Research on tight combination positioning method of coal mine robot based on UWB and IMU

In view of the complex underground environment of coal mine, the existing coal mine robot positioning methods are affected by factors such as non-line-of-sight errors, resulting in low positioning accuracy and low real-time performance, a tight combination positioning method of coal mine robot based on UWB (ultra-wideband) and IMU (inertial measurement Unit) is proposed. First, the UWB module is used to measure the distance between the coal mine robot and the UWB base station, and the real and measured values of the distance between the coal mine robot and the UWB base station are used to train the least square support vector machine (LSSVM) model, and the modified LSSVM model is obtained. Then, the measured value measured by UWB module in the positioning process of coal mine robot is used as the input of LSSVM correction model, and the measured value of UWB is modified by LSSVM correction model to reduce the influence of non-line-of-vision error on positioning accuracy and obtain more accurate distance information. Finally, the range information corrected by the LSSVM correction model is used as the measurement of error state Kalman filter (ESKF), and the position information calculated by the inertial navigation is formed into the measurement equation. ESKF is used to tightly combine the UWB range correction and the distance information calculated by the inertial navigation to complete the status update and obtain more accurate position information. Realize the precise positioning of coal mine robot. The simulation results of UWB base station under different layout schemes show that using LSSVM to modify the model can make UWB ranging information more accurate and improve the positioning accuracy. In the static positioning experiment, when the four UWB base stations are arranged symmetrically, the root-mean-square error of positioning decreases from 0.146 4 m to 0.1398 m. When the four UWB base stations are arranged symmetrically with different heights, the root-mean-square error decreases from 0.300 8 m to 0.200 6 m. When the four base stations are arranged irregularly, the root-mean-square error decreases from 0.317 5 m to 0.314 2 m. Therefore, in the actual scene, the UWB base station should be arranged in equal height symmetry as far as possible. In the dynamic positioning experiment, the fusion positioning trajectory modified by the LSSVM is closer to the real trajectory of the coal mine robot than that before the correction, which verifies that the compact combination positioning method can reduce the non-line-of-sight error and improve the positioning accuracy.

0 Introduction

The working environment of underground coal mine is bad, which poses a great threat to the personal safety of underground workers. Unmanned or under-manned mining of coal mines has increasingly become a research hotspot. The introduction of coal mine robots to replace underground workers to complete dangerous and heavy underground work will play an important role in solving the problem of coal mine safety production [1].

Coal mine robot can be used in mining, transportation and rescue, and accurate positioning is the basis of its intelligent realization. In order to solve the problem of accurate positioning of coal mine robots, the common positioning methods include radio frequency identification positioning technology, ultrasonic positioning technology, Ultra Wide Band (UWB) positioning technology, inertial navigation positioning technology and other methods. Zhang Xiaoli et al. [2] proposed a method of fusion of radio frequency identification information and inertial navigation solution information using extended Kalman filter to realize real-time and high-precision positioning of coal mine robots. Tan Yuxin et al. [3] proposed an ultrasonic network positioning algorithm based on lossless Kalman filter, which combined the ultrasonic network positioning with the heading Angle information and position coordinate information obtained from the positioning of electronic compass and optical code disk to achieve the purpose of reducing positioning errors. Chen Meirong et al. [4] proposed an UWB based hybrid location solution method for underground coal mine, which used the hybrid location method of brainstorming optimization and Taylor series expansion to solve the problem of the need for good initial values for Taylor series expansion. Ma Hongwei et al. [5] proposed a coal mine robot positioning method based on the fusion of strapdown inertial navigation and odometer. Firstly, the strapdown inertial navigation was calibrated by Kalman filter, and then the location information of the coal mine robot calculated by strapdown inertial navigation and the location information calculated by odometer were calibrated by adaptive Kalman filter to obtain combined positioning results with high positioning accuracy. Yang Jinheng et al. [6] proposed a dual inertial navigation (INS) positioning method based on adaptive Kalman filter, and used the acceleration information and angular velocity information obtained from two INS systems to establish a dual INS model. However, due to the complex underground environment, the application of many positioning techniques is limited. The radiofrequency technology can not track the coal mine robot in real time, and the ultrasonic technology will be affected by the Doppler effect, so the positioning accuracy is low and the cost is high. As an emerging technology, UWB has a high time resolution [7]. When it is applied in indoor environment positioning, it has higher stability and positioning accuracy than other positioning technologies. However, when UWB technology is used alone in coal mine robot positioning, the positioning accuracy is affected by UWB base station layout, multipath effect, non-line-of-sight error, etc., and the positioning results have certain fluctuations. When the inertial measurement Unit (IMU) is working, the environmental factors have less interference to it, the update rate is high, and the positioning accuracy is high in a short time, so it can adapt to more complex downhole environment. However, the main disadvantage of using IMU positioning alone is that its positioning errors will accumulate over time, resulting in low positioning accuracy for a long time. In order to realize the accurate positioning of the coal mine robot, Huo Yuanheng [8] combined the UWB ranging information with the IMU positioning information through extended Kalman filter, and the positioning accuracy was improved to a certain extent. However, due to the influence of non-visual range errors and other factors in the UWB ranging process, the positioning results still had certain errors.

In order to reduce the non-line-of-sight error of UWB ranging and realize the accurate positioning of coal mine robot, a compact combination positioning method based on UWB and IMU is proposed in this paper. The Least Square Support Vector Machine (LSSVM) was introduced to correct the UWB ranging information to reduce the non-line-of-sight error of downhole positioning. By using the ErrorState Kalman Filter (ESKF), the modified UWB ranging information and the distance information calculated by the inertial navigation are tightly combined, and the position information of the coal miner is updated, and the precise location is realized. By using the experimental data obtained from UWB module and IMU, the static and dynamic positioning simulation experiments of the compact combination positioning method are carried out in Matlab, and the results prove the reliability of the method.

1 Tight combination positioning method

The principle of the coal mine robot tight combination positioning method based on UWB and IMU is shown in Figure 1. The UWB ranging module was used to obtain the measured distance between the coal mine robot and the UWB base station. The real and measured distance between the coal mine robot and the UWB base station were used to train the LSSVM model, and the LSSVM modified model was obtained, and the UWB ranging information during the positioning process of the coal mine robot was corrected. The acceleration and angular velocity information of the coal mine robot is collected by IMU [9], and the status information of the coal mine robot is solved by inertial navigation. The range information corrected by LSSVM is used as the measurement input of ESKF, and the position information calculated by inertial navigation is used to form its measurement equation, complete the state update, and obtain more accurate position information of the coal mine robot, so as to realize the accurate positioning of the coal mine robot.

2 Experimental Analysis

In order to verify the positioning accuracy of the mine robot tight combination positioning method based on UWB and IMU, experiments were carried out in the simulated underground roadway. In the experiment, firstly, the static positioning experiment is carried out on the different layout of the base station, the influence of the location of the UWB base station on the positioning accuracy is analyzed, and the best scheme of the base station layout is determined. Then the UWB range information before and after the LSSVM correction model is used to carry out the UWB and IMU tight combination positioning experiments respectively, and the positioning errors of the two methods are compared. Finally, the dynamic experiment of UWB and IMU is carried out to compare the motion trajectories of the coalmine robot before and after the modified LSSVM model.

3 Conclusion

(1) A compact combination positioning method for coal mine robot based on UWB and IMU is proposed. The LSSVM correction model is used to correct the measured UWB ranging information and reduce the influence of non-line-of-sight errors on UWB ranging. ESKF is used to combine UWB and IMU tightly to achieve the purpose of improving positioning accuracy.

(2) The influence of UWB ranging information processed by LSSVM modified model on the positioning accuracy of coal mine robot under different layout schemes of UWB base station is studied. The experimental results show that when the UWB base station is arranged with medium height symmetry in the application scene, the root-mean-square positioning error is smaller than that of other arrangements. Therefore, in the actual scene, the UWB base station should be arranged with equal height symmetry as much as possible. The UWB ranging information before and after the LSSVM modified model was modified by UWB and IMU tight combination experiments respectively. The results showed that: After the modified LSSVM model, the root-mean-square error of the combined location is reduced, and the fusion location trajectory after the modified location information is closer to the real trajectory of the coal mine robot than the uncorrected fusion location trajectory. The modified LSSVM model has higher combined positioning accuracy after correcting UWB ranging information, and is more suitable for coal mine robot positioning.

References:

[1] LIU Yulong. Research Status and key technology analysis of Search and rescue robot in Coal mine [J]. Mining Machinery, 2013,41 (3) : 7-12.LIU

[2] Zhang Xiaoli, Wang Zhangzhe. Real-time and high-precision positioning method of downhole inspection robot [J]. Mining Research and Development, 2021,41 (10) : 158-161.

[3] Tan Yuxin, Yang Wei. Ultrasonic network positioning method for underground robot based on UKF [J]. Journal of China Coal Society, 2016,41 (9) : 2396-2404.

[4] Chen Meirong, Wang Kai, Zhang Jiachun, et al. Hybrid solution method of ultra-wideband positioning in coal mine [J]. Industrial and Mine Automation, 2021,47 (3) : 53-59.CHEN Meirong,

Research on downhole robot location method based on strapdown inertial navigation and odometer [J]. Industrial and Mine Automation, 2019,45 (4) : 35-42.

[6] Yang Jinheng, Song Sanyang, Tian Muqin, et al. Dual inertial navigation shearer positioning method based on adaptive Kalman filter [J]. Industrial and Mine Automation, 2021,47 (7) : 14-20, 28.

He Lei, Wei Mingsheng, Qiu Xinyu, et al. Research on underground personnel location algorithm based on UWB [J]. Industrial and Mine Automation, 2022,48 (6) : 134-138.

[7] Huo Yuanheng. Research on indoor unmanned vehicle navigation Technology based on inertial navigation and UWB positioning [D]. Harbin: Harbin Institute of Technology, 2021.

[8] Niu Xiaoji, Kuang Jian, Chen Qijin. Feasibility study of detection and positioning Scheme in small diameter pipeline using MEMS inertial navigation [J]. Chinese Journal of Sensor Technology, 2016,29 (1) : 40-44.

Sensors and Microsystems, 2019, 40 (10) : 147-150.

[10]XU Yuan, LI Yueyang, CHOON K A, et al. Seamlessindoor pedestrian tracking by fusing INS and UWB

[11]measurements via LS-SVM assisted UFIR filter[J].Neurocomputing, 2020, 388:301-308. DOI: 10.1016 / j. eucom 2019.12.121. Jiang Lailai. Research on NO_x emission prediction method of power plant based on LSSVM optimization [D]. Baoding: North China Electric Power University, 2021.JIANG Lailai. Research on NO_x emission predictionmethod of power plant based on optimized LSSVM [D].Baoding: North China Electric Power University, 2021.

[12] Zou Qiang, Lu Fuguang, LAN Kui Bo, et al. Research on UWB/INS combined indoor localization Method based on ISRUKF [J]. Journal of Tianjin University (Natural Science and Engineering Technology Edition), 2022,55 (5) : 496-503.ZOU Qiang,

[13]SHIN E H. Estimation techniques for low-cost inertialnavigation[D]. Calgary: University of Calgary (Canada), 2005.

[14] Qin Yongyuan, Zhang Hongyue, Wang Shuhua. Kalman filter and Integrated navigation Principle [M]. Xi 'an: Northwestern Polytechnical University Press, 2015.

[15]FENG Daquan, WANG Chunqi, HE Chunlong,  et al.Kalman-filter-based integration of IMU and UWB for high-accuracy indoor positioning and navigation[J]. IEEE  Internet of Things Journal, 2020, 7 (4) : 3133-3146.

[16] Xu Aigong, Liu Tao, Sui Xin, et al. UWB/INS compact combination method for indoor positioning and attitude determination [J]. Journal of Navigation and Positioning, 2017, 5 (2) : 14-19.

[17] Wang Chuanyang, Wang Jian. Research on UWB emergency positioning base station layout [J]. Science of Surveying and Mapping, 2019,44 (8) : 174-181. (in Chinese)