Research On High-precision Cooperative Localization Algorithm Under Non-ideal Conditions

As an extension of indoor positioning technology,indoor positioning is widely used in public security,emergency response,environmental monitoring,positioning and navigation,military reconnaissance,and targeting targets.However,due to the existence of some non-ideal conditions,the positioning accuracy will be reduced,such as non-line-of-sight errors,time synchronization errors,incomplete ranging information,multipath effects,multiple access interference,and node topology.How to design an algorithm to improve the positioning accuracy under non-ideal conditions is the key to the current indoor collaborative positioning research.Based on this,this paper mainly studies the indoor co-location problem with non-line-of-sight error,time synchronization error,and incomplete node ranging information.(1)There are many facilities and complex scenes in the indoor environment,and the non-line-of-sight error caused by obstacles will seriously affect the positioning accuracy.This paper firstly summarizes the existing algorithms with non-line-of-sight errors into two categories.The first category is to use statistical methods,geometric relationship methods,machine learning methods,etc.to identify non-line-of-sight errors.The second category is to use the introduction of Balanced parameter method,convex optimization algorithm and multi-dimensional scaling algorithm are used to suppress non-line-of-sight.Then,according to the characteristics of the actual ranging data,a non-line-of-sight identification algorithm is designed,which mainly includes two parts: calibration system error and non-line-of-sight identification.The success rate of distance recognition is 91.7%,and the non-line-of-sight recognition rate in dynamic scenes is 100%.A large number of experiments have verified the feasibility and effectiveness of the algorithm.(2)Due to the influence of crystal vibration,clock offset and clock skew,time synchronization error will be caused.This paper introduces several algorithms to reduce or eliminate the influence of time synchronization error on positioning results,including:iterative arrival time algorithm,linear position line algorithm,classical CHAN algorithm,quadratic programming algorithm,using weighted least squares algorithm to an improved algorithm for sub-programming problems.And it is applied to 2D single target,2D multi-target,3D and various positioning scenarios considering the influence of clock skew and clock offset for simulation test,which overcomes the defect that the previous algorithm simulation test has few application scenarios.The results show that the iterative time-of-arrival algorithm has smaller root mean square error,higher positioning accuracy and stable positioning results,and the algorithm has universal applicability to each positioning scene with time synchronization error.(3)Due to electronic interference,limited detection range and other factors,the distance between some nodes is not measurable,so it is impossible to obtain a complete distance matrix.In addition,since the parameters of each node are different,it is also worth considering whether the transmission range of the data has an influence on the positioning result.In this paper,considering the incompleteness of ranging information and whether it is affected by the communication radius,the RCMDS algorithm and the RSMACOF algorithm are proposed based on the existing matrix completion and MDS algorithms.Compared with other similar combined algorithms,the results show that with the increase of the number of missing nodes pairs and less than 25 pairs,the average positioning error of the two algorithms proposed in this paper increases after 500 Monte Carlo simulations,but it reaches below the meter level,and both show the best positioning performance.