Research On SINS/GIS Based Integrated Inertial Navigation Technology Of Mine Inspection Robot

With the construction of intelligent mines,research on autonomous inspection robots to achieve unmanned mine inspection has become the research focus of mine inspection robots.It is difficult for a single navigation system to meet the requirements of autonomous navigation of robots,the use of integrated navigation technology to achieve precise autonomous navigation of robots is the trend of mine navigation research.In addition,the integrated navigation technology is a complex subject involving motion control,environmental cognitive positioning strategy,multi-sensor fusion,path planning,and also a technical difficulty of the robot.The research on the integrated navigation technology of mine inspection robot will provide a reference for the development of various autonomous robots,and promote the mine intelligent and unmanned process.There are strapdown inertial navigation system(SINS),geographic information system(GIS)and optical encoder(OE)that can be used on for the integrated navigation of inspection robot in the mine.How to effectively integrate them to improve the accuracy and anti-interference ability of the integrated navigation of inspection robot in the mine is the focus of this paper.This paper mainly discusses the navigation positioning principle and error model of the sensor,the multi-sensor integrated filter structure,and the optimization of the federal Kalman filter algorithm,the specific research contents are as follows:Firstly,the common coordinate system and coordinate transformation method of strapdown inertial navigation system are analyzed;The navigation parameter updating algorithm of strapdown inertial navigation system,the positioning algorithm of mine geographic information system and the dead reckoning algorithm of optical encoder are studied,and the error models of them are deduced.Secondly,the algorithm principle of standard Kalman filter and unscented Kalman filter is analyzed.Combined with the actual mathematical model of the integrated navigation system of inspection robot,standard Kalman filter is selected as the sub filter of the integrated navigation system.Centralized filter and federated filter are analyzed,and the federated filter is selected for the data fusion of integrated navigation system by comparison.Aiming at the traditional federation filter fixed allocation factor method is used to optimize,adaptive allocation factor method is used,and simulation results show that the algorithm can improve the navigation and positioning accuracy of the system.Then,aiming at the problem of abnormal system model and noise characteristics when the robot is running in the mine,the improved sage Husa filter is added to the robot federated filter,and an adaptive federated Kalman filter(AFKF)is obtained.AFKF can not only eliminate the influence of measurement gross error on the filtering accuracy,but also estimate the noise covariance on-line when it is abnormal,suppress filter divergence,and improve the accuracy and robustness of the robot integrated navigation system;A * path smoothing optimization algorithm was studied for the problem that the robot planned the path with many turns and is not smooth,and the effectiveness of the optimization algorithm was verified by simulation.Finally,the adaptive federated Kalman filter algorithm proposed in this paper is verified by simulation experiments and robot running experiments.The results show that AFKF can effectively control the system model and the impact of abnormal noise characteristics on the robot integrated navigation system,which makes the mine inspection robot integrated navigation system have better accuracy and robustness.