Study On Error Correction Method Of Pipeline Geographic Coordinate Measurement System

Normalization detection of long distance oil and gas pipeline is an important measure to ensure the pipeline safe operation. Pipeline geographic coordinate measurement system can locate pipeline flaw geographic coordinate and depict pipeline trajectory with in normal cleaning pipeline and flaw detection, so that the maintenance cost can be reduced and the pipeline change of stress / strain can be monitored and evaluated. It is important significance to improve the repair efficiency, safety of pipeline, pipeline operation auxiliary piping prediction of perfect information. The core of pipeline geographic coordinate measurement system is strapdown inertial navigation system, the biggest feature of the system is simple in hardware structure, software is extremely complex and diverse. For the existence of inertial sensitive element errors, the initial value and installation errors, and these in the navigation complex calculation process accumulate unceasingly with time(especially the cumulative position error after two integral computed). This thesis is about study on several kinds of error estimation and compensation, the main work are:(1) Firstly the SINS navigation program was designed, using a quaternion attitude and use 4-order Runge-Kutta method for solving attitude quaternion update. Such method is to improve the accuracy and speed of the numerical solution. Analysis of noise characteristics of MEMS sensitive element, then the random noise model based on AR(1) is established. The method of first mean value and then kalman filter is used to delete impulse noise caused by pipe welds, bumpy motion and random noise. Test results show this method can significantly improve pitch angle, roll angle and height.(2) The installation error of strapdown inertial system on the pipeline pig is one of the important factors affecting the accuracy of pipeline geographic coordinate measurement. This paper studied the influence law of the installation error of the navigation solution, and defined installation error angle is the no coincide between the coordinate axis of the pipe detector and the strapdown inertial system. Installation error transfer model was established, and the SINS pitch and yaw misalignment angle as a state variable, based on the initial segment of pipe laying flat characteristics, establish a direction constraints based on the initial segment of pipeline and odometer speed constraints for observations of SINS/odometer /GPS for magnetic punctuation integrated navigation model, and the nonlinear cubature kalman filtering is used to estimate the installation error of the pitch and yaw angle.Three experiments show that the method can effectively estimate the installation error, installation error estimation precision of 0.2 ’- 0.6’ and can highly solute trajectory error and height error, the precision of measuring geographic coordinate can reach 1.07e-3 degrees; high solution precision can reach 1.59e-3m.(3)Initial alignment accuracy is an important factor which influences the precision of the pipeline geographic coordinate measurement. Aiming at the dynamic initial alignment problem, a kind of new pipeline inertial measuring initial alignment algorithm is put forward. Established the initial alignment observation equation with the pipeline direction angle to pipeline reference point as the center of the 100 meter section, location and pipeline detector operating speed as dynamic measurement view. The combination algorithm with the model error predictive filter and cubature kalman filter is used to evaluate the state vector. The model error predictive filtering algorithm is to correct one-step prediction value of cubature kalman filter system. The algorithm can be real-time prediction model error, and can reduce the state dimension of cubature kalman filter and cut down the estimated time. Experimental results show that: the car dynamic alignment accuracy can reach 0.001°, proposed initial alignment algorithm is suitable for complex running environment, uncertain system model error and strong nonlinear.(4)No real-time GPS information in closed oil and gas pipeline can be used to correct the error of navigation parameters, only the odometer’s velocity information and distance information can be used. So error correction method of two level combination filtering algorithm is proposed. The front stage filter is forward filter based on self-constraint of odometer velocity,navigation error is evaluated by the CKF and KF combination filter, and the position information output from the front stage forward filter is made as the reference trajectory for the later stage filter. The later stage filter is backward filter based on point reference reverse smooth filter due to the location information can be obtained at pipeline magnetic punctuation and pipeline exit, and can further correct and amend the position parameter from the front stage filter. The experiment results prove that the algorithm is advantage,and location precision reach 10e-3 deg.