Research On Error Processing Of Inertial Measurement While Drilling

In order to improve the efficiency of underground resource exploration and exploitation,inertial measurement while drilling technology has been widely used.Inertial measurement while drilling is a micro-inertial measurement unit(MIMU)composed of a three-axis accelerometer and a three-axis gyroscope,which measures and collects drilling tool information in real time,obtains the attitude angle of the drilling tool through attitude calculation,and then observes and controls the wellbore trajectory.Considering that MIMU works in complex environments such as high temperature,vibration,and shock for a long time,and the manufacturing process error of inertial devices,a large amount of interference noise will be aliased in the drilling tool information collected by MIMU,resulting in a large error in the calculated attitude angle of the drilling tool.Therefore,the error processing research of inertial measurement while drilling is very important.The specific research contents of this paper are as follows:(1)The laboratory-made attitude measurement module integrated with MIMU is used to measure the acceleration and angular velocity data of the drilling tool,the error characteristics and attitude error characteristics of the MIMU while drilling are analyzed,which lays theoretical foundation for the subsequent research on the error processing of MIMU.(2)Aiming at the device errors of the MIMU,the device error model to be calibrated is established,and the device error calibration algorithm based on 24-order EKF is proposed.At the same time,the iterative algorithm is used to estimate the residual error of the calibration parameters,and then the MIMU device error calibration model is obtained;The 24-position data acquisition scheme is designed,the improved RLS algorithm is used to preprocess the original data,and the device error is compensated by the calibration model,effectively realize the device error calibration of the MIMU.(3)Aiming at the problem of large error of gyroscope output signal in MWD,an adaptive evaluation parameter is proposed to improve the step size,crossover probability and mutation probability of traditional genetic algorithm to obtain the modal number and penalty parameters of optimal variational mode decomposition,and then realize the optimal decomposition of gyroscope output signal;The sample entropy is introduced to classify the various modes of the gyroscope,the noise modes are removed directly,the remaining modes are processed by time-frequency peak filtering,and the processed modes are reconstructed,effectively realize the error processing of the gyroscope.(4)Aiming at the problem of large accelerometer error caused by vibration and other signals in MWD,the accelerometer measurement signal model is established.With the help of the characteristics of high short-time precision of gyroscope,kalman filter is used to process the vibration signals under white noise.For the vibration signals under colored noise,the polynomial autoregression model is used to transform them into white noise,and then kalman filter is used to process the vibration signals;At the same time,the improved wavelet is proposed to process the shock signal,effectively realize the error processing of the accelerometer.(5)Aiming at the problem of large attitude error in complex environment while drilling,considering that the gyroscope drifts with time and the accelerometer is stable for a long time,the distributed information fusion of sensors is used to establish the PI control model of azimuth drift under vibration;A new neural network based on function approximation theory is designed,and OWDD algorithm is introduced to quickly determine the weight of neural network,and PI control parameters are adjusted in real time,so as to compensate azimuth drift adaptively;Further use the UKF to fuse the accelerometer and gyroscope data,effectively achieve attitude error compensation.53 pictures,4 tables,68 references.