本文提出了两种方法分别可以获得更好的初始猜测和更好的利用传感器测量结果。第一，为了获得更好的初始猜测，利用目标被传感器阵列采集到的磁场数据的大小，可以大体估算出目标在传感器区域内的哪一个子区间，用子区间的中心点作为优化算法的初始猜测开始计算。第二，为每一个传感器增设一个权重，距离目标近的传感器更重要，权重高；距离目标远的传感器测量中的噪音比例高，所以更不重要，权重低。
用于无线胶囊内窥镜的一个新型被动磁定位可穿戴系统
a novel passive magnetic localization wearable system for wireless capsule endoscopy [1]
Paper Link
Authors: Guoliang Shao, etc.
2019，IEEE Sensors Journal
然而，磁定位受苦于来自磁噪音的偏移效应和在优化过程中的不确定初始猜测。在这篇文章中，一个带有噪音消除的定位系统被提出来减少地磁噪音的飘移影响并且以一个简单和有效的方式来获取初始猜测。
However, the magnetic localization suffers from the drifting effect from magnetic noise and the uncertain initial guess during the optimization. In this paper, a localization system with noise cancellation is proposed to reduce the drifting effect of the geomagnetic noise and obtain the initial guess in a simple and effective way.
两个相关的传感器组的测量方差被展示为点线。关系能被看见当距离增加，测量方差缩小。更进一步的，当目标位于三轴的中点时，两个相关的组的测量方差相等。在所提出的基于方差的算法中，定位区域被分为12个子区域在Z轴有3个部分X轴2个Y轴2个。每个子区域的中心在优化过程中被选择为初始猜测。
The measurement variance of two relative sensor groups is plotted as the dotted line. The relationship can be seen that as the distance increases, the measurement variance decreases. Moreover, when the target is placed in the mid-point along three axes, the measurement variances of two variance groups equal. In the proposed variance-based algorithm, the localization region is divided into 12 sub-regions with 3 parts along the Z axis, 2 parts along the X axis, and 2 parts along the Y axis. The center of each sub-region is selected as the initial guess during the optimization process.
距离部分地反映了每个传感器的信噪比测量。当距离小的时候，目标磁铁的磁密贡献测量的主要部分，这给了一个高的信噪比。相反的，大的距离，相关的传感器将给一个低的信噪比测量。传感器权重通过估计的距离被确定。
The distance partially reflects the SNR measurement of each sensor. When the distance is small, the magnetic density from the target magnet contributes the major part of the measurement, which gives a high SNR. Conversely, with a large distance, the corresponding sensor will give a low SNR. The sensor weight is determined by the estimated distance.
[1]: Shao, Guoliang, et al. “A Novel Passive Magnetic Localization Wearable System for Wireless Capsule Endoscopy.” IEEE Sensors Journal 19.9 (2019): 3462-3472.