CN104535063A - Geographic coordinate compensation method for seabed oil-gas pipeline detection positioning system - Google Patents

Abstract

The invention belongs to the seabed oil-gas pipeline detection positioning field, and concretely relates to a geographic coordinate compensation method for a seabed oil-gas pipeline detection positioning system. The method comprises employing a metal weld quantity detected by a current vortex sensor to determine a system-located steel pipe segment, so as to realize coarse positioning; employing a mileage gauge/ultralow-frequency electromagnetic wave correction method to calculate a strapdown inertial navigation system positioned by the pipeline and perform error compensation, taking mileage increment and an ultralow-frequency electromagnetic wave positioning signal as observed value, taking state error as a state variable, and employing Kalman filtering to estimate the state error; and combining a quadratic Bayes curve method and a forward and backward filtering algorithm to reduce the maximum error. The provided method is applicable to precise positioning of any pipeline detection positioning system based on a strapdown inertial navigation system, is capable of substantially reducing the accumulation effect of strapdown inertial navigation system error along with time when single filtering is performed, and is capable of reducing positioning error to one fourth of the original positioning error.

Description

The compensation method of a kind of sea-bottom oil-gas pipeline detection and location systematic geography coordinate

Technical field

The invention belongs to sea-bottom oil-gas pipeline detection and location field, be specifically related to the compensation method of a kind of sea-bottom oil-gas pipeline detection and location systematic geography coordinate.

Background technology

Oil-gas pipeline is the quick and the most most economical means realizing all kinds of hydrocarbon resources long distance transportation.But China's most of oil feed channel is on active service and been has all has been met or exceeded its military service phase, becomes a large hidden danger of current oil and gas pipes.And the leakage of hydrocarbon resources, blast etc. has not only had a strong impact on the Mining Transport of hydrocarbon resources, causes huge economic loss, also can cause environmental pollution, even cause eco-catastrophe.Therefore, the periodic detection of oil and gas pipes, safety assessment and on-call maintenance are to guaranteeing its safety, Effec-tive Function, and promote that oil & gas industry continues to develop at high speed, the tool that prevents the pollution of the environment is of great significance.

Pipe detection positioning system is primarily of detection system and positioning system composition.Positioning system be realize pipe detection key from the sixties in last century, American and Britain, De Deng state drop into multi-million dollar, continually develop out novel high-precision tube road defect detecting device, research and formulate the technical standard of pipe detection and safety evaluation, and by government formulate relevant law, regulation.Also there has been vicennial history in China to the research and development of all kinds of checkout equipment of oil and gas pipes, and part achievement in research is in the practical stage.But in pipe detection positioning system research gordian technique, blockage is kept, for the accurate Position Research realizing pipe detection system brings great difficulty to gordian technique abroad always.

Inertial navigation system divides gimbaled inertial navigation system and strap-down inertial navigation system two class.Gimbaled inertial navigation system because its volume is large, high in cost of production shortcoming is unsuitable in pipe detection positioning system application.The accurate location realizing pipe detection positioning system that appears as of strap-down inertial technology brings dawn.Strapdown inertial navigation system collective amass little, energy consumption is low, good stability, precision are high, low cost and other advantages in one, become the replacer of traditional platform formula inertial navigation system.

There is the inherent characteristic of accumulating in time in the positioning error of inertial navigation system, becomes the main factor of influential system positioning precision.Integrated navigation technology based on Kalman filter is the important method improving system accuracy, therefore intrinsic according to pipeline characteristic, and the integrated navigation and location technology that research is relevant is that the precision improving pipe detection positioning system is provided convenience.Simultaneously, according to the feature of pipe detection positioning system off-line data process, adopt the method estimating system positioning error curve of forward direction filtering and backward filtering respectively, studying two curve compensation methods is the emphasis directions realizing pipe detection positioning system hi-Fix at present.

All do not find to invent similar method introduction therewith in core periodical and patent consulting at present.

Summary of the invention

The object of the invention is can not realize the problem of long-distance oil & gas pipeline hi-Fix for single strapdown inertial navigation system and other assisted location methods in current pipe detection positioning system, propose a kind of oil and gas pipes positioning system geographic coordinate compensation method of the defects detection hi-Fix realized in long-distance oil & gas pipeline.

The object of the present invention is achieved like this:

(1) the metal welding seam number detected by employing current vortex sensor is to determine that the lengths of steel pipes residing for pipe detection system realizes coarse positioning;

(2) mileage gauge/strapdown inertial navigation system of ultra-low frequency electromagnetic wave bearing calibration to location of pipeline is adopted to calculate and carry out error compensation, mileage increment and ultra-low frequency electromagnetic wave positioning signal are as observed reading, using state error as state variable, use Kalman filter estimated state error;

(3) after setting up Kalman filter equation, by the attitude under Department of Geography, speed and site error, three axle gyroscopic drift errors and three axis accelerometer errors are as quantity of state X:

X＝[φ _xφ _yφ _zδv _nxδv _nyδv _nzδλ δL δh ε _xε _yε _z▽ _x▽ _y▽ _z]

φ _x, φ _y, φ _zfor the attitude error of pipe detection system; δ v _nx, δ v _ny, δ v _nzfor pipe detection system is in the velocity error of Department of Geography; δ λ, δ L, δ h are longitude, latitude and height error; ε _x, ε _y, ε _zfor gyroscopic drift error; ▽ _x, ▽ _y, ▽ _zfor accelerometer bias error;

(4) pipe detection system can be expressed as [0 v at carrier system medium velocity vector ₀0] ^t, under being transformed into navigation system be:

$V_{n0}=C_b^n\·{\left[\begin{array}{ccc}0& v_0& 0\end{array}\right]}^T$

In formula, v ₀mileage gauge speed; V _n0the speed of mileage gauge speed under navigation system;

Obtaining position according to ultra-low frequency electromagnetic wave is P _eLF, the speed calculated with inertial navigation and the difference of position are as observed reading:

Wherein, V _nand P _nthe speed that calculates of strapdown inertial navigation system and position.Δ v _i(i=x, y, z) is velocity error, Δ p _i(i=x, y, z) is site error;

(5) the discrete Kalman filter model of forward direction is set up according to quantity of state and observed quantity:

$\begin{array}{c}{X}_k=F_{k-1}{X}_{k-1}+G_{k-1}^w{W}_{k-1}\\ Z_k=H_k{X}_k+V_k\end{array}$

(6) prediction and the renewal of forward direction filtering is realized by evaluated error state:

${\hat{X}}_{f,k}^{-}=F_{k-1}{\hat{X}}_{f,k-1}^{+}+W_k$

$P_{f,k}^{-}=F_{k-1}{P}_{f,k-1}^{+}{F}_{k-1}^T+T_{k-1}{Q}_{k-1}{T}_{k-1}^T$

$K_{f,k}=P_{f,k}^{-}{H}_K^T{(H_k{P}_{f,k}^{-}{H}_K^T+R_k)}^{-1}$

${\hat{X}}_{f,k}^{+}={\hat{X}}_{f,k}^{-}+K_{f,k}(Z_k-H_k{\hat{X}}_{f,k}^{-})$

$P_{f,k}^{+}=(I-K_{f,k}{H}_k)P_{f,k}^{-}{(I-K_{f,k}{H}_k)}^T+K_{f,k}{R}_k{K}_{f,k}^T$

Obtain the attitude after the compensation of forward direction Kalman filter, position and velocity information;

(7) adopt Kalman filter algorithm dorsad to derive from current time k toward previous moment k-1, Kalman filter model is dorsad:

$\begin{array}{c}{\hat{X}}_{b,k-1}^{+}=F_{k-1}^{-1}{\hat{X}}_{b,k-1}^{-}+F_{k-1}^{-1}{W}_{k-1}\\ Z_{k-1}=H_{k-1}{\hat{X}}_{b,k-1}^{+}+V_{k-1}\end{array}$

By dynamic transfer transpose of a matrix:

$F_{b,k}=F_{k-1}^{-1}$

The prediction of Kalman filter and renewal equation dorsad:

${\hat{X}}_{b,k}^{-}=F_{b,k}{\hat{X}}_{b,k}^{-}$

$P_{b,k-1}^{-}=F_{b,k}(P_{b,k-1}^{+}+T_{k-1}{Q}_{k-1}{T}_{k-1}^T)F_{b,k}^T$

$K_{b,k-1}=P_{b,k-1}^{-}{H}_{k-1}^T{(H_{k-1}{P}_{b,k-1}^{-}{H}_{k-1}^T+R_{k-1})}^{-1}$

${\hat{X}}_{b,k-1}^{+}={\hat{X}}_{b,k-1}^{-}+K_{b,k-1}(Z_{k-1}-H_{k-1}{\hat{X}}_{b,k-1}^{-})$

$P_{b,k-1}^{+}=(I-K_{b,k-1}{H}_{k-1})P_{b,k-1}^{-}{(I-K_{b,k-1}{H}_{k-1})}^T+K_{b,k-1}{R}_{k-1}{K}_{b,k-1}^T$

(8) method of secondary Bayes curve reduces maximum error in conjunction with forward and backward filtering algorithm:

(8.1) forward direction filtering geographical line estimated value is resolved.

(8.2) backward filtering geographical line estimated value is obtained.

(8.3) find respectively a bit on two curves, make the two nearest, then find the intermediate point p of 2 ₁.

(8.4) on two curves, a pair p is found respectively ₀and p ₂, and have a segment distance with two closest approaches.

(8.5) by p ₀, p ₁and p ₂build secondary Bayes curve, be shown below:

p _Bezier(t)＝(1-t) ²p ₀+2(1-t)tp ₁+t ²p ₂,t∈[0,1]；

(8.6) geographical compensated curve is built:

$p^n=\left\{\begin{array}{cc}{p}_{\mathrm{Forward\; Estimation}}& [p_{\mathrm{start}},p_0]\\ p_{\mathrm{Bezier}}& [p_0,p_1]\\ p_{\mathrm{Backward\; Estimation}}& [p_1,p_{\mathrm{End}}]\end{array}\right..$

Beneficial effect of the present invention is:

The Technique of Subsea Pipeline Inspection positioning system geographic coordinate compensation method that this patent proposes is suitable for the accurate location of any pipe detection positioning system based on strapdown inertial navigation system; By the mode combined in conjunction with forward direction Kalman filter and backward Kalman filter, the effect that when greatly can reduce single filtering, strapdown inertial navigation system error is accumulated in time, can make positioning error be reduced to original 1/4; In forward direction Kalman filter and rear on the basis of Kalman filter, then in conjunction with the geographic coordinate backoff algorithm of this patent based on secondary Bayes curve, system positioning error can be reduced further; The geographic coordinate backoff algorithm adopted, greatly reduces system positioning error, can provide accurate position for the defect location of long-distance oil & gas pipeline; Compared with original detection and location system, the mode adopting strapdown inertial navigation system and other modes to combine realizes accurate location, and use forward direction Kalman filter and backward Kalman filter to combine, realize the fine compensation of systematic geography coordinate with secondary Bayes curve.The positioning precision of system long-distance oil & gas pipeline greatly can be improved from the angle of software approach.

Accompanying drawing explanation

Fig. 1 sea-bottom oil-gas pipeline detection and location system architecture diagram;

Fig. 2 is based on the mileage gauge/ultra-low frequency electromagnetic wave correcting algorithm block diagram of Kalman filter;

Fig. 3 geographic coordinate compensates schematic three dimensional views;

Fig. 4 geographic coordinate compensates two-dimensional effects figure;

Fig. 5 sea-bottom oil-gas pipeline detection and location system cloud gray model process flow diagram.

Embodiment

Below in conjunction with accompanying drawing citing, patent of the present invention is described in more detail, it should be noted that gyroscope that this system uses, accelerometer, pipeline defect detecting system, ultra-low frequency electromagnetic wave location, mileage gauge and system, control circuit are typical device and are connected with circuit therefore are no longer described its schematic diagram:

First the metal welding seam number detected by employing current vortex sensor is to determine that the lengths of steel pipes residing for pipe detection system realizes coarse positioning.Then mileage gauge/the strapdown inertial navigation system of ultra-low frequency electromagnetic wave bearing calibration to location of pipeline is adopted to calculate and carry out error compensation, mileage increment and ultra-low frequency electromagnetic wave positioning signal are as observed reading, using state error as state variable, use Kalman filter estimated state error.

After setting up Kalman filter equation, by the attitude under Department of Geography, speed and site error, three axle gyroscopic drift errors and three axis accelerometer errors are as quantity of state X:

X＝[φ _xφ _yφ _zδv _nxδv _nyδv _nzδλ δL δh ε _xε _yε _z▽ _x▽ _y▽ _z] (1)

In formula, φ _x, φ _y, φ _zfor the attitude error of pipe detection system; δ v _nx, δ v _ny, δ v _nzfor pipe detection system is in the velocity error of Department of Geography; δ λ, δ L, δ h are longitude, latitude and height error; ε _x, ε _y, ε _zfor gyroscopic drift error; ▽ _x, ▽ _y, ▽ _zfor accelerometer bias error.

Pipe detection system can only travel forward along pipeline, cannot move to side and top.Pipe detection system is identical with mileage gauge speed in carrier system y-axis direction, and the direction of x-axis and z-axis is zero, therefore pipe detection system can be expressed as [0 v at carrier system medium velocity vector ₀0] ^t, under being transformed into navigation system be:

$V_{n0}=C_b^n\·{\left[\begin{array}{ccc}0& v_0& 0\end{array}\right]}^T---\left(2\right)$

In formula, v ₀mileage gauge speed; V _n0the speed of mileage gauge speed under navigation system.Meanwhile, obtaining position according to ultra-low frequency electromagnetic wave is P _eLF, the speed calculate the two and inertial navigation and the difference of position are as observed reading:

$Z=\left[\begin{array}{c}{V}_n-V_{n0}\\ P_n-P_{\mathrm{ELF}}\end{array}\right]=\left[\begin{array}{ccc}{\mathrm{\Δv}}_x& {\mathrm{\Δv}}_y& {\mathrm{\Δv}}_z\\ {\mathrm{\Δp}}_x& {\mathrm{\Δp}}_y& {\mathrm{\Δp}}_z\end{array}\right]---\left(3\right)$

Wherein, V _nand P _nthe speed that calculates of strapdown inertial navigation system and position.Δ v _i(i=x, y, z) is velocity error, Δ p _i(i=x, y, z) is site error.

Then, the discrete Kalman filter model of forward direction is set up according to quantity of state and observed quantity:

$\begin{array}{c}{X}_k=F_{k-1}{X}_{k-1}+G_{k-1}^w{W}_{k-1}\\ Z_k=H_k{X}_k+V_k\end{array}---\left(4\right)$

Next, prediction and the renewal of forward direction filtering is realized by evaluated error state:

${\hat{X}}_{f,k}^{-}=F_{k-1}{\hat{X}}_{f,k-1}^{+}+W_k---\left(5\right)$

$P_{f,k}^{-}=F_{k-1}{P}_{f,k-1}^{+}{F}_{k-1}^T+T_{k-1}{Q}_{k-1}{T}_{k-1}^T---\left(6\right)$

$K_{f,k}=P_{f,k}^{-}{H}_K^T{(H_k{P}_{f,k}^{-}{H}_K^T+R_k)}^{-1}---\left(7\right)$

${\hat{X}}_{f,k}^{+}={\hat{X}}_{f,k}^{-}+K_{f,k}(Z_k-H_k{\hat{X}}_{f,k}^{-})---\left(8\right)$

$P_{f,k}^{+}=(I-K_{f,k}{H}_k)P_{f,k}^{-}{(I-K_{f,k}{H}_k)}^T+K_{f,k}{R}_k{K}_{f,k}^T---\left(9\right)$

The attitude after the compensation of forward direction Kalman filter, position and velocity information can be obtained like this.

But the navigation information that simple forward direction Kalman filter method obtains can be dispersed along with the time, is unfavorable for the accurate location of long-distance oil & gas pipeline detection system.Defect location after completing whole pipe detection task, need off-line to store given data process.Adopt Kalman filter algorithm dorsad to derive toward previous moment k-1 from current time k, mainly transposition is carried out to the dynamic matrix in formula (5).Kalman filter model is dorsad:

$\begin{array}{c}{\hat{X}}_{b,k-1}^{+}=F_{k-1}^{-1}{\hat{X}}_{b,k-1}^{-}+F_{k-1}^{-1}{W}_{k-1}\\ Z_{k-1}=H_{k-1}{\hat{X}}_{b,k-1}^{+}+V_{k-1}\end{array}---\left(10\right)$

Conveniently, dynamic transfer transpose of a matrix is expressed as:

$F_{b,k}=F_{k-1}^{-1}---\left(11\right)$

In order to realize the optimal estimating to error state, the prediction of Kalman filter and renewal equation are dorsad:

${\hat{X}}_{b,k}^{-}=F_{b,k}{\hat{X}}_{b,k}^{-}---\left(12\right)$

$P_{b,k-1}^{-}=F_{b,k}(P_{b,k-1}^{+}+T_{k-1}{Q}_{k-1}{T}_{k-1}^T)F_{b,k}^T---\left(13\right)$

$K_{b,k-1}=P_{b,k-1}^{-}{H}_{k-1}^T{(H_{k-1}{P}_{b,k-1}^{-}{H}_{k-1}^T+R_{k-1})}^{-1}---\left(14\right)$

${\hat{X}}_{b,k-1}^{+}={\hat{X}}_{b,k-1}^{-}+K_{b,k-1}(Z_{k-1}-H_{k-1}{\hat{X}}_{b,k-1}^{-})---\left(15\right)$

$P_{b,k-1}^{+}=(I-K_{b,k-1}{H}_{k-1})P_{b,k-1}^{-}{(I-K_{b,k-1}{H}_{k-1})}^T+K_{b,k-1}{R}_{k-1}{K}_{b,k-1}^T---\left(16\right)$

The position that pipe detection system enters and takes out obtains by GPS, and therefore starting point and ending point position is known.Can the error that dip of high degree by the reckoning of forward direction filtering algorithm and backward filtering algorithm.Positioning error can be reduced to original 1/4 by using the closest approach of forward and backward filter result.The method proposing secondary Bayes curve herein in conjunction with forward and backward filtering algorithm to reduce maximum error, this geographical backoff algorithm step is as follows:

(1) forward direction filtering geographical line estimated value is resolved.

(2) backward filtering geographical line estimated value is obtained.

(3) find respectively a bit on two curves, make the two nearest, then find the intermediate point p of 2 ₁.

(4) on two curves, a pair p is found respectively ₀and p ₂, and have a segment distance with two closest approaches.

(5) by p ₀, p ₁and p ₂build secondary Bayes curve, be shown below:

p _Bezier(t)＝(1-t) ²p ₀+2(1-t)tp ₁+t ²p ₂,t∈[0,1] (17)

(6) geographical compensated curve is built as follows:

$p^n=\left\{\begin{array}{cc}{p}_{\mathrm{Forward\; Estimation}}& [p_{\mathrm{start}},p_0]\\ p_{\mathrm{Bezier}}& [p_0,p_1]\\ p_{\mathrm{Backward\; Estimation}}& [p_1,p_{\mathrm{End}}]\end{array}\right.---\left(18\right)$

Composition graphs 1, sets forth this sea-bottom oil-gas pipeline detection and location system architecture diagram: this systemic-function mainly comprises sea-bottom oil-gas pipeline defect detecting system and sea-bottom oil-gas pipeline detection and location system two large divisions.Wherein sea-bottom oil-gas pipeline defect detecting system is all kinds of defects realizing tube wall, corrosion, effective detection of the correlation parameters such as crackle.Because it is not emphasis of the present invention, it is not described in detail.And intermediate portions is sea-bottom oil-gas pipeline detection and location systems, primarily of electromagnetic eddy sensor, strapdown inertial navigation system, mileage gauge, ultra-low frequency electromagnetic wave location four major part composition.Electromagnetic eddy sensor is used for counting the circular weld of seabed steel pipe, thus position in the pipeline determining roughly the current place of pipe detection system.Then, in single whole pipeline, the geographic coordinate values resolving output with strapdown inertial navigation system realizes the accurate location in pipeline for core.Simultaneously, disperse in time when considering the error of simple inertial navigation system, therefore the velocity information adopting mileage gauge to export uses Kalman filter to revise to the velocity amplitude that inertial navigation system calculates, and improves the precision that strapdown inertial navigation system speed exports.And the SLF electromagnetic wave receiving device installed regularly determines the position of pipe detection system in the ultra-low frequency electromagnetic wave that water surface pipeline maintenance ship is launched and pipe detection positioning system.Use this positional information to adopt Kalman filter to revise to the position of strapdown inertial navigation system, thus improve the position output accuracy of pipe detection system.What deserves to be explained is, the output data of pipe detection positioning system and pipeline defect detecting system are all be kept in the mass data storage hard disk entrained by system, all data calculation process are all adopt the mode of off-line to adopt computing machine to process, and obtain defect information and the defect particular location in the duct of pipeline.

Composition graphs 2, to be described in detail in pipeline accurately in position fixing process, and mileage gauge and ultra-low frequency electromagnetic wave are to the correction problem of strapdown inertial navigation system output speed information and positional information.Connect firmly the strapdown inertial navigation system (SINS) be made up of three-axis gyroscope and three axis accelerometer in pipe detection system and can calculate the current navigation information (attitude of pipe detection system by the mode of navigation calculation, speed, position).But due to the effect that inertial navigation system error is inherently accumulated in time, the time that navigational system is run is longer, and the error of the navigation information exported will be larger.Therefore need to adopt other utility appliance to reduce the effect of this accumulation of error.According to the feature of navigating in pipeline, the velocity information that mileage gauge exports becomes the first-selection that aided inertial navigation system carries out velocity error correction.Meanwhile, in Ocean Oil And Gas Pipeline detects, the accurate location of pipe detection system in some position of pipeline is realized by carrying out SLF electromagnetic wave communication between pipe testing apparatus and pipeline maintenance ship.The position output information of strapdown inertial navigation system is revised by the mode of Kalman filter.Therefore, in Ocean Oil And Gas Pipeline detection and location system, realize being suitable for marine environment and in long-distance oil & gas pipeline, realize accurate location, electromagnetic eddy sensor is adopted to realize coarse localization, strapdown inertial navigation system, mileage gauge and ultra-low frequency electromagnetic wave realize pinpoint mode can provide good method for the location of oil and gas pipes defect.

Composition graphs 3,4, on the system hardware composition and navigator fix basis of Fig. 2, by the mode in conjunction with forward direction Kalman filter (i.e. Fig. 2 introduction) and backward Kalman filter, reduces the positioning error of system.Further, have employed a kind of method being suitable for the secondary Bayes curve of Project Realization and carry out the poorest curve of matching two positioning error curve at closest approach.Can find from the geographical error compensation schematic three dimensional views of Fig. 3, generally forward location curve and backward auditory localization cues are disjoint in three dimensions, but any can be found in two curves respectively to make two curve distances nearest, namely as the intersection point in Fig. 4 two-dimensional curve.The error of single forward direction kalman filtering auditory localization cues can be made in this way greatly to reduce, improve the positioning precision of system.But, in order to improve the precision of strapdown inertial navigation system in oil and gas pipes detection and location system further, can compensate at forward location graph of errors and the rear closest approach place to positioning error curve further, both reductions curve, at the error amount of closest approach, improves the final positioning precision of system.For this reason, the present invention intends adopting secondary Bayes curve to carry out the system positioning error at matching closest approach place.

Concrete steps are:

Step 1, estimates navigational system forward location curve by forward direction kalman filtering, as shown in Figure 3, enters step 2;

Step 2, to estimate after navigational system to auditory localization cues by backward kalman filtering, as shown in Figure 3, enters step 3;

Step 3, at forward location curve and the rear forward direction closest approach and rear to closest approach of finding respectively on auditory localization cues, and the intermediate point p of 2 ₁, enter step 4;

Step 4, forward location curve finds 1 p from forward direction closest approach one segment distance place ₀, on auditory localization cues, find 1 p from rear to closest approach one segment distance place rear ₂, enter step 5;

Step 5, with p ₀, p ₁and p ₂build secondary Bayes curve, shown in (17), enter step 6;

Step 6, according to the secondary Bayes geographic coordinate compensated curve built, builds complete geographic coordinate curve.From initial time to p ₀the auditory localization cues that point adopts forward direction kalman filtering to estimate, from p ₀point is to p ₂point adopts the secondary Bayes geographic coordinate compensated curve built in step 5, from p ₂point adopts backward kalman filtering to estimate auditory localization cues to terminal.Shown in final two-dimensional localization curve tendency as minimum in Fig. 4.

Composition graphs 5, is described in detail to the operational scheme of whole sea-bottom oil-gas pipeline detection and location system: this patent is to realize hi-Fix in the long-distance oil & gas pipeline defects detection of seabed for target.First adopt electromagnetic eddy sensor to realize the counting of defect detecting system to steel pipe circular weld number, realize the coarse localization of pipeline inspection system in pipeline.Next, adopt based on strapdown inertial navigation system, and test the speed with mileage gauge and ultra-low frequency electromagnetic wave location, the mode that kalman filtering is estimated revises every error of strapdown inertial navigation system, improves navigational system in the location of long-distance oil & gas pipeline.When strapdown inertial navigation system auditory localization cues is estimated in kalman filtering, have employed the mode of forward direction kalman filtering method and the combination of backward kalman filtering method.Simultaneously at forward location curve and rear to auditory localization cues junction, have employed the combination that secondary Bayes curve carrys out matching two curve, compensate the positioning error of simple closest approach line compensation way.The operational scheme of this system is as follows:

Step 1, system electrification completes the navigation initial alignment process under initialization and static condition, for real-time navigation provides initial baseline, enters step 2;

Step 2, travels through one time by whole sea-bottom oil-gas pipeline detection and location system in pipeline, completes the preservation of detection to defect and navigation data, enters step 3;

Step 3, coarse positioning.First by the counting of electromagnetic eddy sensor determine pipe inspection positioning system be engraved in time each pipeline place rough position, and the line sections at the defect of correspondence, enters step 4;

Step 4, strapdown inertial navigation system is resolved to the navigation information of output, mileage gauge output speed information and ultra-low frequency electromagnetic wave output position information adopt forward direction kalman filtering and backward kalman filtering to estimate the forward location curve of navigational system and backward auditory localization cues respectively, enter step 5;

Step 5, adopts secondary Bayes curve to realize geographic coordinate to forward location curve and backward auditory localization cues and compensates, construct complete pipe detection positioning system auditory localization cues, enter step 6;

Step 6, in conjunction with the defect of pipeline information of pipeline detection system, realizes pipe curve information position display in the duct.

Claims (1)

1. a sea-bottom oil-gas pipeline detection and location systematic geography coordinate compensation method, is characterized in that:

(1) the metal welding seam number detected by employing current vortex sensor is to determine that the lengths of steel pipes residing for pipe detection system realizes coarse positioning;

(2) mileage gauge/strapdown inertial navigation system of ultra-low frequency electromagnetic wave bearing calibration to location of pipeline is adopted to calculate and carry out error compensation, mileage increment and ultra-low frequency electromagnetic wave positioning signal are as observed reading, using state error as state variable, use Kalman filter estimated state error;

(3) after setting up Kalman filter equation, by the attitude under Department of Geography, speed and site error, three axle gyroscopic drift errors and three axis accelerometer errors are as quantity of state X:

$X=\left[\begin{array}{ccccccccccccccc}{\mathrm{\φ}}_x& {\mathrm{\φ}}_y& {\mathrm{\φ}}_z& {\mathrm{\δv}}_{\mathrm{nx}}& {\mathrm{\δv}}_{\mathrm{ny}}& {\mathrm{\δv}}_{\mathrm{nz}}& \mathrm{\δ\λ}& \mathrm{\δL}& \mathrm{\δh}& {\mathrm{\ϵ}}_x& {\mathrm{\ϵ}}_y& {\mathrm{\ϵ}}_z& {\▿}_x& {\▿}_y& {\▿}_z\end{array}\right]$

φ _x, φ _y, φ _zfor the attitude error of pipe detection system; δ v _nx, δ v _ny, δ v _nzfor pipe detection system is in the velocity error of Department of Geography; δ λ, δ L, δ h are longitude, latitude and height error; ε _x, ε _y, ε _zfor gyroscopic drift error; for accelerometer bias error;

(4) pipe detection system can be expressed as [0 v at carrier system medium velocity vector ₀0] ^t, under being transformed into navigation system be:

$V_{n0}=C_b^n\·{\left[\begin{array}{ccc}0& v_0& 0\end{array}\right]}^T$

In formula, v ₀mileage gauge speed; V _n0the speed of mileage gauge speed under navigation system;

Obtaining position according to ultra-low frequency electromagnetic wave is P _eLF, the speed calculated with inertial navigation and the difference of position are as observed reading:

Wherein, V _nand P _nthe speed that calculates of strapdown inertial navigation system and position, Δ v _i(i=x, y, z) is velocity error, Δ p _i(i=x, y, z) is site error;

(5) the discrete Kalman filter model of forward direction is set up according to quantity of state and observed quantity:

$\begin{array}{c}{X}_k=F_{k-1}{X}_{k-1}+G_{k-1}^w{W}_{k-1}\\ Z_k=H_k{X}_k+V_k\end{array};$

(6) prediction and the renewal of forward direction filtering is realized by evaluated error state:

${\hat{X}}_{f,k}^{-}=F_{k-1}{\hat{X}}_{f,k-1}^{+}+W_k$

$P_{f,k}^{-}=F_{k-1}{P}_{f,k-1}^{+}{F}_{k-1}^T+T_{k-1}{Q}_{k-1}{T}_{k-1}^T$

$K_{f,k}=P_{f,k}^{-}{H}_K^T{(H_k{P}_{f,k}^{-}{H}_K^T+R_k)}^{-1}$

${\hat{X}}_{f,k}^{+}={\hat{X}}_{f,k}^{-}+K_{f,k}(Z_k-H_k{\hat{X}}_{f,k}^{-})$

$P_{f,k}^{+}(I-K_{f,k}{H}_k)P_{f,k}^{-}{(I-K_{f,k}{H}_k)}^T+K_{f,k}{R}_k{K}_{f,k}^T$

Obtain the attitude after the compensation of forward direction Kalman filter, position and velocity information;

(7) adopt Kalman filter algorithm dorsad to derive from current time k toward previous moment k-1, Kalman filter model is dorsad:

${\hat{X}}_{b,k-1}^{+}=F_{k-1}^{-1}{\hat{X}}_{b,k-1}^{-}+F_{k-1}^{-1}{W}_{k-1}$

$Z_{k-1}=H_{k-1}{\hat{X}}_{b,k-1}^{+}+V_{k-1}$

By dynamic transfer transpose of a matrix:

$F_{b,k}=F_{k-1}^{-1}$

The prediction of Kalman filter and renewal equation dorsad:

${\hat{X}}_{b,k}^{-}=F_{b,k}{\hat{X}}_{b,k}^{-}$

$P_{b,k-1}^{-}=F_{b,k}(P_{b,k-1}^{+}+T_{k-1}{Q}_{k-1}{T}_{k-1}^T)F_{b,k}^T$

$K_{b,k-1}=P_{b,k-1}^{-}{H}_{k-1}^T{(H_{k-1}{P}_{b,k-1}^{-}{H}_{k-1}^T+R_{k-1})}^{-1}$

${\hat{X}}_{b,k-1}^{+}={\hat{X}}_{b,k-1}^{-}+K_{b,k-1}(Z_{k-1}-H_{k-1}{\hat{X}}_{b,k-1}^{-})$

$P_{b,k-1}^{+}=(I-K_{b,k-1}{H}_{k-1})P_{b,k-1}^{-}{(I-K_{b,k-1}{H}_{k-1})}^T+K_{b,k-1}{R}_{k-1}{K}_{b,k-1}^T;$

(8) method of secondary Bayes curve reduces maximum error in conjunction with forward and backward filtering algorithm:

(8.1) forward direction filtering geographical line estimated value is resolved;

(8.2) backward filtering geographical line estimated value is obtained;

(8.3) find respectively a bit on two curves, make the two nearest, then find the intermediate point p of 2 ₁;

(8.4) on two curves, a pair p is found respectively ₀and p ₂, and have a segment distance with two closest approaches;

(8.5) by p ₀, p ₁and p ₂build secondary Bayes curve:

p _Bezier(t)＝(1-t) ²p ₀+2(1-t)tp ₁+t ²p ₂,t∈[0,1]；

(8.6) geographical compensated curve is built:

$p^n=\left\{\begin{array}{cc}{p}_{\mathrm{Forward\; Estimation}}& [p_{\mathrm{start}},p_0]\\ p_{\mathrm{Bezier}}& [p_0,p_1]\\ p_{\mathrm{Backward\; Estimation}}& [p_1,p_{\mathrm{End}}]\end{array}\right..$