井下工作人员实时定位,特别是当矿难发生时,能为救援工作提供重要的位置信息,提高了救援效率。基于此,本文在分析现有一些人员定位算法基础上,对基于位置指纹的定位算法进行了研究,并提出了改进的算法。首先,本文分析了几种常用的定位技术以及它们在煤矿井下应用的可行性,得出基于位置指纹的定位技术更加适合在煤矿井下使用。由于位置指纹定位算法是基于信号强度的,本文分析了信号在井下传播的特点,以及信号的多径效应对定位的影响。并指出了指纹定位中的关键问题,为下面的算法研究打下基础。其次,针对矿井巷道狭长的特点,本文引入区域划分的思想,利用井下实验得到的数据确定区域阈值,将被定位区域均匀划分为多个子区域,在进行定位时,根据终端接收的信号强度判断出终端所在的子区域,在子区域内用位置指纹算法对其进行精确定位。最后,通过对K邻近法和贝叶斯匹配算法的优缺点进行分析,提出了一种加权K-Bayes定位匹配方法。此算法先采用加权欧式距离计算出符合条件的K个位置参考点,然后通过设定阈值排除偏差较大参考点的影响,紧接着通过贝叶斯方法计算出剩余参考点处的概率,对结果进行加权处理,并通过实验证明了该算法的可行性。
[Abstract]:Safety is the most important task in coal mine production, but the safety hidden danger still exists due to the special working environment of coal mine. The technology can locate the underground workers in real time, especially when the mine accident occurs, it can provide the important position information for the rescue work, and improve the rescue efficiency. In this paper, based on the analysis of some existing location algorithms, the location algorithm based on location fingerprint is studied, and an improved algorithm is proposed. Firstly, this paper analyzes several common positioning techniques and the feasibility of their application in underground coal mines, and concludes that the location fingerprint based positioning technology is more suitable for use in coal mines. Because the location fingerprint location algorithm is based on the signal intensity, this paper analyzes the characteristics of the signal propagation in the underground and the influence of the multi-path effect of the signal on the location. The key problem of fingerprint location is pointed out, which lays a foundation for the following algorithm research. Secondly, in view of the long and narrow characteristics of mine roadway, this paper introduces the idea of regional division, using the data obtained from underground experiments to determine the regional threshold, and divides the location area into multiple sub-areas evenly. According to the signal strength received by the terminal, the sub-region of the terminal is determined, and the location fingerprint algorithm is used to locate the sub-region accurately. Finally, by analyzing the advantages and disadvantages of K-neighborhood method and Bayesian matching algorithm, a weighted K-Bayes location matching method is proposed. In this algorithm, the weighted Euclidean distance is used to calculate the K position reference points, and then the probability of the remaining reference points is calculated by the Bayesian method by setting a threshold to exclude the influence of large deviation reference points. The results are weighted and the feasibility of the algorithm is proved by experiments.
【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

【参考文献】