| Achieving continuous and high-precision indoor positioning services is a key research direction for indoor positioning.In recent years,indoor positioning service scenarios have been expanded from general indoor open scenarios to special positioning scenarios such as underground narrow spaces.In underground tunnels and corridors,the positioning service faces such challenges as long and narrow space and limited navigation signal propagation,so the study of underground positioning has become a hot spot for indoor pervasive positioning research.Ultra-wideband(UWB)technology has the advantages of high-precision ranging and multi-path resistance,and can provide high-precision positioning results when applied to indoor positioning scenarios.UWB positioning accuracy can reach centimeter level in unobstructed open indoor scenes.However,the deployment of base stations is limited in the narrow underground space.Therefore,optimizing the base station deployment scheme in long and narrow underground spaces,studying the processing method of narrow and long mesh observation data,and improving the positioning method under the restricted base station deployment conditions have become the key technologies for positioning in long and narrow underground spaces.In this paper,we use UWB module to collect ranging data and design a scheme to build an underground control network based on UWB technology.The innovation of the application of sick free network parity in the data processing of narrow underground UWB network is advanced,and the establishment,maintenance,extension and recovery of narrow spatial control network are carried out in terms of theoretical analysis and actual measurement verification.The traditional position iterative solution method is improved to achieve high precision 3D positioning of unknown targets for the situation of base stations close to the common surface faced by underground UWB networks.The main contributions are as follows:(1)The method of parity processing for narrow-length control networks is proposed.Narrow and long control networks have pathological problems,and the solution results of traditional net leveling methods to deal with narrow and long spatial observation networks have large errors,and the leveling solution results are seriously affected by the observation errors.In this paper,we analyze the performance of different pathological network processing methods in narrow scenarios,analyze the effect of different methods from both theoretical basis and test results,and finally choose to use the overall least squares leveling method to level the narrow network,and the results improve the accuracy and reliability of base station point location solution and meet the requirements of location reference service.In the environment of 450m× 5m with strong pathology,the total error of ten reference points reaches 0.134 m,and its accuracy can meet the needs of base station service.(2)The high-precision three-dimensional positioning method is proposed under the condition of base station coplanarity.Underground space base station deployment environment is restricted,and the base station deployment is close to coplanar,so the positioning accuracy of elevation direction is poor in 3D positioning.In this paper,we improve the initial value selection and step size control strategy for the Newton-based iterative solution method to ensure the correct convergence of the iterative algorithm and significantly improve the positioning accuracy.The new method achieves an elevation direction positioning accuracy of 5 cm,which is 63% better. |