Research On The Algorithms Based Seamless Outdoor/Indoor Positioning Based On GNSS/MIMU

People spend more than 80% of their time at indoors,and there is a strong demand for indoor positioning.Outdoor positioning usually adopts satellite navigation positioning(GNSS).Indoor technology can be divided into inertial measurement unit(IMU),Wi-Fi,radio frequency label(RFID),Zigbee,Bluetooth,UWB,etc.Indoor positioning technology based on micro inertial measurement unit(MIMU)is an important development method of indoor positioning technology,which provides users with location information through the algorithm such as dead reckoning or strap down Inertial navigation,with high autonomy and strong anti-interference ability.However,this technology only uses low-cost MIMU,which can only determine its relative position,and the positioning accuracy will rapidly decline with time,which limits the application field of this technology.In order to overcome the limitations of single technology,to achieve seamless positioning of indoor and outdoor ability,this article studies of algorithms of the indoor and outdoor seamless positioning based on GNSS/MIMU,and focuses on the detection of zero velocity state,zero velocity correction,indoor absolute heading and the key technology such as heading correction,the design scheme of the indoor and outdoor seamless,indoor and outdoor seamless positioning software is developed,and verified the accuracy.This paper mainly completes the following five aspects of work.1.Aiming at the problem that the traditional fixed threshold zero-velocity detection algorithm is liable to misjudge and misjudge the zero-velocity state under different motion modes,a dual-threshold zero-velocity detection algorithm suitable for mixed motion mode is proposed.The algorithm uses the feature point recognition of angular velocity change to get the motion pattern attribute and adaptively matches the corresponding acceleration and angular velocity threshold as well as the time window size to carry on the zero velocity state detection.Experimental results show that the success rate of motion pattern recognition over 98%.2.The algorithm for solving indoor absolute heading is studied,the real-time solving model of absolute position information from outdoor to indoor is derived,and the initial position coordinates and absolute heading are provided for MIMU by using mobile phone GNSS positioning information.The experimental results show that under the condition of mobile phone GNSS positioning precision meet the requirements,the final positioning error is 0.7 m,accounting for about 0.7% of the total length of track,the final course offset error is about 1 °.3.In view of the problem that the indoor launched positioning service is unable to obtain the absolute position information accurately,the postmortem calculation model of the absolute position information indoors to outdoors is established.After the users walk to outdoors,the absolute position coordinates and the heading information of the indoor point position can be calculated through the absolute directional inverse calculation.The experimental results show that after using the absolute position calculating model,get the final positioning error is 0.97 m,accounting for 0.35% of the total length of track,eventually return to the starting point of heading deviation is about 0.4 °.4.Aiming at the problem of heading deviation error accumulation caused by low-precision MIMU after long and long walking,a course repeated correction algorithm is proposed,and the general formula of N times course correction is deduced in detail.The experimental results show that the error of heading drift is reduced to 1.86° from 32.18°and the positioning accuracy is improved to 1m compared with the results of repeated course correction.5.Analyzed structure of the algorithm,and designed the mobile app based on the android system,which includes the functions of data reception,data processing,storage and map display.The validity of the software is verified by the long distance of indoor and outdoor walking experiment,and the experimental results show that the positioning accuracy of the software is about 1%d,the real-time display function of the map and the data storage function can be realized.