Research And Design Of Positioning Method Of Small Unmanned Mobile Platform Based On UWB

With the continuous development and expansion of the global transportation industry,people's positioning demand for vehicles,ships and other transportation tools is also increasing.GPS and BDS,as the common real-time satellite positioning and navigation systems in the world,have been widely used in the neighborhood of vehicle trajectory tracking,ship positioning and navigation.However,in some bad scenes such as high-rise buildings and closed Spaces in cities,the positioning performance is greatly reduced due to the weak signal strength of GPS and BDS positioning and navigation satellites.As an emerging ground auxiliary positioning technology,UWB positioning technology,due to its advantages of high positioning accuracy and low construction cost,just makes up for the shortcomings of positioning satellites that cannot accurately locate vehicles and ships in harsh scenes.At the same time,UWB is susceptible to the influence of objects in the environment,resulting in multipath interference.It is easily affected by the spatial layout of anchor points,resulting in poor positioning effect.Therefore,compared with only using satellite or UWB for positioning,using multiple positioning systems for data fusion positioning can ensure the positioning accuracy and stability of the system,which is the key research direction of positioning technology today.The experiment in this paper is a small unmanned mobile platform based on UWB and INS,and five UWB anchor points are deployed.Through static experiment,different UWB LS and GN positioning algorithms are compared and analyzed.Through dynamic experiment,the data fusion algorithm of UWB and INS based on different Kalman filter is compared and analyzed.Due to the complexity of the real environment,there will be a lot of NLOS effects(buildings,trees,human bodies,etc.).At present,most of the relevant studies on UWB positioning are based on LOS scene.Therefore,in this paper,in the dynamic experiment,the human body is used as NLOS shielding,and the UWB ranging error under LOS and NLOS environment is compared and modeled.At the same time,the link correlation and positioning accuracy between UWB anchor points and tag nodes are compared and studied.The main research results are as follows:(1)The root mean square error of the TOA-based LS positioning algorithm is 0.28 cm lower than that of the TDOA-based LS positioning algorithm,and the root mean square error of the TDOA-based GN positioning algorithm is 2.79 cm lower than that of the TDOA-based LS positioning algorithm.(2)The root mean square error of the data fusion positioning algorithm using linear Kalman filter is 4.36 cm lower than that of the GN positioning algorithm based on TOA,and the root mean square error of the data fusion positioning algorithm using extended Kalman filter is5.31 cm lower than that of the GN positioning algorithm based on TOA.(3)In the case of LOS,the UWB ranging error follows the Rice distribution,and the root mean square error of UWB positioning is 8.75 cm,and there are three link pairs with correlation.In the case of NLOS,the UWB ranging error follows lognormal distribution,and the root mean square error of UWB positioning is 60.30 cm,and there is only one link pair with correlation.In conclusion,the experiment verifies that the GN algorithm of UWB positioning has better positioning effect than the LS algorithm,and the Kalman filter data fusion positioning algorithm has higher positioning accuracy than the UWB positioning algorithm.At the same time,the experimental results show that when the human body acts as the NLOS interference on the transmission link,the ranging error and location error of UWB increase significantly,and the correlation between links decreases significantly.