Research On Indoor Fingerprinting Localization Algorithm Based On BIM

Since the industrial civilization,human society,economy and culture have developed rapidly,and countless high-rise buildings have been erected all over the world.In the complex urban environment,people's demand for indoor localization service is also increasing.However,the global satellite navigation and localization technology has been developed to this day and still cannot meet people's demand for high-precision indoor localization.In recent years,wireless localization technologies such as Ultra Wide Band(UWB)have been used for indoor localization.However,in indoor environments where wireless signals are blocked,non-line-of-sight and multipath signals greatly reduce the localization accuracy.Some researchers extracted the features of these multipath signals with non-line-of-sight errors and make them into fingerprint database,and propose many fingerprint location methods.However,in a city full of tall buildings,it is a huge project to construct an offline fingerprint database for each floor of each tall building that needs location services.It requires a lot of labor cost,equipment cost,and a lot of measurement time.Moreover,if indoor furniture,buildings and other objects are moved,replaced,discarded,etc.,or indoor people moving around,it will affect the measured fingerprint library,and even cause the fingerprint library to expire and need to be re-measured,which not only affects localization accuracy,but also greatly increase the cost of indoor localization.In order to reduce the cost of traditional fingerprinting localization to build a fingerprint database,this paper combines Building Information Modeling(BIM)to replace the traditional manual method of measuring the fingerprint database by computer calculation of the fingerprint database,and a fingerprinting localization method using BIM prior information is proposed.The main research contents of this paper are as follows:First of all,this paper uses the prior information of indoor 3D buildings provided by BIM,and proposes a BIM-based distance fingerprint library construction method to solve the shortcomings of traditional fingerprint libraries.This method not only has fast construction time,but also does not require manpower and measurement equipment.Even if the indoor objects or anchor node layout position changes and other factors lead to the expiration of fingerprint database,it can be solved by quickly rebuilding the fingerprint database.Secondly,based on the construction of the BIM distance fingerprint database,fingerprint matching algorithms are adopted on the ultra-wideband Localization data set,and the Pauta Criterion is used to process the singular measured values in the online localization stage to reduce the influence of the wrong measurement data on the localization results.The Test results show that the BIM-based distance fingerprint database construction method proposed in this paper greatly reduces the construction time of the fingerprint database,and the fingerprint location method using BIM prior information has better location performance than the traditional least square location algorithm.