基于 UWB/IMU 融合的室内定位与导航技术研究：Abstract
III
Abstract
Nowadays, indoor positioning technology have become more and more popular
in areas such as sweeping robots, service robots, unmanned (intelligent parking),
VR/AR, pedestrian tracking and unmanned aerial vehicles. Highly accurate location
technology can provide moving target for pose used to explore unknown
environments and complete the automated track or planning. Therefore, studying
issues about indoor high-precision location is so important.
The challenge of indoor location is that single-method positioning system cannot
meet requirements of actual application. The reason maybe is that not enough location
accuracy, or limited application scenarios, or lack of stability, or the high expense and
so on. In order to solve the shortcomings, the paper carries out research on key
technologies of location information fusion combining UWB (Ultra Wideband)
ranging with inertial navigation system and also gives some ideas for the
implementation of engineering system. To realize multi-sensor fusion, we firstly
analyzes the characteristics and inherent defects of single-method positioning system,
and then designs the state estimation model for different scenarios including nonlinear
Kalman filter algorithm for system consisting of multi-anchor ranging and inertial
navigation, and optimization model for system consisting of single-anchor ranging
and pedestrian dead reckoning(PDR). The following description in detail:
Firstly, demonstrate how the number and spatial layout of anchors deployed in
the scenarios impact of location error, and then give instructional recommendations
about actual deployment;
Secondly, put forward a simple and effective parameter calibration method, such
as static zero bias calibration, noise component analysis for low-cost
MEMS(Micro-Electro-Mechanical System) inertial sensor;
Thirdly, study Kalman filter based on UWB ranging and inertial navigation
system, and compare the performance of Extended Kalman filter(EKF) to unscented
Kalman filter(UKF), and realized non-line-of-sight(NLOS) discrimination and
adaptive filtering.
Fourthly, study the key algorithm of single-anchor and Foot-Mounted PDR
评论0