Research On Indoor Positioning Technology Of Mobile Robot Based On UWB/IMU

The acquisition of accurate location information in indoor scenarios is an important factor in the era of the Internet of Things,such as intelligent manufacturing and smart home,which puts forward new challenges and high requirements for low-cost and high-precision indoor positioning and navigation technology.Due to the complex indoor environment,a single sensor cannot meet the positioning requirements.Therefore,this paper deeply studies the indoor positioning technology with Ultra-Wide Band(UWB)technology as the core and inertial measurement unit(IMU)as the auxiliary.First of all,according to the application scenario,the requirements analysis of the positioning scheme is proposed,so as to carry out the scheme design.UWB has the advantages of strong interference immunity and high multipath resolution,but may cause unstable positioning results under non-line-of-sight conditions.IMUs can provide more accurate position solving information in a short period of time,but due to the measurement error and cumulative effect of the sensor,its positioning and navigation errors increase rapidly over time.Secondly,the positioning principle and error factors of UWB and IMU sensors are analyzed,and the positioning algorithms of the two are derived in detail based on this,different data analysis methods are used to process them,and the signal processing algorithms of UWB in non-line-of-sight conditions are analyzed,among which the deep learning method has a good performance effect.Then,Kalman filtering was selected as the core algorithm to tightly combine the UWB/IMU.Graph optimization theory is introduced,and the g2 o solver is used to post-process the sensor data,thereby improving the overall positioning accuracy.In addition,in order to save computing resources,the localization effect of adaptive difference theory on UWB under non-line-of-sight conditions is discussed.Finally,the test results in the actual field show that this paper verifies the positioning accuracy of UWB positioning at the line of sight,and more than 99% of the positioning point error remains within 6cm.In the non-line-of-sight case,the accuracy of the UWB/IMU fusion algorithm is high,and the combined positioning and graph optimization methods can keep the error within 14 cm.Scenarios such as indoor mobile spraying and fast movement of AGVs can meet their daily needs.