Research On Key Technology Of Digital Twin Localization In The Underground Coal Mine Based On UWB

The location of target in the underground coal mine is an important support for the realisation of smart mine.Due to the limitation of conditions,the evaluation of underground localization algorithms is mostly based on simulation environment,without fully considering the impact of factors such as actual tunnel environment and positioning base station deployment on localization accuracy.In addition,due to the complex and harsh underground environment,localization techniques that work well on the ground cannot be directly applied to the underground.To address the above problems,this thesis designs a framework for underground digital twin localization system in coal mine to provide a mine twin environment for the evaluation of localization algorithms,and proposes improved localization algorithms based on Ultra Wide Band(UWB)and Inertial Measurement Unit(IMU)to improve the performance of underground target localization.The main research contents are as follows:(1)Aiming at the problem that the underground localization algorithm is difficult to evaluate in the field,and the simulation experiment based on Gaussian noise assumption cannot guarantee the authenticity of the evaluation results,a framework of digital twin localization system for coal mine is proposed.The framework includes five parts: physical space,digital space,twin service,twin data and connectivity,which can realize real-time acquisition of underground environment and ranging information,and provide users with services such as twin environment customization and twin localization experiments.In order to simulate the actual ranging error of the UWB base station in the twin environment,a ranging behaviour model is proposed in the digital space,which effectively portrays the influence of obstacle and distance on the ranging accuracy.(2)Aiming at the problem of unreasonable Particle distribution and particle depletion in the Extended Particle Filter(EPF)algorithm,firstly,an AF-EPF algorithm based on target acceleration is proposed.The particle state is shifted by using the particle triaxial acceleration probability and the weight offset parameter to improve the diversity of particles.Further,the AF-TP-EPF algorithm is proposed by studying the particle selection optimization strategy based on the constraints of the tunnel topology and the target motion,using the tunnel structure characteristics as well as the acceleration information obtained by IMU to constrain the particles.The superiority of the above algorithms over EPF is verified through experiments.(3)On the basis of the above research,the digital twin localization system in coal mine is developed,and the functions of twin roaming,twin environment customization and twin localization experiment are realized.Aiming at the complex environment and actual twin service requirements in coal mine,variable optimization and model hierarchical synchronous loading are completed based on Vue lifecycle and Promise mechanism,Web Socket and JSON are used for real-time communication and algorithm call,and twin picking,target manipulation and collision detection are realized by means of keyboard and mouse monitoring events.It improves the system performance and user experience.This thesis has 47 figures,9 tables,and 85 references.