Experimental device for verifying positioning accuracy of INS-DR (inertial navigation system-digital radiography) combined pipeline navigation algorithm

Abstract

The invention discloses an experimental device for verifying the positioning accuracy of an INS-DR combined pipeline navigation algorithm, and belongs to the technical field of simulation experiments. Comprising a remote control trolley provided with an inertial measurement unit, an encoder and an upper computer, the inertial measurement unit is used for acquiring motion data generated in the operation process of the remote control trolley and uploading the motion data to the upper computer; the encoder is used for acquiring odometer wheel data generated in the running process of the remote control trolley and uploading the odometer wheel data to the upper computer; the upper computer loads motion data of the remote control trolley and odometer wheel data to an INS-DR combined pipeline navigation algorithm program, and motion track data is obtained through calculation; and comparing the actual motion trail data of the remote control trolley with the motion trail data obtained by resolving, thereby realizing the precision evaluation of the INS-DR combined pipeline navigation algorithm. According to the invention, the motion track of the pipeline detector can be obtained through offline calculation based on the motion information of the detector, and oil and gas pipeline leakage detection and positioning are realized.

Description

Technical Field

[0001] The present invention relates to an experimental device, in particular to an experimental device for verifying the positioning accuracy of an INS-DR combined pipeline navigation algorithm. Background Art

[0002] The petroleum industry is a key sector in the nation's economic development. Pipeline transportation, with its advantages of high capacity, high stability, safety, and low cost, has become a primary method of transporting oil and gas resources. As the mileage of oil and gas pipelines continues to increase, submarine oil and gas pipelines, in particular, face risks such as corrosion, deformation, and leakage due to their long-term service in the harsh marine environment. These risks not only pose safety risks but can also cause severe damage to the marine ecosystem and result in significant economic losses. Therefore, timely leak detection of offshore oil and gas pipelines is crucial for preventing and responding to major marine accidents.

[0003] At present, the detection and positioning technologies in submarine oil and gas pipelines mainly include the following:

[0004] Negative pressure wave method: It is more sensitive to larger pipeline leaks, but cannot perceive subtle negative pressure waves, is insensitive to micro leaks, and is easily affected by various factors, resulting in false alarms.

[0005] Pressure gradient method: It requires multiple pressure detection locations on the pipeline, which is difficult to achieve on site. The positioning accuracy is affected by the accuracy of the instrument. It is often used as an auxiliary detection method.

[0006] Magnetic flux leakage detection method: It can detect girth weld defects, circumferential defects and axial defects in pipelines, but it is not sensitive to minor defects and requires a pipe cleaning process before detection, which results in high detection costs.

[0007] Ultrasonic method: locates pipelines and detects corrosion damage by detecting the propagation speed and reflection of sound waves in the pipeline. It has high accuracy and can detect tiny defects and cracks, but it needs to contact the pipeline surface, which can easily cause probe wear and damage to the pipeline surface.

[0008] Acoustic internal detection: This method utilizes leak noise for detection. When an internal detector passes the leak location, it picks up the noise and uses a mileage wheel and girth weld recognition system to achieve detection and positioning. This method relies on leak noise detection and may be affected by ambient noise, affecting detection accuracy.

[0009] In addition, combined navigation technology using inertial navigation systems (INS) and dead reckoning (DR) is also being applied to pipeline inspection. INS is independent of external information and unaffected by climate and external factors, continuously providing real-time information on the carrier's position, attitude, and velocity. DR uses attitude, heading, and mileage information to infer the carrier's relative position. INS-DR combined navigation has become the preferred option for low-cost autonomous navigation systems. However, due to the current lack of experimental conditions for validating navigation algorithms on actual oil and gas pipelines, the accuracy of existing INS-DR combined navigation algorithms cannot be verified, limiting their practical application. Summary of the Invention

[0010] In light of the shortcomings of existing technologies, this paper provides an experimental device for verifying the positioning accuracy of an INS-DR combined pipeline navigation algorithm. This device employs a sports car experiment to construct a test system that simulates an oil and gas pipeline environment. By collecting signals from the sports car, the accuracy of the navigation and positioning algorithm is verified. By restoring the motion trajectory of the sports car, the device can achieve navigation and positioning for pipeline robots in practical applications.

[0011] The technical means adopted in the present invention are as follows:

[0012] An experimental device for verifying the positioning accuracy of an INS-DR combined pipeline navigation algorithm includes: a remote-controlled vehicle equipped with an inertial measurement unit, an encoder, and a host computer. Single-sided sponge tape is placed at regular intervals along the vehicle's path to simulate welds. The host computer is used to run the INS-DR combined pipeline navigation algorithm program, which is used to calculate the vehicle's motion trajectory based on the vehicle's motion data.

[0013] The inertial measurement unit is used to obtain motion data generated by the remote control car during operation and upload it to the host computer;

[0014] The encoder is used to obtain the odometer data generated by the remote control car during operation and upload it to the host computer;

[0015] The host computer loads the motion data and odometer data of the remote control car into the INS-DR combined pipeline navigation algorithm program, and obtains the motion trajectory data through calculation;

[0016] The upper computer compares the actual motion trajectory data of the remote control car with the motion trajectory data obtained by solution, thereby realizing the accuracy evaluation of the INS-DR combined pipeline navigation algorithm.

[0017] Furthermore, the experimental device also includes a microprocessor and a memory card installed on the remote control car. The memory card stores a data acquisition program. When the data acquisition program is running, the microprocessor collects data sent by the inertial measurement unit and the encoder.

[0018] Furthermore, when the data acquisition program is running, the inertial measurement unit and the encoder are controlled to output pulse signals at a sampling frequency of 100 Hz.

[0019] Furthermore, there are two encoders, which are respectively fixed to the two rear wheels of the remote-controlled car. During the operation of the remote-controlled car, the encoders and the wheels always rotate at the same speed.

[0020] Furthermore, the center point of the inertial measurement unit sensitive element is installed directly above the center point of the equivalent odometer wheel position, and the equivalent odometer wheel position is set to the center point of the line connecting the two rear wheels of the remote control car.

[0021] Furthermore, the motion data collected by the inertial measurement unit includes the vehicle body angular velocity output by the three-axis gyroscope and the vehicle body specific force output by the three-axis accelerometer.

[0022] Compared with the prior art, the present invention has the following advantages:

[0023] The present invention can compare the path calculated by the INS-DR navigation algorithm with the actual path through a sports car experiment, and intuitively verify the accuracy of the combined navigation algorithm.

[0024] The present invention uses a Kalman filter to adaptively adjust the weights of the odometer and the IMU, reducing errors and thus improving positioning accuracy, so that the integrated navigation system can more accurately determine the position and motion state of the vehicle.

[0025] The present invention is verified by experimental data of sports cars with multiple groups of different paths (such as L-path, S-path, rectangular path, and figure-8 path). The combined navigation algorithm can successfully reproduce the car path and obtain the car's speed, position, and posture. The state mean square error of the filter converges, proving the reliability and stability of the algorithm under different motion trajectories. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0027] Figure 1 This is a schematic diagram of the experimental device structure for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm of the present invention.

[0028] Figure 2 A comparison diagram of the path calculated by the existing method and the true path.

[0029] Figure 3 This is a flowchart of the INS-DR combined pipeline navigation algorithm execution in an embodiment of the present invention.

[0030] Figure 4 This is a main block diagram of the strapdown inertial navigation update algorithm in an embodiment of the present invention.

[0031] Figure 5 This is a comparison diagram of the path calculated by the integrated navigation system and the actual path in an embodiment of the present invention.

[0032] Figure 6 This is a graph showing the changes in the posture error, speed error, and position error of the vehicle in an embodiment of the present invention.

[0033] Figure 7 Schematic diagram of pipeline weld identification waveform in an embodiment of the present invention.

[0034] Figure 8 Schematic diagram of lever arm error in dead reckoning according to an embodiment of the present invention.

[0035] Figure 9 Schematic diagram of the sensor installation position in an embodiment of the present invention.

[0036] In the figure: 1. Host computer; 2. Microprocessor; 3. Memory card; 4. Center point of the sensitive element of the inertial measurement unit sensor; 5. Inertial measurement unit; 6. Left encoder; 7. Right encoder; 8. Left rear wheel; 9. Right rear wheel; 10. Equivalent odometer wheel. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0038] like Figure 1As shown, the present invention provides an experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm, comprising: a remote-controlled car equipped with an inertial measurement unit, an encoder, and a host computer. The host computer is used to run the INS-DR combined pipeline navigation algorithm program. The inertial measurement unit is used to obtain the motion data generated by the remote-controlled car during operation and upload it to the host computer. The encoder is used to obtain the odometer data generated by the remote-controlled car during operation and upload it to the host computer. The host computer loads the motion data and odometer data of the remote-controlled car into the INS-DR combined pipeline navigation algorithm program, and obtains the motion trajectory data through solution. The host computer compares the actual motion trajectory data of the remote-controlled car with the motion trajectory data obtained through solution, thereby realizing the accuracy evaluation of the INS-DR combined pipeline navigation algorithm.

[0039] This embodiment provides an experimental device for verifying the accuracy of the INS-DR integrated navigation algorithm. By installing various sensors on a remote-controlled vehicle, motion data during vehicle operation is obtained as input to the INS-DR integrated navigation algorithm. The raw data is then analyzed using an INS-DR integrated navigation program written in MATLAB to obtain the vehicle's speed, position, attitude, and path during operation.

[0040] The experimental device includes: a remote control car, a remote controller, an encoder, a microprocessor (STM32F103RCT6 single-chip microcomputer is selected in this embodiment), a memory card (SD card is preferably used in this embodiment), a card reader, an inertial measurement unit (IMU), a power supply and a host computer.

[0041] As a preferred embodiment of the present invention, this embodiment is equipped with two incremental encoders, whose shafts are respectively fixed to the two rear wheels of the trolley, forming the odometer. During the operation of the trolley, the encoder and the wheel can always be ensured to rotate at the same speed. The data output of the encoder is connected to the single-chip microcomputer; the microprocessor is powered by a 5V power supply, and the single-chip microcomputer has an SD card slot for installing an SD card. Code is written to make the encoder output a pulse signal at a sampling frequency of 100Hz. The pulse signal is then converted into the displacement increments of the two wheels and stored on the SD card, completing the odometer data acquisition. In this embodiment, the data of the inertial measurement unit is sent directly to the host computer, and the encoder data is directly stored on the memory card of the STM32 single-chip microcomputer.

[0042] The center point of the line connecting the two odometry wheels is the equivalent odometry wheel position. The equivalent odometry wheel is a hypothetical wheel structure whose imaginary position is the line connecting the points of tangency between the two wheels and the ground. The line connecting the points of tangency is chosen because it is inherently defined in the odometry wheel coordinate system in integrated navigation. The tangency point provides the coordinate origin of the odometry wheel coordinate system. Furthermore, the speeds measured by the two rear wheels cannot remain consistent during vehicle movement, so the average value is used as the actual vehicle speed. In other words, the equivalent odometry wheel is located at the center of the line connecting the points of tangency, and its speed is the average speed of the two rear wheels.

[0043] In order to reduce the influence of the arm error in the integrated navigation, when installing the IMU, the center point of the IMU sensitive element needs to be installed just above the center point of the equivalent odometer. This eliminates the influence of the arm error in the horizontal direction, leaving only a few millimeters of error in the vertical direction, which can be ignored. The output end of the IMU is connected to the host computer and outputs data at a sampling frequency of 100Hz. The host computer receives the IMU data in real time, which mainly includes the angular velocity (° / s) output by the three-axis gyroscope and the specific force (m / s) output by the three-axis accelerometer. 2 ) and stores it, and the IMU data collection is completed.

[0044] Since the integrated navigation algorithm has requirements for data format, the collected data is preprocessed, including equivalent odometer data and IMU data. According to the non-holonomic constraints, during the operation of the car, its left and right (X direction) and up and down (Z direction) speeds are 0m / s, and it only has front and back (Y direction) speeds. Therefore, the collected odometer data only represents the displacement increment of the car in the Y direction. The equivalent odometer data is the average value of the left and right odometers, and the format is [X; Y; Z], that is, [0; Y; 0]. The gyroscope and accelerometer inputs used in the integrated navigation algorithm are all incremental information (corresponding to the units of rad and m / s, respectively). Since the raw data are the vehicle body angular velocity and specific force information, the raw data are multiplied by the sampling interval (0.01s) for processing. In addition, the angle system of the gyroscope output is converted to the radian system. In this application, a 6-axis inertial measurement unit is used, which includes a 3-axis gyroscope and a 3-axis accelerometer.

[0045] Import the converted data into MATLAB and save it in MATLAB (matrix) format. Columns 1 through 8 contain the IMU three-axis angle increments, IMU three-axis velocity increments, equivalent odometry distance increments, and a timestamp. This data is then fed into the INS-DR integrated navigation algorithm, which calculates the vehicle's attitude angle changes, velocity changes, and trajectory.

[0046] The process of conducting the test by the device given in the embodiment of the present invention is as follows: Figure 3 As shown, it mainly includes the following steps.

[0047] 1. Connect the two encoders to the STM32F103RCT6 microcontroller and the IMU to the host computer.

[0048] 2. Power on the car and IMU.

[0049] 3. The MCU is powered on and starts recording encoder data and storing it in the SD card. The host computer starts and starts recording IMU data.

[0050] 4. The remote control is connected to the car through the Bluetooth module, and the car is controlled to move along an L-shaped path.

[0051] 5. After the L-shaped path is completed, press KEY2 on the MCU to stop recording the encoder data, remove the SD card, insert the card reader and upload it to the computer; operate the host computer to stop recording the IMU data and save the IMU data to the computer. Here, you can use the IMU official data acquisition system page as the host computer. Log in to the host computer page and click Start Acquisition to transfer the IMU data to the computer in real time and save it in EXCEL format.

[0052] 6. Data post-processing: align the timestamps of the encoder data with the IMU data; average the distance increment information measured by the two encoders to obtain the equivalent odometry distance increment; convert the angular velocity / specific force information measured by the IMU into incremental information (angle increment and velocity increment); integrate the IMU and odometry data into the same EXCEL file, with columns 1 to 8 of the table containing the IMU three-axis angle increment, IMU three-axis velocity increment, equivalent odometry distance increment, and timestamp, respectively.

[0053] 7. Open MATLAB to import data, save it in mat (numerical matrix) format, load the data in the navigation algorithm, and start the program to solve.

[0054] 8. Obtain the moving path of the car; obtain the change graph of the car's position, speed, and attitude over time; obtain the change graph of the combined navigation position error, speed error, and attitude error over time.

[0055] 9. Compare the path calculated by the combined navigation algorithm with the actual path to confirm the navigation positioning accuracy.

[0056] This example verifies the accuracy of a combined navigation algorithm through a race car experiment before inspecting an oil and gas pipeline. By comparing the path calculated by the INS-Odometer-Weld navigation algorithm with the actual path, various navigation parameters are successfully obtained and presented as charts in Matlab. An extended Kalman filter is used to adaptively adjust the weights of the odometer and IMU. Position observations are constructed by the difference between the inertial navigation-calculated position and the dead-reckoned position. Attitude observations are constructed by the difference between the pitch and heading angles calculated in real time by the inertial navigation system and the pitch and heading angles calculated by the inertial navigation system at the starting point of each straight pipe section.

[0057] By conducting an L-shaped sports car experiment in the corridor of the laboratory building, it was verified that the combined navigation algorithm can successfully reproduce the car's path, obtain the car's speed, position and posture, and successfully analyze the navigation error state. The state mean square error of the filter converged.

[0058] The first step is to use the IMU data to perform pure inertial navigation solution, which mainly includes attitude update algorithm, velocity update algorithm and position update algorithm.

[0059] (1) Posture update algorithm

[0060]

[0061] Where: and Respectively represent t m-1 and t m Quaternion of attitude transformation at the moment; It is from t m-1 to t m The attitude quaternion changes at each moment, and there is an angular velocity increment and

[0062] The above quaternion attitude update algorithm is based on an assumption: the moving coordinate system rotates along a fixed axis during the update cycle. If this is not the case, directly solving the quaternion using the angular increments will introduce non-commutative errors. To reduce the impact of this non-commutative error, the equivalent rotation vector is first solved using the angular increments, and then the quaternion is updated using this equivalent rotation vector.

[0063] For actual systems, it is easy to excite conical motion during angular motion, resulting in conical error. Therefore, it is necessary to study the equivalent rotation vector algorithm based on conical error compensation. In order to compensate for this error, a multi-sample compensation algorithm is usually used. m-1 ,t m ], the calculation formula of the equivalent rotation vector is:

[0064]

[0065] in, T=t m -t m-1 , represents the attitude quaternion update period; Δθ m is the sum of the angular increments within the time period T, Δθ mi is the angular increment of the i-th sub-sample in the time period T; k N-1 is the cone error compensation coefficient. The cone error compensation coefficients of different samples are shown in Table 1.

[0066] Table 1 Cone error compensation coefficients for different samples

[0067]

[0068] This time we choose twin sampling, then the equivalent rotation vector calculation formula is:

[0069]

[0070] Let φ(T) be abbreviated as φ m , by replacing the angular velocity increment in formula (2) with the equivalent rotation vector, we can obtain the attitude update algorithm that matches the equivalent rotation vector with the quaternion, that is, becomes

[0071]

[0072] The quaternion obtained by attitude update can be converted into Euler angles to obtain the pitch angle, roll angle and heading angle of the car.

[0073] (2) Speed ​​update algorithm

[0074] The inertial force equation is converted into m-1 ,t m ] The internal integration can be used to obtain the inertial navigation velocity update algorithm:

[0075]

[0076] Where: and t m-1 and t m Inertial velocity at the moment; and They are respectively called time periods T = t m -t m-1 The internal navigation system compares the velocity increment and the velocity increment of harmful acceleration.

[0077] The integrand is a time-slow variable, and the midpoint of the interval can be used as t m-1 / 2 =(t m-1 +t m) / 2 approximate calculation, so

[0078]

[0079] It is essentially the integration of two fast variables, which is finally sorted out as

[0080]

[0081] where Δv m is the accelerometer sampling specific force velocity increment within the time period T, It is called the rotation error compensation of velocity, which is caused by the rotation change of the specific force direction in space during the solution period. It is called the paddling error compensation, which can be obtained through a multi-sample algorithm. Here, a two-sample speed paddling error compensation algorithm is selected.

[0082] (3) Location update algorithm

[0083] The differential equation for the position (latitude, longitude, and altitude) of the strapdown inertial navigation system is as follows:

[0084]

[0085] Rewritten into matrix form:

[0086]

[0087] Where: v E 、v N and v U Represents the speed v n The east, north and sky components of v n =[v E v N v U ] T .

[0088] R Mh =R M +h, R Nh =R N +h, R M is the principal radius of curvature of the meridian, R N is the principal curvature radius of the ecliptic circle, L and h are the geographical latitude and altitude respectively.

[0089] By integrating the position differential equation (9), we can get the position update algorithm. Since the calculation error caused by the position update algorithm is generally small, the trapezoidal integration method can be used to discretize equation (10) to obtain

[0090]

[0091] The above is the update algorithm of strapdown inertial navigation, and its main calculation block diagram is as follows: Figure 4 shown

[0092] The comparison diagram between the calculated path and the actual path obtained by the above pure inertial navigation settlement method is as follows: Figure 5 As shown in the figure, the trajectory calculated by a single inertial navigation system deviates significantly from the actual trajectory, with an eastward deviation of 30.75m and a northward deviation of 25.12m, resulting in a positioning error of 39.46%. Therefore, using the inertial navigation system alone for a long time cannot achieve ideal positioning accuracy and must be used in conjunction with other sensors.

[0093] This embodiment will verify the accuracy of the SINS-Odometer-Weld integrated navigation system composed of the inertial navigation system, odometer wheel and girth weld. The specific logic block diagram is as follows: Figure 3 shown.

[0094] The system has all the update algorithms of pure inertial navigation solution, but adds auxiliary positioning of odometer wheel and girth weld to correct the position error (including velocity error) and attitude error in pure inertial navigation solution. The specific process is as follows:

[0095] Dead Reckoning (DR) is a navigation algorithm that uses a known initial position and measured data on the object's motion direction, speed, and time to gradually calculate the current position. The odometry wheel outputs the vehicle's distance increment within a sampling interval. The rear wheels of a vehicle are typically non-steering wheels, which always align with the vehicle's motion. Therefore, they are often used as odometry wheels.

[0096] The odometer measurement coordinate system (m system) satisfies the "right-front-up" right-hand rectangular coordinate system. Assuming that there is no jumping or side sliding during the movement of the car, according to the vehicle kinematic constraints (non-holonomic constraint principle), its right and vertical speeds are considered to be zero. Similarly, the distance increments in these two directions are also zero. Therefore, the odometer is in the time period [t j-1 ,t j ](T j =t j -t j-1 ) can be expressed as

[0097]

[0098] Where: ΔS j is the forward distance increment of the car measured by the odometer, v D Time period T j The average forward speed of the inner car.

[0099] In the actual installation process, there is an installation deviation angle from the M system to the B system, so the distance increment measured by the odometer wheel needs to be and average speed First convert to b system, then convert to n system. Assume α θ and α ψ are the pitch angle and azimuth angle respectively, then the odometer distance increment and average speed of system b are

[0100]

[0101] in: It is the attitude transformation matrix from the m system to the b system.

[0102] The velocity in the carrier coordinate system Attitude matrix calculated by strapdown inertial navigation Converted to the navigation coordinate system, we can get the position differential equation corresponding to the dead reckoning

[0103]

[0104] in: It is t j-1 The car attitude matrix obtained by strapdown inertial navigation at all times.

[0105] Correspondingly, the discretized position update algorithm is as follows:

[0106]

[0107] In the formula: are the displacement increments in the east, north and celestial directions respectively.

[0108] By calculating the deviation on both sides of equation (14), we can obtain the dead reckoning position error equation:

[0109]

[0110] in M pkD =M pvD M vkD ,

[0111] In dead reckoning, the attitude matrix calculated by pure inertial navigation is directly used to transform the coordinates of the odometer measurement to obtain the dead reckoning speed in the n-frame. In this case, the inertial navigation solution and dead reckoning use the same attitude matrix and have the same misalignment angle error. The inertial navigation error and dead reckoning error are combined together to form the following state vector:

[0112]

[0113] like Figure 8 As shown, since the odometer measurement coordinate system (m system) and the carrier coordinate system (b system) are not completely coincident, there is a lever arm between them. The arm error between the odometer wheel and the inertial navigation is

[0114]

[0115] Taking into account the influence of the arm error, when designing the carrier car, the center point of the IMU sensitive element is aligned with the center point of the equivalent odometer measurement. The distance deviation in the x-direction and y-direction is almost 0, and there is only a 9cm deviation in the z-direction, that is, dx = dy = 0. Figure 9 As shown. Since the horizontal attitude angle of the car is generally not large during driving, the influence of the lever arm in the height direction can be ignored. Therefore, for this car model, the influence of the lever arm effect can be ignored when performing integrated navigation. The observation quantity is constructed by the difference between the inertial navigation solution position and the dead reckoning position.

[0116]

[0117] The above is the position correction, and the difference between the inertial navigation solution position and the dead reckoning position constitutes the observation quantity. Therefore, the SINS / Odometer integrated navigation state space model is

[0118]

[0119] in:

[0120] H=[0 3×6 I 3×3 -I 3×3 0 3×9 ]; V is the position measurement noise.

[0121] 2. The pipeline is welded together using standard fittings, and the length of standard fittings is fixed. Each weld seam is assumed to have passed through a certain length of standard fittings. This can be used as a reference for coarse positioning and to correct the distance measurement of the odometer. At the same time, due to the characteristics of non-integrity constraints, the pitch and heading angles of the pipeline robot remain unchanged. Therefore, when passing through a weld seam, the pitch and heading angles at that moment can be used to correct the robot's posture in the entire subsequent straight pipe section.

[0122] After adding weld auxiliary positioning, the observation quantity composed of the robot pitch angle and heading angle is

[0123]

[0124] in pitchINS and yaw INS It is the pitch angle and heading angle calculated by the inertial navigation system in real time. J and yaw J is the pitch angle and heading angle calculated by the inertial navigation system at the starting position of each straight pipe section; V' is the measurement noise of the pitch angle and heading angle. Figure 3 shown.

[0125] In the figure, the extended Kalman filter (EKF) fuses and corrects the data of the carrier attitude, velocity, and position information calculated by the strapdown inertial navigation algorithm and the dead reckoning algorithm, and uses the accelerometer of the IMU to identify the pipeline weld. The specific identification scheme is: when the vehicle passes through the weld, the vehicle body will vibrate. The accelerometer of the IMU sensor on the vehicle is sensitive enough to detect these vibration signals. The specific force output by the accelerometer will jump, especially the accelerometers on the y-axis and z-axis. Because the distance between each weld is the same, the output of the accelerometer will jump periodically, such as Figure 7 shown.

[0126] Each time a weld is identified, the pitch angle and heading angle calculated by the strapdown inertial navigation solution at that time point are used as the observation value of the next straight pipe section to correct the attitude error; at the same time, each time a weld is identified, the total mileage Li is increased by a pipe section length l. At the same time, the time corresponding to the continuous identification of two welds is recorded as t and t'. Because the sampling interval of the odometer is 0.01s, each increment ΔSk ​​within the Δt time period is accumulated to obtain the total distance increment S Δt , assign L to S Δt Used to correct the measurement error of the odometer between two welds; similarly, the total mileage Li+1 obtained by the accelerometer identifying the weld is assigned to the total mileage S measured by the odometer to correct the measurement error of the odometer, and considering that the car is not at the weld when it stops, the distance between the last weld and the robot's stop position is added.

[0127] Finally, we get a comparison chart of the path calculated by the SINS-Odometer-Weld integrated navigation system and the actual path, as shown in the figure below: Figure 5 At the same time, the change diagrams of attitude error, velocity error and position error are obtained, as shown in Figure 6As can be seen from the figure, the trajectory calculated by the SINS / Odometer / Weld integrated navigation system has greatly improved its accuracy compared to the trajectory calculated by a single strapdown inertial navigation system. Its trajectory is more consistent with the actual trajectory, with an eastward deviation of 0.85m and a northward deviation of 0.82m. The positioning error is 1.2%, and the positioning accuracy is improved by 38.26%. In the experimental environment of 100 meters, the positioning accuracy can reach 98.8%, verifying the feasibility and accuracy of the SINS / Odometer / Weld integrated navigation system.

[0128] from Figure 6 It can be seen from the figure that the attitude error and velocity error converge rapidly. This is mainly because the odometer and IMU data are fused through the extended Kalman filter algorithm, and the attitude and velocity of the integrated navigation system are constrained under the action of the weld, and the attitude and velocity errors are estimated and corrected in time. The celestial error in the position error tends to expand. This is because the altitude channel of the strapdown inertial navigation system calculates the altitude through the double integration of the vertical acceleration, and the zero bias of the accelerometer will cause the altitude error to accumulate rapidly over time. The extended Kalman filter model only considers horizontal observations, not vertical observations, so the altitude error cannot be compensated.

[0129] Before inspecting oil and gas pipelines, the accuracy of the integrated navigation algorithm can be verified through a race car experiment. By comparing the path calculated by the INS-DR navigation algorithm with the actual path, various navigation parameters can be successfully obtained and presented as charts in Matlab. This experimental setup primarily serves to verify the feasibility of the integrated navigation algorithm. A small car is used in place of a real-world pipeline robot. The pipeline robot also has odometry wheels and an inertial measurement unit (IMU), but the velocities of its three wheels are averaged to obtain the equivalent odometry wheel speed. Therefore, the navigation algorithm used to restore the car's path is the same as that used in real-world pipelines. If the car's navigation algorithm can restore the car's path, it proves that the navigation algorithm can be applied to real-world pipelines. Simply replace the raw data with the pipeline robot's raw data. A Kalman filter can adaptively adjust the weights of the odometry wheels and IMU to reduce errors and improve positioning accuracy. Through verification of multiple sets of sports car experimental data (such as L path, S path, rectangular path, 8-shaped path, etc.), the combined navigation algorithm can successfully reproduce the car path, obtain the car's speed, position and attitude, and can successfully analyze the navigation error state (such as speed error, position error, attitude error, zero bias, lever arm error, etc.), and the state mean square error of the filter has converged.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm, characterized by: include: A remote-controlled car equipped with an inertial measurement unit (IMU), an encoder, and a host computer. Single-sided sponge tape is placed at regular intervals along the car's path to simulate welds. The host computer is used to run an INS-DR combined pipeline navigation algorithm program, which calculates the car's trajectory based on the car's motion data. The inertial measurement unit is used to obtain motion data generated by the remote control car during operation and upload it to the host computer; The encoder is used to obtain the odometer data generated by the remote control car during operation and upload it to the host computer; The host computer loads the motion data and odometer data of the remote control car into the INS-DR combined pipeline navigation algorithm program, and obtains the motion trajectory data through calculation; The upper computer compares the actual motion trajectory data of the remote control car with the motion trajectory data obtained by solution, thereby realizing the accuracy evaluation of the INS-DR combined pipeline navigation algorithm.

2. The experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm according to claim 1 is characterized in that: The experimental device further includes a microprocessor and a memory card installed on the remote control car. The memory card stores a data acquisition program. When the data acquisition program is running, the microprocessor collects data sent by the inertial measurement unit and the encoder.

3. The experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm according to claim 2 is characterized in that: When the data acquisition program is running, the inertial measurement unit and the encoder are controlled to output pulse signals at a sampling frequency of 100 Hz.

4. The experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm according to claim 1 is characterized in that: There are two encoders, which are respectively fixed to the two rear wheels of the remote-controlled car. During the operation of the remote-controlled car, the encoders and the wheels always rotate at the same speed.

5. The experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm according to claim 1 is characterized in that: The center point of the inertial measurement unit sensitive element is installed directly above the center point of the equivalent odometer wheel position, and the equivalent odometer wheel position is set to the center point of the line connecting the two rear wheels of the remote control car.

6. The experimental device for verifying the positioning accuracy of the INS-DR combined pipeline navigation algorithm according to claim 1 is characterized in that: The motion data collected by the inertial measurement unit includes the vehicle body angular velocity output by the three-axis gyroscope and the vehicle body specific force output by the three-axis accelerometer.

Images