School of Electronics and Information, Northwestern Polytechnical University, Xi'an 710072, China.
Shaanxi Key Laboratory of Integrated and Intelligent Navigation, Xi'an 710000, China.
Sensors (Basel). 2018 Nov 2;18(11):3748. doi: 10.3390/s18113748.
The development of smart cities calls for improved accuracy in navigation and positioning services; due to the effects of satellite orbit error, ionospheric error, poor quality of navigation signals and so on, it is difficult for existing navigation technology to achieve further improvements in positioning accuracy. Distributed cooperative positioning technology can further improve the accuracy of navigation and positioning with existing GNSS (Global Navigation Satellite System) systems. However, the measured range error and the positioning error of the cooperative nodes exhibit larger reductions in positioning accuracy. In response to this question, this paper proposed a factor graph-aided distributed cooperative positioning algorithm. It establishes the confidence function of factor graphs theory with the ranging error and the positioning error of the coordinated nodes and then fuses the positioning information of the coordinated nodes by the confidence function. It can avoid the influence of positioning error and ranging error and improve the positioning accuracy of cooperative nodes. In the simulation part, the proposed algorithm is compared with a mainly coordinated positioning algorithm from four aspects: the measured range error, positioning error, convergence speed, and mutation error. The simulation results show that the proposed algorithm leads to a 30⁻60% improvement in positioning accuracy compared with other algorithms under the same measured range error and positioning error. The convergence rate and mutation error elimination times are only 1 / 5 to 1 / 3 of the other algorithms.
智慧城市的发展需要提高导航和定位服务的准确性；由于卫星轨道误差、电离层误差、导航信号质量差等因素的影响，现有导航技术难以进一步提高定位精度。分布式协同定位技术可以在现有的全球导航卫星系统（GNSS）基础上进一步提高导航和定位的精度。然而，协同节点的测量距离误差和定位误差会显著降低定位精度。针对这一问题，本文提出了一种基于因子图的分布式协同定位算法。它利用协同节点的测距误差和定位误差建立因子图理论的置信函数，然后通过置信函数融合协同节点的定位信息。这样可以避免定位误差和测距误差的影响，提高协同节点的定位精度。在仿真部分，从测量距离误差、定位误差、收敛速度和突变误差四个方面对所提出的算法与主要的协同定位算法进行了比较。仿真结果表明，在所测距离误差和定位误差相同的情况下，与其他算法相比，所提出的算法可将定位精度提高 30%至 60%。同时，该算法的收敛速度和突变误差消除次数仅是其他算法的 1/5 到 1/3。
