Research And Application Of Trenchless Pipeline Trajectory Mapping Technology Based On Inertial Navigation System

City underground pipelines,due to their environmental friendliness,reliability,and high efficiency,have gradually become the"lifeline"of urban areas.However,the increasingly dense underground pipelines have brought considerable challenges to later maintenance and subsequent pipeline laying work.Therefore,many governments of cities in China have issued regulations requiring the construction party to provide pipeline trajectory information after underground pipeline laying.Due to the limitations of the current non-excavation construction skills,there are often large differences could be shown between the pipeline design trajectory,the guiding trajectory and the final laying pipeline trajectory.Hence,a pipeline trajectory measurement system is needed to obtain accurate position information of non-excavation laying pipelines.However,it is difficult to measure the trajectory of small-diameter non-metallic pipelines laid by trenchless engineering,and indirect detection methods cannot meet the accuracy requirements.To solve this problem,this thesis focus on an underground pipeline trajectory measurement technology based on inertial navigation system,and develops a system that includes a trajectory measuring instrument,an automatic rope puller and a PC software.It also establishes an error propagation model based on this system,and proposes a series of algorithms for data preprocessing,data fusion and trajectory correction.The results show that the maximum errors of elevation and plane are controlled within 0.15%and 0.2%,respectively.The main work and achievements of this thesis are as follows:（1）The research and development of a pipeline trajectory measurement system was completed.The trajectory measuring instrument is the core device of this system,which is the carrier of the trajectory information acquisition sensors.Besides ensuring the smooth and efficient operation of the sensors in the pipeline,it also realizing the functions of sensor control,data temporary storage and upload.Based on these requirements,the electronic circuit and mechanical structure of the trajectory measuring instrument are designed and developed.In addition,an automatic rope puller is the power mechanism of this system.In order to overcome the shortcomings of manual rope pulling and ordinary winch rope pulling,this thesis develops an automatic rope arranging system and a tension monitoring system.The former realizes that the rope was tightly and neatly arranged on the rope reel,avoiding the situation that the rope was knotted or even slipped out of the reel due to chaotic arrangement.The latter realizes overload protection function by real-time monitoring of rope tension,avoiding rope breakage or even instrument damage caused by forced pulling when the instrument was stuck by foreign objects in the pipeline.Finally,through MATLAB platform,upper computer software was written to realize communication between upper computer and trajectory measuring instrument,data transmission and trajectory calculation functions.（2）The sources of error for this system were analyzed,and the differential equations between each error source and the final trajectory error were established.The modeling descriptions of each error source were given,and the error propagation simulation analysis was carried out.The following conclusions were obtained:Firstly,at a sampling frequency of 200Hz,the point spacing is very small,and the accuracy of the full-angle half-distance method is unsatisfying.The maximum elevation error and maximum plane error are 0.26%and 1.58%,respectively.The curvature radius method,full-angle full-distance method,and equal-angle full-distance method have similar accuracy,with maximum elevation error and maximum plane error of 0.13%and0.79%,respectively.Considering both accuracy and computational complexity,the full-angle full-distance method is an optimal algorithm.Secondly,the trajectory error is positively correlated with the gyroscope random walk,and the correlation is stronger in large span ranges（one order of magnitude or more）and weaker in small span ranges（within one order of magnitude）.The reason is the gyro bias walk shows randomness;The odometer pulse measurement error shows less randomness,so it has a stronger positive correlation with the trajectory error;The wheel diameter error is relatively stable and shows a strict positive correlation with the trajectory error.Finally,for this system（using tactical-grade IMU,odometer pulse error of about10%,wheel diameter error of about 1mm）,the trajectory errors（maximum elevation error and maximum plane error）caused by gyroscope random walk,odometer pulse error and wheel diameter error all reached more than 0.9%,which are the main sources of trajectory error.This means that,it is necessary to suppress these three errors.（3）For the attitude angle error,this thesis uses the extended Kalman filter algorithm to suppress it,where both pitch angle and roll angle can be obtained by fusing acceleration data and angular velocity data.When the system noise covariance matrixQ=10^-6·E_4×4and observation noise covariance matrixR=1 00·E _3×3,the fusion effect is better;azimuth angle can be obtained by fusing magnetic field data and angular velocity data.WhenQ=10^-6·E _4×4 andR=3·E_3×3,the fusion effect is better.For odometer pulse error and wheel diameter error,both can be suppressed by using the average correction algorithm and geometric correction algorithm proposed in this thesis.As a result,the trajectory error is reduced from more than 0.9%to less than 0.2%.（4）In order to test the accuracy and stability of this system,an experimental site with two pipelines was set up,and 10 tests were conducted in each pipeline.Among them,the maximum elevation errors of 10 tests in pipeline 1 were all controlled within 0.15%,with an average value of 0.1073%;9 out of 10 tests showed maximum planar errors within0.2%,and 1 was 0.2031%,with an average value of 0.1740%.The maximum elevation errors of 10 tests in pipeline 2 were all controlled within 0.15%,with an average value of0.1225%;the maximum planar errors were all controlled within 0.2%,with an average value of 0.1815%.This indicates that this system shows good accuracy performance.In addition,this system was also applied in the field,and the results showed that the system had good sealing performance,fast data upload speed,easy-to-operate software,and could quickly produce maps at the construction site,indicating its good practicability and reliability.