Research On UWB High-precision Positioning Algorithm Based On Vehicle Motion Characteristics

With the rapid popularization of intel igent connected cars,the auto-driving industry has attracted more and more researchers.When facing the requirements of autonomy and intel igence,realizing high-precision,high-real-time,and high-reliability positioning of vehicles is an important technical link in the autonomous driving scheme.RTK technology based on GNSS and ground-based enhancement is one of the commonly used positioning methods in open-air areas.However,GNSS cannot provide navigation and positioning information normally in complex environments with obstructions such as underground parking lots,high-rise buildings,and indoors.Therefore,research on highprecision indoor positioning technology has become a hot spot today.Ultra-wideband positioning technology(UWB)and inertial navigation positioning technology(IMU)stand out among many indoor positioning technologies.Among them,UWB positioning technology has high time resolution and can provide high-positioning single-point positioning coordinates.Each positioning calculation can be regarded as a random observation of the vehicle position,but there will be no cumulative error;the IMU is a positioning technology that is not affected by the outside,and can independently calculate the state of the carrier’s motion by relying only on its own accelerometer and gyroscope.However,in the process of trajectory calculation,there will be cumulative errors caused by two integrations,which cannot be long time to continuously achieve track tracking.Both UWB and IMU cannot overcome their technical difficulties by themselves.In order to solve the inherent shortcomings of a single sensor,the joint positioning technology makes full use of the advantages of multiple sensors through fusion algorithms to achieve better positioning results.It has a high degree of attention in the frontiers of scientific research and engineering practice.Based on the above requirements,this article first introduces the principles of ultrawideband positioning technology and inertial navigation positioning technology.Using theoretical calculations and experimental verification methods to analyze the error transmission characteristics of the two positioning technologies.And aiming at the differences and complementarities between the two,a multi-sensor combination positioning method is proposed.In addition,the use of the model can make the combined positioning algorithm more effective because the car has a universal and mature motion model.In this paper,a combined Kalman filter algorithm is designed to perform multi-source information fusion including vehicle motion model,IMU sensor and UWB system,and provides reliable and stable location services for vehicles.What’s more,this article also designs an embedded platform for the implementation of the above algorithm,and carries out the performance verification and analysis of the combined positioning system on the built experimental platform.According to the analysis of the actual vehicle test data,the comprehensive output positioning trajectory is better than the positioning effect of the UWB single positioning.Compared with the result of UWB single positioning,the positioning accuracy under the fusion algorithm is improved by 48.45%,and the positioning error is basically within ±10cm.Reliable attitude information can be provided at the same time,which meets the requirements for high precision,high real-time,and high reliability of auto-driving cars.