论文部分内容阅读
针对地磁匹配中经常会出现相似点,造成定位偏差较大的问题,该文提出利用智能手机识别用户室内行为方式的方法,为地磁匹配算法提供筛选条件。开发了智能手机传感器数据采集工具,获取用户在室内环境下的行为数据。原始数据首先利用一阶低通滤波和平滑滤波算法进行去噪处理,再经过数据分割和特征提取后,应用于行为识别过程。行为识别模型的建立主要使用两种方法,K最近邻算法和隐式马尔可夫模型,并研究了两种方法的不足以及改进途径。通过针对识别准确度的对比实验,在输入最合适的数据的条件下,隐式马尔可夫模型的准确度略优于K最近邻算法。两种方法的识别准确率均在95%以上,能够有效地提高地磁定位精度。利用室内用户行为数据辅助地磁室内定位,很好地改善了地磁数据单一、定位精度较低的问题。
Aiming at the problem that geomagnetic match often appears similar points, resulting in large positioning deviation, this paper proposes a method of using smartphone to identify user indoor behavior, and provides screening conditions for geomagnetic matching algorithm. Developed a smart phone sensor data acquisition tool to obtain the user’s behavior data in the indoor environment. The original data is first denoised by first-order low-pass filtering and smoothing filtering algorithm, and then applied to the behavior identification process after data segmentation and feature extraction. The establishment of behavior recognition model mainly uses two methods, K nearest neighbor algorithm and Hidden Markov Model, and studies the shortcomings of the two methods and ways to improve. By contrast experiments for recognition accuracy, the accuracy of Hidden Markov Model is slightly better than that of K nearest neighbor algorithm when the most suitable data is input. The recognition accuracy of the two methods is more than 95%, which can effectively improve the accuracy of geomagnetic localization. Utilizing the indoor user behavior data to assist the geomagnetic indoor positioning, the problem of single geomagnetic data and low positioning accuracy is well improved.