Autonomous operation detection robot and autonomous operation detection method for tank bottom plate in oil

Abstract

The present invention discloses an autonomous operation detection robot for tank bottom plates in oil and an autonomous operation detection method, which relate to the field of petrochemical industry. The autonomous operation detection robot for tank bottom plates in oil includes a motion module, a positioning and posture recognition module, an obstacle avoidance module, a detection module and a control module; wherein the motion module is configured to adjust and control the robot's motion direction, speed and posture under the control of the control module; the positioning and posture recognition module is configured to recognize the robot's position and posture under the control of the control module; the obstacle avoidance module is configured to avoid obstacles under the control of the control module; and the detection module is configured to detect bottom plate corrosion under the control of the control module. The present invention can realize bottom plate corrosion detection under the condition that large oil sludge is not visible, improve detection efficiency and reduce operation and maintenance costs.

Description

Technical Field

[0001] The present invention relates to the technical field of petrochemical industry, and in particular to an autonomous operation detection robot and an autonomous operation detection method for a tank bottom plate in oil. Background Art

[0002] Large atmospheric pressure storage tanks for oil reserves are important and commonly used large storage equipment in the field of petrochemicals. They are used in the oil refining industry to store various raw oils, semi-finished oils, finished oils, aromatic products and liquefied gas, etc. The bottom plate of the storage tank is usually made of welded steel plates, which plays a role in transferring the weight of the oil and the tank. Its outer surface is in contact with the foundation and is easily affected by moisture, and the inner surface is often in contact with moisture and impurities deposited in the oil, so the bottom plate is easily corroded.

[0003] At present, great importance is attached to the service safety of large storage tanks. Most of them adopt the method of regular tank opening inspection to ensure their safety. The main mode of tank opening inspection is after oil unloading, and the method of magnetic leakage detection is used for the bottom plate of the tank. However, this mode is costly, inefficient, time-consuming, highly polluting, and has poor timeliness, which seriously affects the operation and maintenance effect of the tank. Summary of the invention

[0004] The purpose of the present invention is to provide an autonomous operation inspection robot and an autonomous operation inspection method for the bottom plate of a storage tank in oil, which can realize the bottom plate corrosion inspection under the condition that large oil sludge is not visible, improve the inspection efficiency and reduce the operation and maintenance cost.

[0005] To achieve the above object, the present invention provides the following solution.

[0006] On the one hand, the present invention provides a tank bottom plate in-oil autonomous operation detection robot (referred to as robot), comprising: a motion module, a positioning and posture recognition module, an obstacle avoidance module, a detection module and a control module; the control module is respectively connected to the motion module, the positioning and posture recognition module, the obstacle avoidance module and the detection module;

[0007] The motion module is configured to adjust and control the motion direction, speed and posture of the robot under the control of the control module;

[0008] The positioning and posture recognition module is configured to realize the recognition of the position and posture of the robot under the control of the control module;

[0009] The obstacle avoidance module is configured to achieve obstacle avoidance under the control of the control module;

[0010] The detection module is configured to realize bottom plate corrosion detection under the control of the control module; the detection module includes a detection component, a communication component and a detection control board; the detection component includes one or a combination of leakage magnetic detection sensors, electromagnetic ultrasonic detection sensors, guided wave detection sensors and pulsed eddy current detection sensors; the detection component and the communication component are both connected to the detection control board; the detection control board is connected to the control module via the communication component for transmitting the detection results back to the control module.

[0011] Optionally, the motion module includes a wheel set and a drive motor; the wheel set is installed on both sides of the long side of the robot rectangle, including a combination of one or more of crawler wheels, universal wheels and Mecanum wheels; the main shaft of the wheel set is connected to the drive motor, and the wheel set is controlled by the rotation of the drive motor, thereby adjusting and controlling the movement direction, speed and posture of the robot; the drive motor is connected to the main control board of the control module, and the main control board is controlled by an external remote control, or by the host computer of the control module.

[0012] Optionally, the positioning and attitude recognition module includes an acoustic signal generator installed on the robot, an acoustic signal receiver installed on the tank wall, and a gyroscope placed inside the robot; the positioning and attitude recognition module determines the position and horizontal attitude of the robot based on the transmitted and received acoustic signals through robot positioning and attitude solution, and solves the robot's pitch attitude based on the gyroscope.

[0013] Optionally, the number of the acoustic signal receivers installed on the tank wall is greater than 8, and at least 3 acoustic signal receivers are covered in the quadrant formed by every 90-degree angle of the tank wall circumference, and when the robot is located in a quadrant, the 3 adjacent acoustic signal receivers are used for positioning.

[0014] Optionally, the obstacle avoidance module includes a pair of sound signal generators and sound signal receivers; the sound signal generator transmits a sound signal in front of the robot, and when the sound signal encounters an obstacle, it returns, and the signal receiver receives the return signal, and the distance of the obstacle is judged by the time difference between the transmission and reception of the sound signal; the robot's route is autonomously planned by the upper computer of the control module or controlled by an external remote control to avoid obstacles.

[0015] Optionally, the tank bottom plate autonomous operation inspection robot further includes a dredging module; the dredging module is connected to the control module, and is used for cutting away the oil sludge on the traveling path to expose the bottom plate to be inspected.

[0016] Optionally, the tank bottom plate in-oil autonomous operation detection robot also includes a retracting module; the retracting module is connected to the retracting cable, and is used to place the robot into the bottom plate or retract it from the bottom plate.

[0017] On the other hand, the present invention also provides an autonomous operation detection method based on the tank bottom plate in oil autonomous operation detection robot, comprising:

[0018] The control module starts the positioning and posture recognition module to realize the recognition of its own position and posture, and adjusts the starting position and posture of the autonomous operation to the predetermined position and posture by controlling the motion module;

[0019] The control motion module enables the robot to move according to the pre-planned autonomous route;

[0020] If the obstacle avoidance module detects unexpected obstacles during the driving process, the control module will plan the best route and modify the driving route by controlling the motion module to avoid the obstacle.

[0021] During the movement, the control detection module performs the required detection on the tank bottom plate and transmits the detection results back. During this period, the positioning and posture recognition module continuously positions the robot to achieve a one-to-one correspondence between the position information and the detection results.

[0022] After the detection task is completed, the robot is returned to the starting position according to the set route by controlling the motion module.

[0023] Optionally, before the control module starts the positioning and posture recognition module, it also includes:

[0024] Place the robot into the base plate through the retractable module.

[0025] Optionally, during the traveling process, the detection module is controlled to perform required detection on the bottom plate of the storage tank and transmit the detection result back, specifically including:

[0026] During the travel, the control module turns on the magnetic flux leakage detection sensor and the electromagnetic ultrasonic detection sensor to perform magnetic flux leakage detection and electromagnetic ultrasonic thickness measurement, collects magnetic flux leakage detection signals and electromagnetic ultrasonic thickness measurement signals as detection results and transmits them to the control module;

[0027] When the magnetic flux leakage and electromagnetic ultrasound detection results are transmitted back to the control module and show no signal or the pitch angle measured by the gyroscope is greater than the threshold, the magnetic flux leakage detection sensor and the electromagnetic ultrasound detection sensor are turned off, and the pulsed eddy current detection sensor is started to perform pulsed eddy current thickness measurement, and the pulsed eddy current thickness measurement signal is collected and transmitted to the control module as the detection result;

[0028] At a specific location or when encountering unexpected obstacles, the guided wave detection sensor is started to perform single-point thickness measurement, other sensors are deactivated, and the guided wave detection signal is collected and transmitted to the control module as the detection result.

[0029] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0030] The tank bottom plate autonomous operation detection robot and autonomous operation detection method of the present invention, by setting a motion module, a positioning and posture recognition module, an obstacle avoidance module and a control module that match the detection module, enable the motion module to adjust and control the robot's motion direction, speed and posture under the control of the control module; the positioning and posture recognition module can recognize the robot's position and posture under the control of the control module; the obstacle avoidance module can avoid obstacles under the control of the control module; the detection module can detect bottom plate corrosion under the control of the control module; the purpose of autonomous operation detection in the oil in a large oil mud invisible environment is achieved, and it has the characteristics of low cost, high efficiency, short time consumption, green and pollution-free, and good timeliness, which greatly improves the operation and maintenance detection efficiency of the storage tank and reduces the operation and maintenance cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 This is a schematic diagram of the module composition of the tank bottom plate autonomous operation inspection robot;

[0033] Figure 2 This is a structural diagram of the first embodiment of the autonomous operation detection robot for the bottom plate of the tank; wherein Figure 2 Part (a) is its bottom view. Figure 2 Part (b) is a top view thereof;

[0034] Figure 3 This is a structural diagram of the second embodiment of the tank bottom plate autonomous operation detection robot; wherein Figure 3 Part (a) is its bottom view. Figure 3 Part (b) is a top view thereof;

[0035] Figure 4 This is a schematic diagram of the structure of the first embodiment of the positioning and posture recognition module;

[0036] Figure 5 This is a schematic diagram of the structure of the second embodiment of the positioning and posture recognition module;

[0037] Figure 6 It is a structural schematic diagram of an embodiment of a dredging module;

[0038] Figure 7 is a flow chart of a path tracking algorithm stored in a control module;

[0039] Figure 8Schematic diagram of an example of coordinate filtering and trajectory estimation;

[0040] Fig. 9 This is a schematic diagram of an example of command control;

[0041] Fig.10 Schematic diagram of the coordination process between the control module and the detection module. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments 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 creative work are within the scope of protection of the present invention.

[0043] In the present invention, the "in oil" detection means "under or in the oil", which is relative to the traditional tank opening detection. The current tank detection method is to export the oil in the target tank to another tank, and then perform tank cleaning detection on the target tank. The robot of the present invention can realize autonomous operation detection directly in the oil without opening and cleaning the tank. The existing technology is mainly based on tank opening detection. The main difference between tank opening detection and in oil detection is that during oil detection, due to the large amount of oil sludge accumulated on the bottom plate, the robot's movement ability is limited, and the oil detection environment is an invisible environment, so the positioning, obstacle avoidance, and navigation functions related to conventional vision and optical measurement technologies will be limited. Existing in oil detection robots rarely consider the impact of the oil sludge environment. Most of them are limited to theoretical research or overly idealized in oil detection robots, and do not consider the problem of autonomous operation detection after the robot is put into the tank. In this regard, the present invention focuses on the problem of autonomous operation detection of the robot in the oil after it is lowered into the tank, and provides an autonomous operation detection robot and an autonomous operation detection method for the tank bottom plate in the oil. The robot has the corresponding functions required for autonomous operation detection, and can realize bottom plate corrosion detection under conditions where large oil sludge is not visible, thereby improving detection efficiency and reducing operation and maintenance costs.

[0044] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] In some embodiments, see Figure 1The present invention provides an autonomous operation detection robot for the bottom plate of a storage tank in oil, which mainly includes: a motion module 1, a positioning and posture recognition module 2, an obstacle avoidance module 3, a detection module 4 and a control module 5. In some embodiments, the robot may also be equipped with a retractable module 6 connected to the retractable cable, and / or include a dredging module 7 (depending on the oil sludge situation), and the whole needs to meet the explosion-proof requirements. The control module 5 is respectively connected to the motion module 1, the positioning and posture recognition module 2, the obstacle avoidance module 3, the detection module 4, the retractable module 6 and the dredging module 7. The control module 5 includes an upper computer and a main control board connected to each other. Among them, the motion module 1 is configured to adjust and control the robot's motion direction, speed and posture under the control of the control module. The positioning and posture recognition module 2 is configured to recognize the robot's position and posture under the control of the control module. The obstacle avoidance module 3 is configured to avoid obstacles under the control of the control module. The detection module 4 is configured to detect the corrosion of the bottom plate under the control of the control module. The retractable module 6 is connected to the retractable cable and is used to put the robot into the bottom plate or retract it from the bottom plate. The dredging module 7 is used to cut the sludge on the road to expose the bottom plate to be inspected.

[0046] Figure 2 and Figure 3 Two different embodiments of the tank bottom plate autonomous oil inspection robot of the present invention are shown respectively. Figure 2 Part (a) and Figure 3 Part (a) of the diagram is a bottom view. Figure 2 Part (b) and Figure 3 Part (b) of the diagram is a top view. Figure 2 and Figure 3 , the motion module 1 includes a wheel set and a drive motor. The wheel set is installed on both sides of the long side of the robot rectangle, including but not limited to a combination of one or more of crawler type, universal wheel and Mecanum wheel. The main shaft of the wheel set is connected to the drive motor, and the wheel set is controlled by the rotation of the drive motor, thereby adjusting and controlling the movement direction, speed and posture of the robot. The drive motor is connected to the main control board of the control module 5, and the main control board can be controlled by an external remote control or by the host computer of the control module.

[0047] See also Figure 4, the positioning and posture recognition module 2 includes a plurality of acoustic signal generators 201 installed on the robot, a plurality of acoustic signal receivers 202 installed on the tank wall, and a gyroscope (not shown in the figure) placed inside the robot. The positions of the plurality of acoustic signal generators 201 distributed on the robot do not overlap, but the heights are the same. The plurality of acoustic signal receivers 202 are distributed circumferentially on the outer wall of the storage tank, and the height of the plurality of acoustic signal receivers 202 from the ground is equal to the height of the acoustic signal generator 201. The number of acoustic signal receivers 202 installed on the tank wall is at least 3, and the plurality of acoustic signal receivers 202 are evenly distributed circumferentially on the outer wall of the storage tank. The three acoustic signal receivers 202 closest to the acoustic signal generator 201 are selected, and the sound source is positioned according to the three-point positioning method. Based on the transmitted and received acoustic signals, the position and horizontal posture of the robot can be determined through robot positioning and posture solution. Furthermore, the pitch attitude of the robot can be solved based on the pitch angle data measured by the gyroscope.

[0048] Specifically, the acoustic signal generator 201 is installed on the robot, at least at two non-overlapping locations on the robot, and the installation location is adjustable. Different acoustic signal generator locations use different location and posture calculation algorithms. Figure 4 As shown, if the two sound signal generators 201 are located on both sides of the center point of the robot, the robot position can be obtained by connecting the two positioning points and finding the center position during the solution; the connection direction of the two sound signal generators 201 can be determined as the long side direction of the robot, that is, the movement direction of the robot. Figure 2 and Figure 3 As shown, the center point of the line connecting the two sound signal generators 201 is not the center position of the robot, so a slightly complicated calculation is required, and the direction of the line is not parallel to the movement direction of the robot, and a slightly complicated solution process is also required to determine the robot posture. In practical applications, the above methods are all feasible and do not affect the robot position and posture solution. In theory, as long as the positions of the two sound signal generators 201 do not overlap, the robot position and posture solution can be performed. In fact, the farther the distance between the two sound signal generators 201, the better. Figure 4 Two acoustic signal generators are distributed along the long side on the center line of the long side; Figure 2 and Figure 3It is planned according to the mechanical structure design of the robot. Specifically, the two sound signal generators are designed to be foldable. When the positioning and posture recognition module 2 is not working, the sound signal generator is laid down to lower the height of the robot without causing obstacles to the robot's movement. Therefore, in the mechanical space, there needs to be enough space to place the sound signal generator. The two sound signal generators are placed at any position on the upper cover of the robot. Since the positions of the two positioning points are known, the position of the center point of the upper cover relative to the two sound signal generators is also known, and the posture of the robot is also known. At this time, the robot is placed in the tank and the positions of the two sound signal generators are determined by the positioning algorithm. The position and posture of the robot can be determined by posture solution.

[0049] Figure 4 In the embodiment shown, the positioning and posture recognition module 2 is composed of two acoustic signal generators 201 and four acoustic signal receivers 202 installed on the tank wall. The two acoustic signal generators 201 can be installed on the top of the support rod or on the bottom of the robot, with the same height, which can be adjusted as needed. Figure 3 The robot embodiment in the embodiment is provided with a three-section support rod, the function of which is to protrude the sound signal generator outside the robot to prevent the robot from blocking signal transmission. The three-section support rod is designed to be foldable, which can be propped up when in use and laid down when not in use, without causing any obstacles to the movement of the robot. The four sound signal receivers 202 are distributed circumferentially on the outer wall of the storage tank, and the height of the four sound signal receivers 202 from the ground is equal to the height of the sound signal generator 201. The four sound signal receivers 202 are evenly distributed on the outer wall of the storage tank, and the three sound signal receivers 202 closest to the sound signal generator 201 are selected, and the sound source is located according to the three-point positioning method. The positioning position and posture of the robot are confirmed by positioning the two sound signal generators 201. Specifically, the position of the robot is defined as the center point of the rectangle on the robot cover, and the two sound signal generators 201 are located on the axis of the symmetry axis of the long side of the rectangle, and are equidistant from the center point. At this time, as long as the positions of the two sound signal generators 201 are obtained, the long side direction of the rectangle on the robot cover is determined by connecting the two, which is the movement direction of the robot. The midpoint of the line connecting the two can be solved as the center of the rectangle, which is the position of the robot.

[0050] like Figure 4 As shown, if the center position of the line connecting the two acoustic signal generator 201 positioning points can be confirmed as the position of the robot, the direction of the robot (i.e., the direction of movement) can be determined by the direction of the line connecting the two positioning points. The direction of the robot is based on the direction of the long side symmetry axis. Figure 2 and Figure 3 As shown, assuming that the line connecting the two positioning points is not parallel to the long side symmetry axis, when the angle between the line connecting the two positioning points and the symmetry axis is known, the direction of the symmetry axis can also be obtained after the direction of the line is determined. Figure 4 In the illustrated embodiment, the number of the acoustic signal receivers 202 is 4, but may also be 3, 5, 6, 7 or 8.

[0051] Figure 5 Another arrangement of the acoustic signal generator 201 and the acoustic signal receiver 202 in the positioning and posture recognition module 2 is shown, wherein the center point of the connection line of the two acoustic signal generators 201 is not the center position of the robot, and the connection direction is not parallel to the movement direction of the robot. The acoustic signal generator 201 adopts a low-frequency pulse acoustic signal generator, which is installed on the robot, and the signal frequency band is about 35kHz. One of the acoustic signal generators 201 is located at a corner of one long side of the robot, and the other acoustic signal generator 201 is located at the midpoint of another long side of the robot. Optionally, the acoustic signal generator 201 can be installed on the top of the erectable support rod, and the emitted acoustic signal propagates circumferentially in the horizontal plane of the storage medium; the acoustic signal generator 201 can also be installed at the bottom of the robot, and the emitted acoustic signal is emitted perpendicular to the bottom plate of the storage tank, and the acoustic signal enters the bottom plate and then propagates circumferentially along the bottom plate of the storage tank.

[0052] The acoustic signal receiver 202 uses a low-frequency acoustic signal receiver, which is distributed circumferentially on the outer wall of the storage tank. The acoustic signal receiver 202 is at the same height from the ground as the acoustic signal generator 201. The number of the acoustic signal receivers 202 is at least 3, which is used to locate the position of the acoustic signal source. Optionally, the acoustic signal receivers 202 are as large as possible, so that at least 3 acoustic signal receivers 202 are covered in each quadrant formed by a 90-degree angle, so that when the robot is located in a quadrant, the adjacent 3 acoustic signal receivers can be selected for positioning, avoiding serious signal attenuation when the acoustic signal is transmitted to the remote receiver.

[0053] like Figure 5 As shown, when there are 8 sound signal receivers 202, each 90-degree quadrant of the first / second / third / fourth quadrant can be covered by at least 3 sound signal receivers 202. During use, first, one of the sound signal generators 201 sends out a group of positioning pulses. After several milliseconds, each sound signal receiver 202 will receive a pulse signal successively. The propagation distance of the sound signal can be calculated by calculating the time between the signal emission and the peak value of the received pulse signal. According to the three-point positioning principle, the positioning point of the sound source can be calculated. After the positioning is completed, another sound signal generator 201 sends out a positioning signal again, and the above process is repeated to calculate the positioning point of the second sound source. According to the coordinates of the two positioning points of the two sound signal generators 201, the direction of the robot can be calculated.

[0054] See also Figure 4The obstacle avoidance module 3 includes a pair of acoustic signal generators and acoustic signal receivers, which are respectively called obstacle avoidance acoustic signal generators 301 and obstacle avoidance acoustic signal receivers 302. The obstacle avoidance acoustic signal generator 301 transmits an acoustic signal to the robot in front of it. When the acoustic signal encounters an obstacle, it returns. The obstacle avoidance acoustic signal receiver 302 receives the returned acoustic signal, and the distance of the obstacle is judged by the time difference between the transmission and reception of the acoustic signal. The robot's route is autonomously planned by the upper computer of the control module 5 or controlled by an external remote controller to avoid obstacles. The obstacle avoidance module 3 runs continuously during the detection process.

[0055] The detection module 4 is usually located at the abdomen of the robot body, and the various components it includes are arranged as needed. Specifically, the detection module 4 includes a detection component, a communication component and a detection control board. The detection component includes a magnetic flux leakage detection sensor, an electromagnetic ultrasonic detection sensor, a guided wave detection sensor, a pulsed eddy current detection sensor, or a combination of the above. The detection component and the communication component are both connected to the detection control board. The detection control board is connected to the control module 5 via the communication component, and is used to transmit the detection results back to the control module 5.

[0056] When the number of sensors is greater than 1, each sensor is integrated through a detection control board, which has a communication component for transmitting the detection results back to the control module 5 and then back to the host computer, and the detection control board is directly connected and communicated with the main control board of the control module 5. When the leakage magnetic field, electromagnetic ultrasonic, guided wave and pulsed eddy current sensors are integrated, due to the mutual interference between the permanent magnet and the excitation coil, there is the following relationship between these integrated sensors: the leakage magnetic field detection sensor can work when the robot is moving; the electromagnetic ultrasonic detection sensor can work when the robot is stationary or moving; the guided wave detection sensor can work when the robot is stationary; the leakage magnetic field and electromagnetic ultrasonic detection sensors can work at the same time; the electromagnetic ultrasonic and guided wave detection sensors can work at the same time; the three cannot work at the same time.

[0057] Guided wave testing is to excite and receive guided waves on components through the magnetostrictive effect, and determine the defect location based on the defect echo. The defect is a point defect, such as corrosion pits. Magnetic flux leakage testing measures the size of the leakage magnetic field at the defect to determine the corrosion pit. Electromagnetic ultrasonic testing excites and receives ultrasonic waves to measure wall thickness under the combined action of Lorentz force, magnetizing force and magnetostrictive strain. When the pulsed eddy current detection sensor is working, other sensors need to stop working; this is because other sensors are equipped with permanent magnets and coils. In order to avoid electromagnetic interference, pulsed eddy current can only work alone and is used to detect sub-surface defects of materials, mainly uniform corrosion. Magnetic leakage and guided waves are mainly used to detect corrosion defects, mainly point defects. Theoretically, the range that guided waves can detect is larger than that of magnetic leakage. For example, the distance within 1 meter around is its detection range, while magnetic leakage can only detect the area directly below the sensor. The main function of electromagnetic ultrasound and pulsed eddy current is thickness measurement, that is, to detect uniform corrosion of the bottom plate of the tank. The difference between the two is that the pulsed eddy current can theoretically achieve a relatively large lift, such as more than 100 mm, while the electromagnetic ultrasound lift is relatively small, such as a few millimeters. Therefore, when the oil sludge in the tank reaches a certain thickness, the electromagnetic ultrasound will not be able to detect the signal. At this time, pulsed eddy current is required for thickness measurement.

[0058] The control module 5 includes a main control board and a host computer; the host computer controls the corresponding module by sending instructions to the main control board, or the host computer directly sends instructions to the corresponding module to control the module. For example, the motion module 1 is controlled by controlling the driving motor of the wheel group; the start and stop of the positioning and posture recognition module 2 is controlled by controlling the sound signal generator 201 and the sound signal receiver 202; the start and stop of the detection module 4 are controlled by sending instructions to the detection control board of the detection module 4; the silt removal module 7 is controlled by controlling the silt removal wheel motor, etc. The positioning and posture recognition module 2 must cooperate with the detection module 4 to obtain the position information corresponding to the detection result, and it is also necessary to cooperate to complete the posture correction during the operation of the robot.

[0059] like Figure 3 As shown, the desilting module 7 is arranged in the front part of the robot's movement direction. Since the sludge is usually thick and sticky, it hinders the robot's movement. Therefore, desilting wheels are arranged at both ends of the robot to facilitate the movement of the robot. In some embodiments, the structure of the desilting wheel is as follows: Figure 6 As shown, it is a paddle wheel structure, the gear teeth are symmetrical, tilted toward the symmetry axis, and the tilt angle range is 5 to 15 degrees. When working, it rotates, using the cutting force of the gear teeth to cut the oil mud on the road, and the gear teeth rotate to clean the cut oil mud, exposing the bottom plate to be tested. The function of ordinary wheel sets is only to travel.

[0060] like Figure 2 and Figure 3As shown, the control module 5 and the detection module 4 are generally placed in the internal space of the robot. Figure 2 The robot in the illustrated embodiment is characterized in that: the motion module 1 is wheeled; the obstacle avoidance module 3 is provided only at the front end in the motion direction; the sensor of the positioning and posture recognition module 2 is arranged diagonally; and the dredging module 7 is not included. Figure 3 The robot in the illustrated embodiment is characterized in that: the motion module 1 is tracked; the obstacle avoidance module 3 is provided only at the front end in the motion direction; the sensors of the positioning and posture recognition module 2 are arranged on both sides of the robot; and a dredging wheel is included as a dredging module 7.

[0061] Based on the tank bottom plate in oil autonomous operation detection robot, the present invention also provides an autonomous operation detection method, including the following S1 to S5.

[0062] S1. After the robot is placed on the bottom plate through the retractable module 6, the control module 5 starts the positioning and posture recognition module 2 to recognize its own position and posture, and adjusts the starting position and posture of the autonomous operation to the predetermined position and posture by controlling the motion module 1.

[0063] Before the robot enters the tank for inspection, it will determine the priority position to start inspection through the tank body drawing and determine the posture. The posture is the direction of the robot's head and tail, which is convenient for subsequent inspections. Therefore, the predetermined position and posture are determined according to actual needs. After the robot is placed on the bottom plate through the cable, the position at the bottom may deviate from the predetermined position and posture because the cable is long and soft. Therefore, it is necessary to control the motion module 1 to move to the predetermined position and move the wheels to complete the posture adjustment.

[0064] S2. Control the motion module 1 to make the robot move according to the pre-planned autonomous route.

[0065] The control module 5 stores a path tracking algorithm, which can autonomously plan the robot's route, so that the robot can move according to the pre-autonomously planned route by controlling the motion module 1. The flow chart of the path tracking algorithm is shown in FIG. Figure 7 As shown. By setting the state variables flagA and flagB, it is determined whether it is in the correction state. When entering the correction state, the corresponding state is set to 1, and when exiting the correction state, it is set to 0. The specific process of the path tracking algorithm is as follows.

[0066] S2.1. Eliminate outliers. At the moment, the robot's own position coordinates are taken as the coordinates of one of the acoustic signal generators in the positioning and posture recognition module 2, denoted as The direction angle is calculated by the angle between the two consecutive positioning points of the two acoustic signal transmitters, the connecting line of the two acoustic signal generators and the front of the robot, and is recorded as ; The original value of the robot's position (abbreviation for position and posture) is denoted as { , }.exist At this moment, if abnormal noise occurs, which may cause the change of position or direction to exceed the allowed value, the measured value of the pose is discarded and replaced by The velocity, angular velocity and posture of the moment Predict at every moment and get the predicted pose value { , }; Otherwise, update the pose.

[0067] S2.2, Direction Angle Filtering. Since there are errors in the positioning points of the two acoustic signal generators in front and behind the robot, the Kalman filter is used to reduce the error of the robot's direction angle. The direction angle estimation value is .

[0068] S2.3, coordinate filtering. On a two-dimensional plane, the robot coordinate point motion trajectory that originally moved smoothly will oscillate and reciprocate due to positioning errors. In order to eliminate the distortion of the motion trajectory, the segmented fitting method is used. The following takes straight line fitting as an example. A group of points (short for positioning points) is fitted piecewise, and the coordinates are a set , the line segment defined within the point set is obtained through the linear fitting formula ; At this time, the robot coordinates are ,Right now .from Point The intersection point of the perpendicular line is the estimated value of the current robot position. ; The estimated value of the pose is { , }.

[0069] S2.4, estimate the path (or route). Assume that the path update ratio is , discard the set The front points and round them, then update ( - ) points, and get a new set And fit the line segment The existing point set ,in and are the estimated values ​​of the starting point and the current point, +1 for point set In order to make the curve continuous and local, the B-spline curve is used to draw the robot's motion trajectory, which is used to give the mapping of the detection data in motion.

[0070] S2.5, command control. The robot's previous path point is , the target path point is . Calculate the robot's current path point based on the geometric relationship With path Vertical deviation of the path d , the direction deviation between the current direction and the path direction is γ. When the path vertical deviation is used as the control variable, the robot's position error is smaller, but the robot's travel direction is in an adjustment state for a long time, so the travel route will oscillate slightly around the path, which is not conducive to full coverage detection. When the direction deviation is used as the control variable, the robot's travel direction is more stable, but there is an accumulated error in the position. Therefore, it is necessary to combine the two methods. Let γ max is the maximum allowable directional deviation, γ min is the minimum permissible directional deviation; d max is the maximum allowable position error, d min is the minimum allowable position error. min and d < d min When γ>γ, the left and right drives of the robot output the same speed; max and d < d min When γ<γ, the direction deviation is used as the control variable to input into the PID controller A to correct the movement direction of the robot. The speed ratio of the left and right motors is proportional to the output of the controller until γ<γ min ;when d > d max When the vertical deviation of the path d As the control variable, it is input into PID controller B until d < d min .

[0071] S2.6, stop. If the coordinates of the robot are within a circle with a radius equal to the positioning error for multiple consecutive times, it is determined that the robot has stopped due to manual or abnormal circumstances, and the path tracking is stopped.

[0072] The following two examples are used for specific explanation. Figure 8 For an example of coordinate filtering and trajectory estimation, the number of points in the set =4, path update proportionality factor =0.5. The point set M1 is { , }, and the straight line is obtained by fitting With estimated points , where the straight line and Vertical. After that, clear the first two points in the point set M1 and get { }, continue to collect positioning points until a new point set M2{ , }, and get If the degree of the B-spline curve is =3, then when =8 to get the point set { , }, the first B-spline curve can be obtained as the trajectory (i.e. the route); when =10, the second B-spline curve is obtained. Due to locality, it will not be updated. and The line segments between .

[0073] Fig. 9 This is an example of command control. The horizontal direction is the forward direction. Assume that the robot has a larger load on the right side due to uneven ground in the travel area, resulting in a deviation to the right. The moving direction of the initial position 1 is straight ahead. When reaching position 2, the robot deviates to the right. At this time, γ<γ min , the left and right drives continue to maintain the same speed. When reaching position 3, γ>γ max , start PID controller A to adjust the left and right drivers and increase the right speed. When reaching position 4, γ is corrected. d < d min And γ<γ min , the left and right drivers output the same speed. When reaching position 5, γ>γ max If it happens again, continue to correct the direction. When you reach position 6, d>d max , PID controller B is started to adjust the left and right drivers, increasing the speed of the right driver to correct the position error. When reaching positions 7 and 8, the robot begins to deviate to the left. When reaching position 9, d <d min And γ<γ min , the left and right drivers output the same speed.

[0074] S3. During the process of traveling, if the obstacle avoidance module 3 finds unexpected obstacles, the control module 5 will plan the best route and modify the route to avoid obstacles by controlling the motion module 1. Specifically, the internal structure of the tank is obtained in advance through the tank processing drawings, and the route is planned autonomously; when encountering unexpected obstacles, the algorithm needs to be used to achieve the best route planning to avoid obstacles and return to the original route.

[0075] S4. During the process of traveling, the control detection module 4 performs the required detection on the bottom plate of the storage tank, and transmits the detection results back at the same time. During this period, the positioning and posture recognition module 2 continuously positions the robot to achieve a one-to-one correspondence between the position information and the detection results. The detection of the bottom plate is determined according to the needs. For example, if the thickness of the bottom plate needs to be detected, electromagnetic ultrasound can be used, and the detection result transmitted back is the ultrasonic echo data. Different detection sensors transmit different types of data, and different data are transmitted back as detection results according to the characteristics of the detection sensor. The data transmitted back is generally raw data, which is transmitted back to the host computer through the main control board, and the final data processing and result analysis are performed on the host computer.

[0076] See also Fig.10 , the coordination between the control module 5 and the detection module 4 is as follows: the host computer of the control module 5 selects different control channels through the logic control component, starts the excitation end of a sensor in the detection module 4 to transmit the excitation signal, receives the echo signal through the receiving end of the integrated probe, and uploads the echo signal to the signal conditioning component of the control module 5, performs signal acquisition through the AD acquisition component, and uploads the acquired signal to ARM through the logic control component, and is stored by the storage component. The integrated probe is the main part of the detection module 4, and the receiving end is used to receive the signal. The signal conditioning component is located on the main control board. The logic control component is the core component on the main control board and is used to control other modules. The logic control component controls different sensors to perform different actions through different communication protocols sent by the host computer. The excitation ends of different detection sensors are different, some have excitation ends, and some do not. For example, magnetic leakage is passive detection and does not need to send excitation signals; the excitation ends of ultrasound, guided wave, and pulsed eddy current are different, and the excitation signals are also different. The host computer sends an action command, which has been agreed upon as a control command for a sensor through the protocol. The logic control component judges and transmits the command to the corresponding sensor. After receiving the command, the corresponding sensor performs the corresponding detection action according to the protocol. There are two types of detection actions: one is passive detection, such as the above-mentioned magnetic leakage, which starts receiving signals, and the other is active detection, which requires sending excitation signals and then receiving echoes.

[0077] If the integrated sensor is an integrated probe of magnetic flux leakage, electromagnetic ultrasound, guided wave and pulsed eddy current, the detection process is as follows:

[0078] After the robot is lowered into the tank and adjusted to a predetermined position and posture, it moves along a predetermined route. At this time, the magnetic flux leakage detection sensor and the electromagnetic ultrasonic detection sensor are turned on to perform magnetic flux leakage and electromagnetic ultrasonic thickness measurement, and the magnetic flux leakage detection signal and the electromagnetic ultrasonic thickness measurement signal are collected as the detection result and transmitted to the control module 5;

[0079] When the detection results of magnetic flux leakage and electromagnetic ultrasound are transmitted back to the host computer and show no signal, or the pitch angle measured by the gyroscope is greater than the threshold value, the above two sensors are turned off, and the pulsed eddy current detection sensor is started to perform pulsed eddy current thickness measurement, and the pulsed eddy current thickness measurement signal is collected and transmitted to the control module 5 as the detection result; this is because when the lift-off is high, the two detection methods of magnetic flux leakage and electromagnetic ultrasound fail, and the pulsed eddy current can achieve thickness measurement under large lift-off conditions; and when the pitch angle is greater than the threshold value, it means that the sludge is too thick, and the magnetic flux leakage and electromagnetic ultrasound data may be missed or misdetected, so it is necessary to start the pulsed eddy current for further detection; the threshold value of the pitch angle is set according to the actual detection situation;

[0080] At a specific location (an area suspected of being defective and inaccessible to the robot, such as some complex terrain areas or obstructed areas in the tank, at this time, a specific location is selected for the robot to stop and the guided wave detection is started) or when encountering unexpected obstacles, the guided wave detection sensor is started to perform single-point thickness measurement, other sensors are disabled, and the guided wave detection signal is collected and transmitted to the control module 5 as the detection result.

[0081] The host computer of the control module 5 stores the original signal, and the stored data can be analyzed and processed immediately or later. The result of the single-point thickness measurement is the thickness of the bottom plate directly below the detection module 4. By comparing different periods or the original thickness, it is analyzed whether the bottom plate has been thinned. Thinning means corrosion has occurred. Different detection methods have different detection advantages. For the same sample to be tested, different detection methods can obtain different detection results. The detection results obtained by different detection methods corroborate each other, and finally obtain reliable detection conclusions, realize the bottom plate corrosion detection under the condition that large sludge is not visible, and improve the detection accuracy.

[0082] S5. After the detection task is completed, the robot is returned to the starting position according to the set route by controlling the motion module 1, that is, the predetermined position before the detection begins.

[0083] The tank bottom plate in-oil autonomous operation detection robot and autonomous operation detection method of the present invention achieve the purpose of in-oil autonomous operation detection in a large oil sludge invisible environment through movement, positioning and posture recognition, obstacle avoidance, retraction and silt removal modules that match the detection capability. It has the characteristics of low cost, high efficiency, short time consumption, green and pollution-free, and good timeliness, which is greatly beneficial to the operation and maintenance of storage tanks.

[0084] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An autonomous operation detection method based on an autonomous operation detection robot for a tank bottom plate in oil, characterized in that: The tank bottom plate in-oil autonomous operation detection robot comprises: a motion module, a positioning and posture recognition module, an obstacle avoidance module, a detection module and a control module; the control module is respectively connected to the motion module, the positioning and posture recognition module, the obstacle avoidance module and the detection module; The motion module is configured to adjust and control the motion direction, speed and posture of the robot under the control of the control module; The positioning and posture recognition module is configured to realize the recognition of the position and posture of the robot under the control of the control module; The obstacle avoidance module is configured to achieve obstacle avoidance under the control of the control module; The detection module is configured to realize bottom plate corrosion detection under the control of the control module; the detection module includes a detection component, a communication component and a detection control board; the detection component includes one or a combination of a magnetic flux leakage detection sensor, an electromagnetic ultrasonic detection sensor, a guided wave detection sensor and a pulsed eddy current detection sensor; the detection component and the communication component are both connected to the detection control board; the detection control board is connected to the control module via the communication component, and is used to transmit the detection result back to the control module; The autonomous operation detection method comprises: The control module starts the positioning and posture recognition module to realize the recognition of its own position and posture, and adjusts the starting position and posture of the autonomous operation to the predetermined position and posture by controlling the motion module; The control motion module enables the robot to move according to the pre-planned autonomous route; If the obstacle avoidance module detects unexpected obstacles during the driving process, the control module will plan the best route and modify the driving route by controlling the motion module to avoid the obstacle. During the movement, the control detection module performs the required detection on the tank bottom plate and transmits the detection results back. During this period, the positioning and posture recognition module continuously positions the robot to achieve a one-to-one correspondence between the position information and the detection results. After the detection task is completed, the robot is returned to the starting position according to the set route by controlling the motion module; The positioning and attitude recognition module includes an acoustic signal generator installed on the robot, an acoustic signal receiver installed on the tank wall, and a gyroscope placed inside the robot; the positioning and attitude recognition module determines the position and horizontal attitude of the robot through robot positioning and attitude solution based on the transmitted and received acoustic signals, and solves the robot pitch attitude based on the gyroscope; The control module stores a path tracking algorithm, and the specific process of the path tracking algorithm is as follows: exist At the moment, the robot's own position coordinates are taken as the coordinates of one of the acoustic signal generators in the positioning and posture recognition module, denoted as The direction angle is calculated by the two positioning points of the two acoustic signal generators, the angle between the connecting line of the two acoustic signal generators and the front of the robot, and is recorded as ; The original value of the robot's posture is denoted by { , };exist At this moment, if abnormal noise occurs, causing the change in position or direction to exceed the allowed value, the measured value of the pose is discarded and replaced by The velocity, angular velocity and posture of the moment Predict at every moment and get the predicted pose value { , }; Otherwise, update the pose; Since there are errors in the positioning points of the two acoustic signal generators in front and behind the robot, the Kalman filter is used to reduce the error of the robot's direction angle; the direction angle estimation value is ; The segmented fitting method is used to eliminate the distortion of the motion trajectory. A group of points is fitted piecewise, and the coordinates are a set , the line segment defined within the point set is obtained through the linear fitting formula ; At this time, the robot coordinates are ,Right now ;from Point The intersection point is the estimated value of the current robot position. ; The estimated value of the pose is { , }; Assume the path update ratio is , discard the set The front points and round them, then update ( - ) points, and get a new set And fit the line segment ; Existing point set ,in and are the estimated values ​​of the starting point and the current point, +1 for point set The number of midpoints; In order to make the curve continuous and local, the B-spline curve is used to draw the robot's motion trajectory, which is used to give the mapping of the detection data in motion; The robot's previous path point is , the target path point is ; Calculate the robot's current path point based on geometric relationships With path Vertical deviation of the path d , the direction deviation between the current direction and the path direction is γ; let γ max is the maximum allowable directional deviation, γ min is the minimum permissible directional deviation; d max is the maximum allowable position error, d min is the minimum allowable position error; when γ<γ min and d < d min When γ>γ max and d < d min When γ is used as the control variable, the PID controller A is input to correct the movement direction of the robot. The speed ratio of the left and right motors is proportional to the output of the controller until γ<γ min ;when d > d max When the vertical deviation of the path d As the control variable, it is input into PID controller B until d < d min ; If the coordinates of the robot are located within a circle with a radius equal to the positioning error for multiple consecutive times, it is determined that the robot has stopped due to manual or abnormal circumstances, and path tracking is stopped.

2. The autonomous operation detection method according to claim 1, characterized in that: The motion module includes a wheel set and a drive motor; the wheel set is installed on both sides of the long side of the robot rectangle, including a combination of one or more of crawler wheels, universal wheels and Mecanum wheels; the main shaft of the wheel set is connected to the drive motor, and the wheel set is controlled by the rotation of the drive motor, thereby adjusting and controlling the movement direction, speed and posture of the robot; the drive motor is connected to the main control board of the control module, and the main control board is controlled by an external remote control, or by the host computer of the control module.

3. The autonomous operation detection method according to claim 1, characterized in that: The number of the acoustic signal receivers installed on the tank wall is greater than 8, and each quadrant formed by every 90-degree angle of the tank wall is covered with at least 3 acoustic signal receivers. When the robot is located in a quadrant, the three adjacent acoustic signal receivers are used for positioning.

4. The autonomous operation detection method according to claim 1, characterized in that: The obstacle avoidance module includes a pair of sound signal generators and sound signal receivers; the sound signal generator transmits a sound signal to the front of the robot, and when the sound signal encounters an obstacle and returns, the signal receiver receives the return signal, and the distance of the obstacle is judged by the time difference between the transmission and reception of the sound signal; the robot's route is autonomously planned by the upper computer of the control module or controlled by an external remote control to avoid obstacles.

5. The autonomous operation detection method according to claim 1, characterized in that: It also includes a dredging module; the dredging module is connected to the control module and is used to cut away the sludge on the road to expose the bottom plate to be detected.

6. The autonomous operation detection method according to claim 1, characterized in that: It also includes a retractable module; the retractable module is connected to the retractable cable and is used to put the robot into the base plate or retract it from the base plate.

7. The autonomous operation detection method according to claim 1, characterized in that: Before the control module starts the positioning and posture recognition module, it also includes: Place the robot into the base plate through the retractable module.

8. The autonomous operation detection method according to claim 1, characterized in that: During the traveling process, the detection module is controlled to perform the required detection on the bottom plate of the storage tank and transmit the detection results back, specifically including: During the travel, the control module turns on the magnetic flux leakage detection sensor and the electromagnetic ultrasonic detection sensor to perform magnetic flux leakage detection and electromagnetic ultrasonic thickness measurement, collects magnetic flux leakage detection signals and electromagnetic ultrasonic thickness measurement signals as detection results and transmits them to the control module; When the magnetic flux leakage and electromagnetic ultrasound detection results are transmitted back to the control module and show no signal or the pitch angle measured by the gyroscope is greater than the threshold, the magnetic flux leakage detection sensor and the electromagnetic ultrasound detection sensor are turned off, and the pulsed eddy current detection sensor is started to perform pulsed eddy current thickness measurement, and the pulsed eddy current thickness measurement signal is collected and transmitted to the control module as the detection result; At a specific location or when encountering unexpected obstacles, the guided wave detection sensor is started to perform single-point thickness measurement, other sensors are deactivated, and the guided wave detection signal is collected and transmitted to the control module as the detection result.

Images