Sea-turtle-imitating robot for submarine pipeline leakage monitoring and multi-mode control method of sea-turtle-imitating robot
By designing imitation turtle robots and adopting multimodal control methods, the problem of leakage monitoring in the submarine pipeline is solved, real-time monitoring and leakage detection of submarine pipelines is achieved, and economic losses and marine environmental pollution are reduced.
Patent Information
- Application Number
- CN202411513353.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-28
- Publication Date
- 2025-02-14
AI Technical Summary
The existing technology is difficult to effectively monitor the leakage of subsea pipelines, resulting in economic losses, safety issues and marine environmental pollution.
Design a turtle imitation robot, equipped with components such as trunk, head, pectoral fin, caudal fin, detection control unit, etc., adopt multimodal control method, and realize the submarine pipeline through sensor data such as vision sensor, sonar sensor and attitude sensor. Real-time monitoring and leak detection.
It realizes timely detection and monitoring of subsea pipeline leakage, reduces economic losses and marine environmental pollution, and has low maintenance costs.
Smart Images

Figure CN119429038A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of submarine pipeline leakage monitoring, and in particular to a turtle-like robot for submarine pipeline leakage monitoring and a multi-modal control method thereof. Background Art
[0002] Submarine pipelines are the main component of offshore oil and gas field development and production systems, and are also the safest, fastest, most economical and reliable way to transport offshore oil and gas. They mainly transport large amounts of oil and gas continuously on the seabed through closed pipelines. Therefore, they are also known as the "lifeline" of offshore oil and gas projects. When transporting oil and gas, submarine pipelines expand under the action of temperature, pressure and Poisson's effect, generating additional stress on the pipe wall, and as the length of the pipeline increases, the additional stress of the pipeline continues to accumulate. When the accumulated stress is greater than the critical buckling load of the pipeline, the overall buckling of the pipeline may lead to structural failure of the pipeline, such as fracture, fatigue or expansion buckling. At the same time, most submarine pipelines are made of steel materials, and the pipelines are in seawater for a long time, and the environment in which the pipelines are located is often acidic. Therefore, corrosion is also prone to occur inside and outside the pipeline, causing pipeline failure. When the pipeline is damaged and leaks, it will not only cause economic losses, but also may bring safety problems, endangering personal safety and the marine environment. Relying on marine mobile platforms to monitor submarine pipeline leakage is one of the means to detect pipeline damage and leakage. Its manufacturing and maintenance costs are high, and the equipment investment is large. It is difficult to achieve comprehensive and effective monitoring of submarine pipelines, and often the leakage point cannot be found in time. Therefore, there is an urgent need for a robot for submarine pipeline leakage monitoring to promote the development of submarine pipeline leakage monitoring. Summary of the invention
[0003] In view of the above-mentioned deficiencies in the prior art, one object of the present invention is to propose a turtle-like robot for monitoring submarine pipeline leakage, so as to solve the problem of lack of effective monitoring means for damage and leakage of submarine pipelines, which may easily cause economic losses, safety problems and endanger the marine environment.
[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0005] A turtle-like robot for monitoring submarine pipeline leakage comprises a trunk, a head, a top cover, a pectoral fin plate, a caudal fin plate, a pectoral fin drive device, a center of gravity adjustment mechanism and a detection control unit. The trunk is an elliptical cone-shaped shell structure, and the top cover is a spherical shell that matches the top of the trunk. The top cover is fixedly installed above the trunk, and a sealed upper chamber is formed inside.
[0006] The detection control unit includes a visual sensor, an image recognition module, a communication module, an altimeter, a sonar sensor, a nine-axis attitude sensor and a controller. The head is arranged in front of the torso through the neck, the visual sensor is arranged on the head, and the image recognition module, the communication module, the altimeter, the sonar sensor, the nine-axis attitude sensor and the controller are arranged inside the upper chamber.
[0007] There are two pectoral fin plates, which are symmetrically arranged on the left and right sides of the front of the trunk. There are two pectoral fin driving devices, which are arranged on the front of the trunk and are symmetrically arranged on the left and right sides.
[0008] The pectoral fin drive device includes a first servo, a transmission mechanism and an angle adjustment assembly. The transmission mechanism is arranged on the inner side of the trunk through a bearing bracket. The output end of the first servo is connected to the end of the pectoral fin plate on the same side through the transmission mechanism to drive the pectoral fin plate to swing back and forth. During the swinging process of the pectoral fin plate, the angle adjustment assembly controls the angle posture of the pectoral fin plate.
[0009] The middle and rear part of the trunk is a sealed lower chamber, the pectoral fin drive device is located outside the front side of the lower chamber, the center of gravity adjustment mechanism includes a counterweight block, a linear guide mechanism and a servo motor, the counterweight block is slidably arranged in the lower chamber, and the servo motor drives the counterweight block to move forward and backward through the linear guide mechanism.
[0010] There are two tail fins, which are symmetrically arranged on the left and right sides of the rear of the trunk. The front end of each tail fin is rotatably connected to the rear end of the trunk through a second steering gear, and each second steering gear drives the corresponding tail fin to swing up and down around its execution end.
[0011] Furthermore, the trunk includes a belly plate, an upper plate and an annular side plate, the upper plate is located directly above the belly plate, the lower end of the annular side plate is fixedly connected to the outer edge of the belly plate, and the upper end thereof is fixedly connected to the outer edge of the upper plate.
[0012] A partition is provided at the inner front part of the annular side plate, and the partition cooperates with the web plate, the upper plate and the annular side plate to form the lower chamber.
[0013] The front part of the annular side plate is provided with two strip holes which are symmetrically opened on the left and right sides, and the strip holes are located on the front side of the partition plate.
[0014] Furthermore, the head includes a light-transmitting cover shell and a frame, the frame is located on the inner side of the light-transmitting cover shell, the rear end of the frame is connected to the front end of the neck, and the rear end of the neck is connected to the front end of the trunk.
[0015] There are two visual sensors, which are symmetrically arranged on both sides of the inside of the light-transmitting cover shell and are installed on the frame. The signal end of the visual sensor is connected to the controller through communication with the image recognition module.
[0016] Furthermore, the transmission mechanism includes a vertical shaft, a horizontal shaft, a crank, a connecting rod and a rocker arm. The vertical shaft is rotatably arranged on the bearing bracket, and one end of the rocker arm is fixedly connected to the lower end of the vertical shaft.
[0017] The first steering gear is fixedly installed on the rear side of the vertical shaft, one end of the crank is fixedly connected to the output shaft of the first steering gear, and the other end is movably connected to the other end of the rocker arm through the connecting rod.
[0018] The upper end of the vertical shaft is fixedly connected to a first shaft sleeve, the horizontal shaft is inserted into the first shaft sleeve and rotates with it, one end of the horizontal shaft passes through the trunk and is fixedly connected to one end of the pectoral fin plate on the same side, the crank rotates around the output shaft of the first servo, and the vertical shaft is driven to rotate through the crank and the connecting rod, and the vertical shaft drives the horizontal shaft and the pectoral fin plate to swing back and forth.
[0019] Furthermore, the angle adjustment assembly includes an arc guide rail, a drive arm, a guide rod and a ball head. The arc guide rail is in the shape of a 1 / 4 arc and is located on the side where the two bearing brackets are close to each other. It is fixedly connected to the bearing bracket through a guide rail frame, and both end surfaces of the arc guide rail adopt a rounded structure.
[0020] The driving arm is an L-shaped structure, the middle position of its vertical part is fixedly connected to the other end of the horizontal axis, and the ball head is installed at the end of the horizontal part of the driving arm. When the horizontal axis reciprocates around the vertical axis, the ball head slides and fits with the upper surface of the arc guide rail.
[0021] The guide rod is fixedly connected to the top of the first sleeve through a guide rod bracket, which is located directly above the other end of the horizontal axis. A through hole is opened at the upper end of the vertical part of the driving arm. The guide rod is movably inserted into the through hole, and both ends of the guide rod are fixedly connected to the guide rod bracket. There is an annular gap between the guide rod and the inner wall of the through hole.
[0022] Furthermore, the linear guide mechanism includes a linear slide rail, a guide slider and a lead screw. The linear slide rail is longitudinally installed at the inner bottom of the trunk, and the bottom of the counterweight block is slidably matched with the linear slide rail through the guide slider.
[0023] The lead screw is passed through the inner side of the counterweight block and matched with its thread. The front end of the lead screw is matched with the rotation of the trunk through the bearing seat, and the rear end is coaxially fixedly connected with the output shaft of the servo motor. The signal end of the servo motor is connected to the controller.
[0024] Furthermore, the left and right sides of the rear end of the trunk are provided with two concave mounting parts, and a U-shaped bracket is fixed to the rear side of each mounting part.
[0025] The second servo adopts a double-output shaft servo, the outer shell of the second servo is fixedly connected to the front end bolts of the tail fin through a servo bracket, the two output shaft ends are fixedly connected to the U-shaped bracket, the signal end of the second servo is connected to the controller for communication, and in the working state, the second servo drives the tail fin to swing back and forth around its output shaft.
[0026] Another object of the present invention is to provide a multi-modal control method for a turtle-like robot.
[0027] A multi-modal control method of a turtle-like robot is used for the above-mentioned turtle-like robot for submarine pipeline leakage monitoring. The multi-modal control method of the turtle-like robot comprises the following steps:
[0028] Step 1: Set the coordinate points of the route that the turtle robot follows the submarine pipeline to ((x'1, y'1, z'1), (x'2, y'2, z'2), ..., (x' n ,y' n ,z' n )), where (x' n ,y' n ,z' n ) is the coordinate of the tracking path point of the turtle-like robot at the nth time point;
[0029] Step 2: Obtain the current actual position coordinates of the turtle-like robot according to the altimeter and the sonar sensor, and perform difference calculation between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path point at this time;
[0030] Step 3, establish a state-action value table of the turtle-like robot, and obtain the action output in the state-action value table according to the difference calculated in step 2;
[0031] Step 4: According to the action output obtained in step 3, the turtle-like robot is controlled to realize multi-modal motion on the seabed and track the submarine pipeline line.
[0032] Furthermore, the process of obtaining the state-action value table includes the following steps:
[0033] Step 3.1: Define the current actual position coordinates of the turtle-like robot as η k =(x k ,y k ,z k ), where x k ,y k and z k are the actual current position coordinates of the turtle-like robot on the x-axis, y-axis, and z-axis at the kth time point;
[0034] Step 3.2: The coordinates of the tracking path points of the turtle-like robot at k time points are η' k =(x' k ,y' k ,z' k ), where x' k ,y' k and z' kare the coordinates of the tracking path points of the turtle-like robot at the kth time point on the x-axis, y-axis, and z-axis respectively;
[0035] Step 3.3, define the state space S, where S = [s1, s2, ..., s n ,…,s N ], N is a positive integer, N represents the number of states contained in S, s n =[E xn ,E yn ,E zn ] is the nth state in S, E xn Indicates the error interval in the x-axis direction, E yn Indicates the error interval in the y-axis direction, E zn Represents the error range in the z-axis direction;
[0036] The difference between the actual position coordinates of the turtle-like robot at the kth time point and the coordinates of the tracking path point is: e k =[x k -x' k ,y k -y' k ,z k -z' k ], let e k =[e x ,e y ,e z ], judge e k Located in the state, e x ∈E xn , e y ∈E yn , e z ∈E zn When k ∈s n , indicating that the turtle-like robot is in the nth state at this time;
[0037] Step 3.4: Define the action space A = [a1, a2, ..., a m ,…,a M ], M is a positive integer, indicating the number of actions contained in A; define action group a m =[F m1 ,θ m1 ,F m2 ,θ m2 ,F m3 ,θ m3 ,F m4 ,θ m4 ,L fr ], where F m1 is the thrust of the first servo on the left side when performing the mth action, θ m1F is the rudder angle of the first servo on the left side when executing the mth action group; m2 is the thrust of the first servo on the right side when performing the mth action, θ m2 F is the rudder angle of the first servo on the right side when executing the mth action group; m3 is the thrust of the second servo on the left side when performing the mth action, θ m3 F is the rudder angle of the second servo on the left when executing the mth action group; m4 is the thrust of the second servo on the right side when performing the mth action, θ m4 is the rudder angle of the second servo on the right side when executing the mth action group; L fr represents the position of the weight block when performing the mth action, where L fr 1 indicates that the counterweight is located at the front end of the screw; L fr -1 indicates that the counterweight is located at the rear end of the screw; L fr 0 indicates that the counterweight is in the middle of the screw;
[0038] Step 3.5: Set the reward function and calculate the reward of the turtle-like robot in the nth state s n Take action a m Then the reward value r obtained according to the reward function n+1 ;
[0039] Step 3.6, obtaining the state-action value table of the turtle-like robot according to the Q-Learning algorithm;
[0040] The calculation formula of the Q-Learning algorithm is:
[0041] In the formula, Q(s n ,a m ) is the state-action value, α is the learning rate, 0<α≤1; γ is the discount factor; r n+1 Indicates in s n State takes action a m Then transfer to the new state s n+1 The reward value obtained when Indicates that in state s n+1 The maximum value among all possible actions corresponding to the state-action value;
[0042] During the training process, the state-action value table is continuously updated according to the Q-Learning algorithm formula until all values in the state-action value table converge, the training ends, and the final state-action value table is formed.
[0043] Furthermore, the calculation formula of the reward function in step 3.5 is: n+1 =r e +r n+r o ;
[0044] In the formula, r e To reach the new state s n+1 The reward value when Where e xmax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path points in the x-axis direction, e ymax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path points in the y-axis direction, e zmax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path point in the z-axis direction;
[0045] r n is the reward obtained when reaching the path tracking point,
[0046] r o For collision rewards, L is the new state s reached by the turtle-like robot t+1 L0 is the straight-line distance from the obstacle, and L0 is the set safety distance.
[0047] By adopting the above technical scheme, the beneficial technical effect of the present invention is: the turtle-like robot for submarine pipeline leakage monitoring of the present invention realizes inspection along the submarine pipeline through posture control, uses the image information of the submarine pipeline to determine whether a leakage occurs, has low maintenance cost, can find the leakage point in time, reduce economic losses and reduce pollution to the marine environment. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Figure 1 The present invention is a schematic structural diagram of a turtle-like robot for monitoring submarine pipeline leakage.
[0049] Figure 2 It is a cross-sectional view of a turtle-like robot for monitoring submarine pipeline leakage according to the present invention.
[0050] Figure 3 yes Figure 1 Diagram of the internal structure of a part of the human body, showing the torso.
[0051] Figure 4 yes Figure 1 Another partial schematic diagram shows a combination of a pectoral fin plate and a pectoral fin drive device.
[0052] Figure 5 yes Figure 1 Another part of the schematic diagram shows the center of gravity adjustment mechanism.
[0053] Figure 6 It is a structural schematic diagram of the pectoral fin driving device of the present invention.
[0054] Figure 7 It is a schematic diagram of the combined structure of the tail fin plate, the second steering gear and related parts of the present invention. DETAILED DESCRIPTION
[0055] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and examples.
[0056] In the description of the present invention, unless otherwise specified, "plurality" means two or more than two; the orientations or positional relationships indicated by the terms "upper", "lower", "left", "right", "inner", "outer", "front end", "rear end", "head", "tail", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the mechanism or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.
[0057] Embodiment 1, combination Figures 1 to 7 A turtle-like robot for monitoring submarine pipeline leakage includes a trunk 1, a head 2, a top cover 3, a pectoral fin plate 4, a tail fin plate 5, a pectoral fin driving device, a center of gravity adjustment mechanism 6 and a detection control unit. The trunk 1 is a bilaterally symmetrical elliptical cone shell structure. Specifically, the trunk 1 includes a web 11, an upper plate 12 and an annular side plate 13. The upper plate 12 is located directly above the web 11. The lower end of the annular side plate 13 is fixedly connected to the outer edge of the web 11, and the upper end thereof is fixedly connected to the outer edge of the upper plate 12. The annular side plate 13 is an annular shell that gradually shrinks inward from bottom to top. The upper plate 12 and the web 11 respectively close the upper and lower ends of the annular side plate 13.
[0058] The top cover 3 is in the shape of a spherical shell that matches the top of the trunk 1. It is fixedly installed above the trunk 1, and a sealed upper chamber 31 is formed inside. Specifically, an annular sealing ring is embedded in the upper end surface of the annular side plate 13, and the top cover 3 is buckled on the top of the annular side plate 13 and is detachably fixedly connected by bolts. The bottom of the top cover 3 is sealed with the upper end surface of the annular side plate 13 through the annular sealing ring.
[0059] The middle and rear part of the trunk 1 is a sealed lower chamber 101, and the inner front part of the annular side plate 13 has a vertically arranged partition 14, and the partition 14 cooperates with the web 11, the upper plate 12 and the annular side plate 13 to form the lower chamber 101. The front part of the annular side plate 13 has two strip holes 131 opened symmetrically on the left and right, and the strip holes 131 are located on the front side of the partition 14.
[0060] The head 2 is arranged in front of the torso 1 through the neck 21. Specifically, the head 2 includes a light-transmitting cover shell and a frame. The frame is located on the inner side of the light-transmitting cover shell, and its rear end is connected to the front end of the neck 21. The rear end of the neck 21 is connected to the front end of the torso 1.
[0061] The detection control unit includes a visual sensor, an image recognition module, a communication module, an altimeter, a sonar sensor, a nine-axis attitude sensor and a controller. The visual sensor is arranged on the head 2, and the image recognition module, the communication module, the altimeter, the sonar sensor, the nine-axis attitude sensor and the controller are installed inside the upper chamber 31. The nine-axis attitude sensor obtains the position information and attitude information of the turtle-like robot during the swimming process, and sends it to the controller. The controller controls the movement of the two pectoral fins and the two tail fins of the turtle-like robot and adjusts the center of gravity through instructions.
[0062] There are two visual sensors, which are symmetrically arranged on both sides of the transparent cover and installed on the frame. The signal end of the visual sensor is connected to the controller through the image recognition module. When the turtle-like robot swims along the submarine pipeline, the visual sensor continuously collects image information of the submarine pipeline below and transmits the image information to the image recognition module. The image recognition module determines whether the submarine pipeline at that location has a leak and sends the recognition result to the controller. The controller sends the result to the receiving terminal on land through the communication module, and the staff on land take corresponding measures.
[0063] There are two pectoral fin plates 4 symmetrically arranged on the left and right sides of the outside of the trunk 1 , and there are two pectoral fin drive devices, which are installed at the front of the inner side of the trunk 1 and are arranged symmetrically about the central axis of the trunk 1 .
[0064] The pectoral fin drive device is located outside the front side of the lower chamber 101, and the pectoral fin drive device includes a first servo 41, a transmission mechanism and an angle adjustment assembly. The transmission mechanism is arranged on the inner side of the trunk 1 through a bearing bracket 42. The output end of the first servo 41 is connected to the end of the pectoral fin plate 4 on the same side through the transmission mechanism, driving the pectoral fin plate 4 to swing back and forth. During the swinging of the pectoral fin plate 4 back and forth, the angle adjustment assembly controls the angle posture of the pectoral fin plate 4.
[0065] Specifically, the transmission mechanism includes a vertical shaft 43, a horizontal shaft 44, a crank 45, a connecting rod 46 and a rocker arm 47. The vertical shaft 43 is rotatably mounted on the bearing bracket 42, and one end of the rocker arm 47 is fixedly connected to the lower end of the vertical shaft 43. The first steering gear 41 is located at the rear side of the vertical shaft 43, and is fixedly mounted on the upper surface of the web 11 of the trunk 1, or the upper surface of the web 11 is fixed by the bearing bracket 42. One end of the crank 45 is fixedly connected to the output shaft of the first steering gear 41, and the other end is rotatably connected to one end of the connecting rod 46, and the other end of the connecting rod 46 is rotatably connected to the other end of the rocker arm 47. In the working state, the output shaft of the first steering gear 41 drives the crank 45 to rotate horizontally around its axis, and the other end of the crank 45 drives the vertical shaft 43 to reciprocate within a certain angle range through the connecting rod 46 and the rocker arm 47.
[0066] The upper end of the vertical shaft 43 is fixedly connected with a first shaft sleeve 48, the horizontal shaft 44 is inserted into the first shaft sleeve 48 and rotates with it, one end of the horizontal shaft 44 passes through the trunk 1 and is fixedly connected to one end of the pectoral fin plate 4 on the same side, the crank 45 rotates around the output shaft of the first steering gear 41, and the vertical shaft 43 is driven to rotate through the crank 45 and the connecting rod 46, and the vertical shaft 43 drives the horizontal shaft 44 and the pectoral fin plate 4 to swing back and forth. The pectoral fin plate 4 adopts a crescent-shaped flat plate, one end of which is fixedly connected to one end of the horizontal shaft 44, and the vertical shaft 43 drives the horizontal shaft 44 and the pectoral fin plate 4 to swing back and forth around the axis of the vertical shaft 43 to perform paddling, so that the turtle-like robot can move forward on the seabed.
[0067] The angle adjustment assembly includes an arc guide rail 71, a drive arm 72, a guide rod 73 and a ball head 74. The arc guide rail 71 is in the shape of a 1 / 4 arc and is located on the side where the two bearing brackets 42 are close to each other. The arc guide rail 71 is fixedly connected to the bearing bracket 42 through a guide rail frame 711, and both end surfaces of the arc guide rail 71 adopt a rounded structure.
[0068] The driving arm 72 is an L-shaped structure, and the middle position of its vertical part is fixedly connected to the other end of the horizontal axis 44. The ball head 74 is installed at the end of the horizontal part of the driving arm 72. When the horizontal axis 44 reciprocates around the vertical axis 43, the ball head 74 slides with the upper surface of the arc guide rail 71.
[0069] The guide rod 73 is fixedly connected to the top of the first sleeve 48 through a guide rod bracket 75, which is located directly above the other end of the horizontal axis 44. A through hole is opened at the upper end of the vertical part of the driving arm 72. The guide rod 73 is movably inserted into the through hole, and both ends of the guide rod 73 are fixedly connected to the guide rod bracket 75. There is an annular gap between the guide rod 73 and the inner wall of the through hole.
[0070] During the swinging process of the horizontal axis 44, the vertical parts of the driving arm 72 are deflected to both sides respectively, and the ball head 74 slides along the upper surface of the arc guide rail 71. When the ball head 74 reaches the end of the arc guide rail 71, the driving arm 72 changes the deflection state to adjust the angle of the pectoral fin 4. During the sliding process of the ball head 74 along the upper surface of the arc guide rail 71, the guide rod 73 limits the upper end of the driving arm 72 to maintain the posture of the pectoral fin 4. Specifically, during the swinging process of the pectoral fin 4 from front to back, the front side is kept tilted downward at 45 degrees to paddle, driving the turtle-like robot forward, and during the swinging process of the pectoral fin 4 from back to front, it is kept in a horizontal state to reduce the resistance of the water.
[0071] The center of gravity adjustment mechanism 6 includes a counterweight 61, a linear guide mechanism and a servo motor 62. The counterweight 61 is slidably arranged in the lower chamber 101. The servo motor 62 drives the counterweight 61 to move forward and backward through the linear guide mechanism. Specifically, the linear guide mechanism includes a linear guide rail 63, a guide slider 64 and a lead screw 65. The linear guide rail 63 is longitudinally installed at the inner bottom of the trunk 1. The bottom of the counterweight 61 is slidably matched with the linear guide rail 63 through the guide slider 64.
[0072] The lead screw 65 is passed through the inner side of the counterweight 61 and is threadedly engaged with the counterweight 61. The front end of the lead screw 65 is rotationally engaged with the trunk 1 through the bearing seat 66. The rear end is coaxially fixedly connected to the output shaft of the servo motor 62. The signal end of the servo motor 62 is connected to the controller.
[0073] There are two tail fins 5, which are symmetrically arranged on the left and right sides behind the trunk 1. The front end of each tail fin 5 is rotatably connected to the rear end of the trunk 1 through a second steering gear 51, and each second steering gear 51 drives the corresponding tail fin 5 to swing up and down around its execution end.
[0074] Specifically, the left and right sides of the rear end of the trunk 1 have two concave mounting parts 15, and a U-shaped bracket 52 is bolted to the rear side of each mounting part 15. The second steering gear 51 adopts a double-output shaft steering gear, and the housing of the second steering gear 51 is fixedly connected to the front end of the tail fin plate 5 by bolts through the steering gear bracket 53, and the two output shaft ends are fixedly connected to the U-shaped bracket 52. The signal end of the second steering gear 51 is connected to the controller for communication. In the working state, the second steering gear 51 drives the tail fin plate 5 to swing back and forth around its output shaft, driving the turtle-like robot to move forward in the water.
[0075] When the turtle-like robot is in the forward mode, the two pectoral fins 4 on the left and right sides thereof keep paddling synchronously under the drive of the pectoral fin driving device, and the turtle-like robot moves forward. When it is necessary to accelerate forward, the two second steering gears 51 respectively drive the tail fin 5 to swing up and down to accelerate the forward movement of the turtle-like robot.
[0076] When the turtle-like robot is in the left-turn mode, the pectoral fin plate 4 on its right side reciprocates back and forth to achieve paddling, and the pectoral fin plate 4 on its left side remains horizontal, so that the turtle-like robot turns left.
[0077] When the turtle-like robot is in the right-turn mode, the pectoral fin plate 4 on its left side reciprocates back and forth to achieve paddling, and the pectoral fin plate 4 on its right side remains horizontal, so that the turtle-like robot turns right.
[0078] When the turtle-like robot is in the sinking mode, the servo motor 61 drives the screw 65 to rotate, and moves the counterweight 61 to the front end of the screw 65, so that the head of the turtle-like robot sinks and the tail rises. At the same time, the two pectoral fins 4 of the turtle robot are 45° downward to the horizontal direction, and the two tail fins 5 of the turtle-like robot are swung up and down to realize the sinking movement.
[0079] When the turtle-like robot is in the floating mode, the servo motor 61 drives the screw 65 to rotate, and moves the counterweight 61 to the rear end of the screw 65, so that the head of the turtle-like robot is tilted up and the tail is sunken. At the same time, the two pectoral fins 4 of the turtle robot are upward at 45 degrees to the horizontal direction, and the two tail fins 5 of the turtle-like robot are swung up and down to achieve floating movement.
[0080] Example 2: In the actual pipeline leakage monitoring process, the turtle-like robot needs to move along the top of the submarine pipeline and maintain a relative height h with the submarine pipeline to ensure that the visual sensor can effectively monitor whether the submarine pipeline is leaking. If it is too far away from the pipeline on the seabed, it will easily affect the pipeline leakage monitoring effect of the sensor above the turtle-like robot; if it is too close to the submarine pipeline, the movement of the turtle-like robot may cause the stirring of mud and sand, which may also affect the monitoring effect. At the same time, due to the uneven seabed terrain and the presence of some obstacles, the turtle-like robot is required to have five motion modes of moving forward, floating, sinking, turning left, and turning right. The turtle-like robot patrols along the submarine pipeline to achieve effective monitoring of the leakage points of the submarine pipeline.
[0081] A multi-modal control method for a turtle-like robot is used for the above-mentioned turtle-like robot for submarine pipeline leakage monitoring. Before use, the turtle-like robot needs to be trained in multi-modal control so that it can accurately move along the submarine pipeline line.
[0082] The multi-modal control method of the turtle-like robot comprises the following steps: Step 1, setting the coordinate points of the route of the turtle-like robot to track the submarine pipeline as ((x'1, y'1, z'1), (x'2, y'2, z'2), ..., (x' n ,y' n ,z' n )), where (x' n ,y' n,z' n ) is the coordinate of the tracking path point of the turtle-like robot at the nth time point;
[0083] Step 2: Obtain the current actual position coordinates of the turtle-like robot according to the altimeter and the sonar sensor, and perform difference calculation between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path point at this time;
[0084] Step 3, establish a state-action value table of the turtle-like robot, and obtain the action output in the state-action value table according to the difference calculated in step 2;
[0085] The process of obtaining the state-action value table includes the following steps:
[0086] Step 3.1: Define the current actual position coordinates of the turtle-like robot as η k =(x k ,y k ,z k ), where x k ,y k and z k are the actual current position coordinates of the turtle-like robot on the x-axis, y-axis, and z-axis at the kth time point;
[0087] Step 3.2: The coordinates of the tracking path points of the turtle-like robot at k time points are η' k =(x' k ,y' k ,z' k ), where x' k ,y' k and z' k are the coordinates of the tracking path points of the turtle-like robot at the kth time point on the x-axis, y-axis, and z-axis respectively;
[0088] Step 3.3, define the state space S, where S = [s1, s2, ..., s n ,…,s N ], N is a positive integer, N represents the number of states contained in S, s n =[E xn ,E yn ,E zn ] is the nth state in S, E xn Indicates the error interval in the x-axis direction, E yn Indicates the error interval in the y-axis direction, E zn Represents the error range in the z-axis direction;
[0089] The difference between the actual position coordinates of the turtle-like robot at the kth time point and the coordinates of the tracking path point is: e k =[x k -x'k ,y k -y' k ,z k -z' k ], let e k =[e x ,e y ,e z ], judge e k Located in the state, e x ∈E xn , e y ∈E yn , e z ∈E zn When k ∈s n , indicating that the turtle-like robot is in the nth state at this time;
[0090] Step 3.4: Define the action space A = [a1, a2, ..., a m ,…,a M ], M is a positive integer, indicating the number of actions contained in A; define action group a m =[F m1 ,θ m1 ,F m2 ,θ m2 ,F m3 ,θ m3 ,F m4 ,θ m4 ,L fr ], where F m1 is the thrust of the first servo on the left side when performing the mth action, θ m1 F is the rudder angle of the first servo on the left side when executing the mth action group; m2 is the thrust of the first servo on the right side when performing the mth action, θ m2 F is the rudder angle of the first servo on the right side when executing the mth action group; m3 is the thrust of the second servo on the left side when performing the mth action, θ m3 F is the rudder angle of the second servo on the left when executing the mth action group; m4 is the thrust of the second servo on the right side when performing the mth action, θ m4 is the rudder angle of the second servo on the right side when executing the mth action group; L fr represents the position of the weight block when performing the mth action, where L fr 1 indicates that the counterweight is located at the front end of the lead screw. At this time, the tail of the turtle robot is tilted up; L fr -1 indicates that the counterweight is at the rear end of the lead screw. At this time, the head of the turtle robot is tilted up. fr0 indicates that the counterweight is in the middle of the lead screw. At this time, the turtle robot is in a horizontal balance state. By setting action group a m The 9 parameters of can realize the motion modes of the turtle-like robot: sinking, floating, moving forward, turning left, and turning right. The value of M affects the control accuracy of the motion mode of the turtle-like robot. The larger the M, the more actions can be selected, and the higher the control accuracy.
[0091] Step 3.5: Set the reward function and calculate the reward of the turtle-like robot in the nth state s n Take action a m Then the reward value r obtained according to the reward function n+1 ;
[0092] The calculation formula of the reward function is: n+1 =r e +r n +r o ;
[0093] In the formula, r e To reach the new state s n+1 The reward value when Where e xmax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path points in the x-axis direction, e ymax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path points in the y-axis direction, e zmax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path point in the z-axis direction;
[0094] r n is the reward obtained when reaching the path tracking point,
[0095] r o For collision rewards, L is the new state s reached by the turtle-like robot n+1 L0 is the straight-line distance from the obstacle, and L0 is the set safety distance.
[0096] Step 3.6: Obtain the state-action value table of the turtle-like robot according to the Q-Learning algorithm, and update the state-action value table through training to make it converge. The state-action value table is shown in Table 1, which records the state-action value after taking action a in state s.
[0097] Table 1 State-action value table
[0098]
[0099] The calculation formula of the Q-Learning algorithm is:
[0100]
[0101] In the formula, Q(s n ,a m ) is the state-action value; α is the learning rate, 0<α≤1; γ is the discount factor; r n+1 Indicates in s n State takes action a m Then transfer to the new state s n+1 The reward value obtained when Indicates that in state s n+1 The maximum value of all possible actions corresponding to the state-action value; during the training process, the state-action value table is continuously updated according to the Q-Learning algorithm formula until all values in the state-action value table converge, the training is completed, and the final state-action value table is formed.
[0102] Step 4: According to the action output obtained in step 3, the turtle-like robot is controlled to realize multi-modal motion on the seabed and track the submarine pipeline line.
[0103] The turtle-like robot selects the combination of servo thrust and rudder angle "actions" to be executed according to the greedy strategy. The environment will respond to the "actions", that is, give corresponding reward values according to the reward function. After the turtle-like robot takes action, it will transfer from the current state to the next state. In the next state, the turtle-like robot will continue to select the action combination to be executed and obtain the environmental feedback reward, and will also transfer to the new state. Repeating the above process forms an interactive learning process between the action combination and the environment.
[0104] Afterwards, the trained motion output is used to control the turtle-like robot to achieve multimodal motion, thereby realizing the path tracking control of the turtle-like robot. The motion sequence a taken by the turtle-like robot during the path tracking process m →a m+1 →a m+2 ···The corresponding modal sequence is the modal switching process. For example, after a m , a m+1 , a m+2 After three actions, the turtle robot reaches the first path tracking point. Assume a m The corresponding mode is forward, a m+1 The corresponding mode is left turn, a m+2 The corresponding mode is forward, and after switching between forward, left turn, and forward modes, the turtle-like robot can reach the first path tracking point.
[0105] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0106] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.
Claims
1. A turtle-like robot for monitoring submarine pipeline leakage, characterized in that: It includes a trunk, a head, a top cover, a pectoral fin plate, a caudal fin plate, a pectoral fin drive device, a center of gravity adjustment mechanism and a detection control unit. The trunk is an elliptical shell structure, and the top cover is a hemispherical shell arranged above the trunk, and a sealed upper chamber is formed inside the top cover. The detection control unit includes a visual sensor, an image recognition module, a communication module, an altimeter, a sonar sensor, a nine-axis attitude sensor and a controller. The head is arranged in front of the torso through the neck, the visual sensor is arranged on the head, and the image recognition module, the communication module, the altimeter, the sonar sensor, the nine-axis attitude sensor and the controller are arranged inside the upper chamber; There are two pectoral fin plates, which are symmetrically arranged on the left and right sides of the front of the trunk; there are two pectoral fin drive devices, which are arranged on the front of the trunk and are symmetrically arranged on the left and right sides; The pectoral fin drive device includes a first steering gear, a transmission mechanism and an angle adjustment component. The transmission mechanism is arranged on the inner side of the trunk through a bearing bracket. The output end of the first steering gear is connected to the end of the pectoral fin plate on the same side through the transmission mechanism to drive the pectoral fin plate to swing back and forth. During the swinging process of the pectoral fin plate, the angle adjustment component controls the angle posture of the pectoral fin plate. The middle and rear part of the trunk is a sealed lower chamber, the pectoral fin drive device is located outside the front side of the lower chamber, the center of gravity adjustment mechanism includes a counterweight block, a linear guide mechanism and a servo motor, the counterweight block is slidably arranged in the lower chamber, and the servo motor drives the counterweight block to move forward and backward through the linear guide mechanism; There are two tail fins, which are symmetrically arranged on the left and right sides of the rear of the trunk. The front end of each tail fin is rotatably connected to the rear end of the trunk through a second steering gear, and each second steering gear drives the corresponding tail fin to swing up and down around its execution end.
2. The turtle-like robot for monitoring submarine pipeline leakage according to claim 1, characterized in that: The trunk includes a belly plate, an upper plate and an annular side plate, wherein the upper plate is located directly above the belly plate, the lower end of the annular side plate is fixedly connected to the outer edge of the belly plate, and the upper end thereof is fixedly connected to the outer edge of the upper plate; A partition is provided at the inner front part of the annular side plate, and the partition cooperates with the web plate, the upper plate and the annular side plate to form the lower chamber; The front part of the annular side plate is provided with two strip holes which are symmetrically opened on the left and right sides, and the strip holes are located on the front side of the partition plate.
3. The turtle-like robot for monitoring submarine pipeline leakage according to claim 1, characterized in that: The head includes a light-transmitting cover shell and a frame, wherein the frame is located inside the light-transmitting cover shell, and the rear end of the frame is connected to the front end of the neck, and the rear end of the neck is connected to the front end of the trunk; There are two visual sensors, which are symmetrically arranged on both sides of the inside of the light-transmitting cover shell and are installed on the frame. The signal end of the visual sensor is connected to the controller through communication with the image recognition module.
4. The turtle-like robot for monitoring submarine pipeline leakage according to claim 1, characterized in that: The transmission mechanism includes a vertical shaft, a horizontal shaft, a crank, a connecting rod and a rocker arm. The vertical shaft is rotatably arranged on the bearing bracket, and one end of the rocker arm is fixedly connected to the lower end of the vertical shaft. The first steering gear is fixedly installed at the rear side of the vertical shaft, one end of the crank is fixedly connected to the output shaft of the first steering gear, and the other end is movably connected to the other end of the rocker arm through the connecting rod; The upper end of the vertical shaft is fixedly connected to a first shaft sleeve, the horizontal shaft is inserted into the first shaft sleeve and rotates with it, one end of the horizontal shaft passes through the trunk and is fixedly connected to one end of the pectoral fin plate on the same side, the crank rotates around the output shaft of the first servo, and the vertical shaft is driven to rotate through the crank and the connecting rod, and the vertical shaft drives the horizontal shaft and the pectoral fin plate to swing back and forth.
5. The turtle-like robot for monitoring submarine pipeline leakage according to claim 4, characterized in that: The angle adjustment assembly includes an arc guide rail, a driving arm, a guide rod and a ball head. The arc guide rail is in the shape of a quarter arc and is located on the side where the two bearing brackets are close to each other. It is fixedly connected to the bearing bracket through a guide rail frame. Both end surfaces of the arc guide rail adopt a rounded structure. The driving arm is an L-shaped structure, the middle position of the vertical part of which is fixedly connected to the other end of the horizontal axis, and the ball head is installed at the end of the horizontal part of the driving arm. When the horizontal axis reciprocates around the vertical axis, the ball head slides and cooperates with the upper surface of the arc guide rail; The guide rod is fixedly connected to the top of the first sleeve through a guide rod bracket, which is located directly above the other end of the horizontal axis. A through hole is opened at the upper end of the vertical part of the driving arm. The guide rod is movably inserted into the through hole, and both ends of the guide rod are fixedly connected to the guide rod bracket. There is an annular gap between the guide rod and the inner wall of the through hole.
6. The turtle-like robot for monitoring submarine pipeline leakage according to claim 1, characterized in that: The linear guide mechanism includes a linear slide rail, a guide slider and a lead screw. The linear slide rail is longitudinally installed at the inner bottom of the trunk, and the bottom of the counterweight block is slidably matched with the linear slide rail through the guide slider. The lead screw is passed through the inner side of the counterweight block and matched with its thread. The front end of the lead screw is matched with the rotation of the trunk through the bearing seat, and the rear end is coaxially fixedly connected with the output shaft of the servo motor. The signal end of the servo motor is connected to the controller.
7. The turtle-like robot for monitoring submarine pipeline leakage according to claim 1, characterized in that: The left and right sides of the rear end of the trunk are provided with two concave mounting parts, and a U-shaped bracket is fixed to the rear side of each mounting part; The second servo adopts a double-output shaft servo, the outer shell of the second servo is fixedly connected to the front end bolts of the tail fin through a servo bracket, the two output shaft ends are fixedly connected to the U-shaped bracket, the signal end of the second servo is connected to the controller for communication, and in the working state, the second servo drives the tail fin to swing back and forth around its output shaft.
8. A multi-modal control method for a turtle-like robot, characterized in that: A turtle-like robot for monitoring submarine pipeline leakage as claimed in any one of claims 1 to 7; The multi-modal control method of the turtle-like robot comprises the following steps: Step 1: Set the coordinate points of the route that the turtle robot follows the submarine pipeline to ((x'1, y'1, z'1), (x'2, y'2, z'2), ..., (x' n ,y' n ,z' n )), where (x' n ,y' n ,z' n ) is the coordinate of the tracking path point of the turtle-like robot at the nth time point; Step 2: Obtain the current actual position coordinates of the turtle-like robot according to the altimeter and the sonar sensor, and perform difference calculation between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path point at this time; Step 3, establish a state-action value table of the turtle-like robot, and obtain the action output in the state-action value table according to the difference calculated in step 2; Step 4: According to the action output obtained in step 3, the turtle-like robot is controlled to realize multi-modal motion on the seabed and track the submarine pipeline line.
9. The multi-modal control method of a turtle-like robot according to claim 8, characterized in that: The process of obtaining the state-action value table includes the following steps: Step 3.1: Define the current actual position coordinates of the turtle-like robot as η k =(x k ,y k ,z k ), where x k ,y k and z k are the actual current position coordinates of the turtle-like robot on the x-axis, y-axis, and z-axis at the kth time point; Step 3.2: The coordinates of the tracking path points of the turtle-like robot at k time points are η' k =(x' k ,y' k ,z' k ), where x' k ,y' k and z' k are the coordinates of the tracking path points of the turtle-like robot at the kth time point on the x-axis, y-axis, and z-axis respectively; Step 3.3, define the state space S, where S = [s1, s2, ..., s n ,…,s N ], N is a positive integer, N represents the number of states contained in S, s n =[E xn ,E yn ,E zn ] is the nth state in S, E xn Indicates the error interval in the x-axis direction, E yn Indicates the error interval in the y-axis direction, E zn Represents the error range in the z-axis direction; The difference between the actual position coordinates of the turtle-like robot at the kth time point and the coordinates of the tracking path point is: e k =[x k -x' k ,y k -y' k ,z k -z' k ], let e k =[e x ,e y ,e z ], judge e k Located in the state, e x ∈E xn , e y ∈E yn , e z ∈E zn When k ∈s n , indicating that the turtle-like robot is in the nth state at this time; Step 3.4: Define the action space A = [a1, a2, ..., a m ,…,a M ], M is a positive integer, indicating the number of actions contained in A; define action group a m =[F m1 ,θ m1 ,F m2 ,θ m2 ,F m3 ,θ m3 ,F m4 ,θ m4 ,L fr ], where F m1 is the thrust of the first servo on the left side when performing the mth action, θ m1 F is the rudder angle of the first servo on the left side when executing the mth action group; m2 is the thrust of the first servo on the right side when performing the mth action, θ m2 F is the rudder angle of the first servo on the right side when executing the mth action group; m3 is the thrust of the second servo on the left side when performing the mth action, θ m3 F is the rudder angle of the second servo on the left when executing the mth action group; m4 is the thrust of the second servo on the right side when performing the mth action, θ m4 is the rudder angle of the second servo on the right side when executing the mth action group; L fr represents the position of the weight block when performing the mth action, where L fr 1 indicates that the counterweight is located at the front end of the screw; L fr -1 indicates that the counterweight is located at the rear end of the screw; L fr 0 indicates that the counterweight is in the middle of the screw; Step 3.5: Set the reward function and calculate the reward of the turtle-like robot in the nth state s n Take action a m Then the reward value r obtained according to the reward function n+1 ; Step 3.6, obtaining the state-action value table of the turtle-like robot according to the Q-Learning algorithm; The calculation formula of the Q-Learning algorithm is: In the formula, Q(s n ,a m ) is the state-action value, α is the learning rate, 0<α≤1; γ is the discount factor; r n+1 Indicates in s n State takes action a m Then transfer to the new state s n+1 The reward value obtained when Indicates that in state s n+1 The maximum value among all possible actions corresponding to the state-action value; During the training process, the state-action value table is continuously updated according to the Q-Learning algorithm formula until all values in the state-action value table converge, the training ends, and the final state-action value table is formed.
10. The multi-modal control method of a turtle-like robot according to claim 9, characterized in that: The calculation formula of the reward function in step 3.5 is: n+1 =r e +r n +r o ; In the formula, r e To reach the new state s n+1 The reward value when Where e xmax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path points in the x-axis direction, e ymax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path points in the y-axis direction, e zmax is the maximum difference between the current actual position coordinates of the turtle-like robot and the coordinates of the tracking path point in the z-axis direction; r n is the reward obtained when reaching the path tracking point, r o For collision rewards, L is the new state s reached by the turtle-like robot t+1 L0 is the straight-line distance from the obstacle, and L0 is the set safety distance.
Citation Information
Cited By
Water area pollution monitoring and treatment integrated platform based on bionic robot
CN121454018A