Multi-diameter complex mechanism pipeline inspection and repair intelligent device

Abstract

The present application relates to the technical field of pipeline equipment inspection, repair and adaptive control, and specifically discloses a multi-diameter complex mechanism pipeline inspection and repair intelligent control device based on a multi-modal adaptive model, which comprises a welding gun repair device, a scanning device, a steering device, a cleaning gun, a driving device, a signal transceiving device and a multi-modal adaptive control system. The multi-modal adaptive control system comprises a hydraulic cylinder, a PLC control unit, a motor, a servo driver and a touch screen, etc. The hydraulic device is controlled by a rotary transformer, and the transformer is connected with a servo frequency converter driver to realize unit interconnection. The present application has the function of automatically cleaning the pipe wall and can move in a complex pipeline. After discovering a problem, the pipeline leak can be repaired at any time, and the effect is good. The system error and response are reduced, and the operation efficiency and control accuracy of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of pipeline equipment technology, and specifically discloses an intelligent device for inspecting and repairing pipelines with complex structures of multiple diameters. Background Technology

[0002] Currently, existing pipelines have complex structures and varied shapes, making it difficult for pipeline inspection devices to operate under conditions involving multiple diameters or changing diameters. Furthermore, the complex structures, especially in sections with diameter changes, bends, localized deformation, or foreign object accumulation, pose a significant challenge to the stable operation of these inspection devices. It is difficult to guarantee diverse inspection functions for different pipelines, and achieving 360° all-around detection during the inspection process is challenging.

[0003] Meanwhile, the internal environment of pipelines is complex. In these confined spaces, traditional inspection devices often experience problems such as jamming, slippage, or collision with the pipe wall due to insufficient mechanical self-adaptation or a single control strategy. This not only interrupts the inspection task but may also damage the inner wall of the pipeline and the device itself.

[0004] Although some devices for pipeline inspection exist, the intelligence and predictive capabilities of their control systems still need improvement. They struggle to comprehensively and accurately acquire dynamic information on pipeline morphology and device posture changes. Control models are often fixed, unable to adjust motion strategies online based on pipeline diameter variations, bend curvature, or sudden obstacles (such as sediment). Their anti-interference capabilities are weak: they exhibit poor dynamic response to fluid disturbances or local deformation within the pipe, easily leading to stability issues. Maintaining efficient and reliable operation in unpredictable dynamic environments is also difficult. Furthermore, even when a problem is detected, only a signal can be sent to indicate the leak; timely repair is not possible. Additionally, water or oil-contaminated pipes are frequently polluted, requiring internal cleaning to ensure accurate image recognition.

[0005] Therefore, there is a need to develop a device that can automatically clean pipe walls, adapt to complex pipelines, and repair leaks in real time to overcome the aforementioned problems. Summary of the Invention

[0006] The technical solution adopted in this invention is as follows: In a first aspect, the present invention provides an intelligent device for the inspection and repair of pipelines with complex structures of multiple diameters, including a welding torch repair device 1, a scanning device 2, a steering device 3, a cleaning torch 4, a driving device 5, and a signal transceiver device 6. The welding torch repair device 1 is connected to one end of the scanning device 2, and the other end of the scanning device 2 is connected to the steering device 3. The steering device 3 is connected to the cleaning torch 4. The cleaning torch 4 is hollow inside and has a through groove on its periphery. The driving device 5 is located inside the cleaning torch 4 and is connected to the steering device 3. The driving device 5 extends through the through groove to the outside of the periphery of the cleaning torch 4 and contacts the pipeline. The position of the device is adjusted by the driving device 5. The welding torch repair device 1, the scanning device 2, the steering device 3, the cleaning torch 4, and the driving device 5 are communicatively connected to the signal transceiver 6, which is used to locate the specific position and movement status of the pipeline fault repair device and related information on operation commands. The welding torch repair device 1 includes a welding torch 11, a rotary gear device 12, and an axial rotating seat 13; the welding torch 11 is fixed on the rotary gear device 12, and the rotary gear device 12 can be set to a fixed angle after rotating into position to prevent rotation; the end of the rotary gear device 12 is connected to the rotating seat 13 through a universal joint. The steering device 3 includes several hydraulic cylinders, which are equidistantly arranged along the circumferential edges of the scanning device 2 and the driving device 5. The hydraulic cylinders are connected to telescopic rods. The angle of the driving device 5 is changed according to the different extension lengths of the different hydraulic cylinders to complete the turning function. The cleaning mechanism 4 includes a slide 41, a cleaning head 42, a water tank 43, and a water pipe ring 44. The cleaning mechanism 4 is provided with a base connected to the steering device 3, and several slides 41 are vertically arranged on the base. The water tank 43 and the water pipe ring 44 are arranged on the slides 41 and move along the slides 41. Several cleaning heads 42 are arranged on the outer side of the water pipe ring 44. The cleaning heads 42, the water tank 43, and the water pipe ring 44 are connected and water is supplied by the water tank 43.

[0007] The drive device 5 includes several front wheel drive devices 51 and rear wheel drive devices 52; both the front wheel drive device 51 and the rear wheel drive device 52 are provided with drive wheels 521 and rotating arms. A telescopic structure 53 is connected to the middle of the rotating arm. The length of the spring is adjusted according to different pipe inner diameters. Through the movement of the telescopic structure 53, the drive wheel 521 is made to fit against the pipe wall.

[0008] Preferably, the scanning device 2 includes a scanning fixing device 21, a scanning protective cover 22, and a scanning device instrument 23; the scanning protective cover 22 is a transparent cover, and the scanning device instrument 23 is located inside the scanning protective cover 22; the scanning fixing device 21 is located outside the scanning protective cover 22, and one end passes through the scanning protective cover 22 and is fixedly connected to the scanning device instrument 23. The other end of the scanning fixture 21 is connected to the axial rotating seat 13.

[0009] Preferably, the drive wheel 521 uses a rubber tire with raised dots to achieve an anti-slip function.

[0010] Preferably, the drive wheel 521 is a caster wheel.

[0011] Preferably, the telescopic structure 53 adopts a spring rod structure.

[0012] Preferably, the cleaning mechanism 4 includes a slide 41, a cleaning head 42, a water tank 43, and a water pipe ring 44; the cleaning mechanism 4 is provided with a base connected to the steering device 3, and a plurality of slides 41 are vertically arranged on the base; the water tank 43 and the water pipe ring 44 are arranged on the slides 41 and move along the slides 41; the water pipe ring 44; a plurality of cleaning heads 42 are arranged on the outer side; the cleaning heads 42, the water tank 43 and the water pipe ring 44 are connected and water is supplied by the water tank 43.

[0013] Furthermore, the cleaning head 42 is a retractable head.

[0014] Furthermore, the water pipe ring 44 is a rotatable mechanism.

[0015] Preferably, the steering device 3 and the drive device 5 are connected by a hinge.

[0016] Preferably, the steering device 3 is provided with three sets of hydraulic cylinders, which are distributed at 120° intervals.

[0017] Preferably, the front-wheel drive device 51 and the rear-wheel drive device 52 are each provided in three groups, which are distributed at 120° equidistant intervals.

[0018] Secondly, the present invention provides a control method for an intelligent device for inspecting and repairing multi-diameter complex pipeline structures, comprising the following steps: S1. Task decision and data acquisition steps: Receive the inspection task instruction and acquire real-time images and defect data of the inner wall of the pipe through the scanning device (2); at the same time, acquire the motion status data of the device itself; S2, Multimodal decision-making step: Based on the real-time images, defect data and motion state data, the control system makes task decisions to determine whether to perform inspection, cleaning or repair work, and activates the corresponding control model and operation parameters. S3. Steps for generating cooperative control instructions: S31. Motion Coordination Command Generation: Based on the motion state data and pipeline geometry data, a model predictive control algorithm is used to generate coordinated motion control commands for the steering device (3) and the drive device (5) so that the device moves and turns along the pipeline. S32, Operation control instruction generation: In response to the task decision, a corresponding operation control instruction is generated; when the decision is a cleaning operation, an instruction is generated to control the extension, rotation and water spraying of the cleaning head (42) of the cleaning mechanism (4); when the decision is a repair operation, an instruction is generated to control the positioning, angle adjustment and start / stop of the welding torch (11) of the welding torch repair device (1). S4. Command execution and closed-loop feedback steps: The coordinated motion control command and the operation control command are respectively sent to the corresponding actuators to drive the device to complete the movement, cleaning or repair actions, and the control command is adjusted in real time based on sensor feedback.

[0019] Preferably, the generation of cooperative motion control commands in step S31 adopts a master-slave dynamic coupling strategy, specifically including: The desired trajectory is generated by the drive unit (5), and the steering unit (3) tracks this attitude through its hydraulic servo system and uses a dynamic coupling compensator based on the yaw angle coupling error. e c Calculate the hydraulic cylinder pressure compensation amount ΔP to eliminate tracking error.

[0020] Preferably, the dynamic coupling compensator calculates the pressure compensation amount ΔP according to the following formula: Coupling error e c = ψ leader ψ follower ; ψ_leader represents the desired yaw angle calculated by the drive unit (5) based on the current curvature of the pipeline or the planned path; ψ_follower represents the actual yaw angle measured in real time by the attitude sensor (such as a gyroscope) on the steering device (3); Where Kp, Kd, ​​and Ki are control gains; c p， c d This is the learning rate.

[0021] Preferably, the model predictive control algorithm in step S31 adopts a multimodal adaptive model predictive control scheme, including: A model library containing dynamic models of various typical pipeline operating conditions is constructed; the high-level decision-maker activates the prediction model that best matches the current operating condition from the model library based on real-time environmental perception data; and the cost function weights of the model prediction control algorithm are fine-tuned online using reinforcement learning.

[0022] Furthermore, in step S31, fuzzy adaptive decision-making is employed: The pipe curvature κ, hydraulic cylinder pressure deviation ΔP, and turning yaw angle error e are considered. ψ As input variables; The hydraulic cylinder cooperative gain K is calculated and output using fuzzy inference rules. c Along with the damping compensation coefficient β, it is used to adjust the control parameters in real time.

[0023] Furthermore, Fuzzy yaw error e ψ and curvature k Using Gaussian membership functions: Defuzzification output cooperative gain: w j For rule weights, M For the number of rules.

[0024] The beneficial effects of this invention are as follows: This invention features automatic pipe wall cleaning and adaptive movement in complex pipelines, enabling it to detect and repair leaks promptly with excellent results. The multimodal adaptive control system comprises hydraulic cylinders, a PLC control unit, motors, servo drivers, and a touchscreen. Three hydraulic devices form a group, controlled by a rotary transformer, which is connected to the servo frequency converter for unit interconnection. Three motors are connected to their corresponding transformers and then to the PLC control unit, which acts as the master station, with the motion control unit as the slave station, forming a complete closed-loop control system. Simultaneously, a hydraulic and motor coordinated control method is provided for a multi-diameter complex pipeline inspection and repair control device. This invention reduces system errors and responsiveness, improving system operating efficiency and control accuracy. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the device structure in Embodiment 1 of the present invention.

[0026] Figure 2 This is a schematic diagram of the welding torch repair device according to Embodiment 1 of the present invention.

[0027] Figure 3 This is a schematic diagram of the scanning device structure according to Embodiment 1 of the present invention.

[0028] Figure 4 This is a schematic diagram of the cleaning gun structure in Embodiment 1 of the present invention.

[0029] Figure 5 This is a schematic diagram of the drive device structure in Embodiment 1 of the present invention.

[0030] Figure 6 This is a schematic diagram of the control process in Embodiment 1 of the present invention.

[0031] Figure 7 This is a schematic diagram of the information interaction process in Embodiment 1 of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the following embodiments. It should be noted that this invention is not limited to the following embodiments.

[0033] Example 1 An intelligent device for inspecting and repairing pipelines with complex structures of multiple diameters includes a welding torch repair device 1, a scanning device 2, a steering device 3, a cleaning torch 4, a driving device 5, and a signal transceiver device 6. The welding torch repair device 1 is connected to one end of the scanning device 2, and the other end of the scanning device 2 is connected to the steering device 3. The steering device 3 is connected to the cleaning torch 4. The cleaning torch 4 is hollow inside and has a through groove on its periphery. The driving device 5 is located inside the cleaning torch 4 and is connected to the steering device 3. The driving device 5 extends through the through groove to the outside of the periphery of the cleaning torch 4 and contacts the pipeline. The position of the device is adjusted by the driving device 5. The welding torch repair device 1, the scanning device 2, the steering device 3, the cleaning torch 4, and the driving device 5 are communicatively connected to the signal transceiver 6, which is used to locate the specific position and movement status of the pipeline fault repair device and related information on operation commands. The welding torch repair device 1 includes a welding torch 11, a rotary gear device 12, and an axial rotating seat 13; the welding torch 11 is fixed on the rotary gear device 12, and the rotary gear device 12 can be set to a fixed angle after rotating into position to prevent rotation; the end of the rotary gear device 12 is connected to the rotating seat 13 through a universal joint. The scanning device 2 includes a scanning fixing device 21, a scanning protective cover 22, and a scanning device instrument 23; the scanning protective cover 22 is a transparent cover, and the scanning device instrument 23 is located inside the scanning protective cover 22; the scanning fixing device 21 is located outside the scanning protective cover 22, with one end passing through the scanning protective cover 22 and fixedly connected to the scanning device instrument 23; the other end of the scanning fixing device 21 is connected to the axial rotating seat 13.

[0034] The steering device 3 includes three hydraulic cylinders, which are equidistantly arranged along the circumferential edges of the scanning device 2 and the driving device 5. The hydraulic cylinders are connected to telescopic rods. The angle of the driving device 5 is changed according to the different extension lengths of the different hydraulic cylinders to complete the turning function. The steering device 3 is equipped with three sets of hydraulic cylinders, which are equidistantly distributed at 120° intervals.

[0035] The cleaning mechanism 4 includes a slide rail 41, a cleaning head 42, a water tank 43, and a water pipe ring 44. The cleaning mechanism 4 has a base connected to the steering device 3, and several slide rails 41 are vertically arranged on the base. The water tank 43 and the water pipe ring 44 are disposed on the slide rails 41 and move along the slide rails 41. Several cleaning heads 42 are arranged on the outer side of the water pipe ring 44. The cleaning heads 42, water tank 43, and water pipe ring 44 are connected and water is supplied by the water tank 43. The cleaning head 42 is a retractable head, and the water pipe ring 44 is a rotatable mechanism.

[0036] The drive unit 5 includes several front-wheel drive units 51 and rear-wheel drive units 52. Each front-wheel drive unit 51 and rear-wheel drive unit 52 is equipped with a drive wheel 521 and a rotating arm. A telescopic structure 53 is connected to the middle of the rotating arm. The length of the spring is adjusted according to different pipe inner diameters. Through the movement of the telescopic structure 53, the drive wheel 521 is brought into contact with the pipe wall. There are three sets of each of the front-wheel drive units 51 and rear-wheel drive units 52, evenly distributed at 120° intervals.

[0037] The drive wheel 521 is a universal wheel device, which uses rubber tires with raised dots to achieve anti-slip function.

[0038] The telescopic structure 53 adopts a spring rod structure.

[0039] The three hydraulic cylinders of steering device 3 are evenly distributed circumferentially at 120°, achieving adaptive contact with the inner wall of the pipe and stable turning. Let the shape function of the inner wall of the pipe be... R (θ, z), where θ is the circumferential angle and z is the axial position. The extension / retraction amount of each hydraulic cylinder is denoted as... x i (t)( i =1,2,3), the contact force between the hydraulic cylinder and the inner wall of the pipe is denoted as F i (t). The centering condition of the pipeline inspection device within the pipeline can be expressed as: The dynamic equations of the steering mechanism hydraulic cylinder are established to achieve adaptive control of the system. The dynamics of the hydraulic system can be described as follows: in: m For equivalent quality, c The damping coefficient is... k The elastic coefficient, A The effective area of ​​the hydraulic cylinder piston. P For hydraulic pressure, d i ( t This represents an external disturbance. Flexible steering component hydraulic system: The joint is driven by three hydraulic cylinders to achieve flexible steering of the intelligent device for pipeline inspection and repair of complex mechanisms.

[0040] Hydraulic cylinder force model: in, P i For the first i The pressure of each hydraulic cylinder A i The piston rod area is... k f The coefficient of friction, B v It uses viscous damping. The extension / retraction length and length difference of each hydraulic cylinder adapt to changes in pipe diameter, assisting in stability during steering.

[0041] The relationship between the length difference between hydraulic cylinders and the displacement of the hydraulic cylinders is as follows: x 0 represents the initial displacement. L This refers to the telescopic length.

[0042] Hydraulic actuator model Hydraulic cylinder flow-pressure equation: C d For flow coefficient, A v For the valve orifice area, r This refers to the density of the hydraulic oil.

[0043] Piston dynamics: F load,i This refers to the reaction force of the pipe wall on the inspection device.

[0044] Inspection device steering dynamics Equations of motion in the yaw direction (based on the Newton-Euler equations): I z For rotational inertia, d i For the hydraulic cylinder's lever arm, i For the installation angle of the hydraulic cylinder, F fluid For fluid resistance (simplified based on the Navier-Stokes equations). F friction This refers to the frictional force of the pipe wall.

[0045] Fluid resistance model (incompressible flow) A front The area facing the airflow of the inspection device.

[0046] The collaborative control strategy in this embodiment is as follows: The control method includes the following steps: S1. Task decision and data acquisition steps: Receive the inspection task instruction and acquire real-time images and defect data of the inner wall of the pipe through the scanning device (2); at the same time, acquire the motion status data of the device itself; S2, Multimodal decision-making step: Based on the real-time images, defect data and motion state data, the control system makes task decisions to determine whether to perform inspection, cleaning or repair work, and activates the corresponding control model and operation parameters. S3. Steps for generating cooperative control instructions: S31. Motion Coordination Command Generation: Based on the motion state data and pipeline geometry data, a model predictive control algorithm is used to generate coordinated motion control commands for the steering device (3) and the drive device (5) so that the device moves and turns along the pipeline. S32, Operation control instruction generation: In response to the task decision, a corresponding operation control instruction is generated; when the decision is a cleaning operation, an instruction is generated to control the extension, rotation and water spraying of the cleaning head (42) of the cleaning mechanism (4); when the decision is a repair operation, an instruction is generated to control the positioning, angle adjustment and start / stop of the welding torch (11) of the welding torch repair device (1). S4. Command execution and closed-loop feedback steps: The coordinated motion control command and the operation control command are respectively sent to the corresponding actuators to drive the device to complete the movement, cleaning or repair actions, and the control command is adjusted in real time based on sensor feedback.

[0047] The generation of cooperative motion control commands in step S31 adopts a master-slave dynamic coupling strategy, specifically including: The desired trajectory is generated by the drive unit (5), and the steering unit (3) tracks this attitude through its hydraulic servo system and uses a dynamic coupling compensator based on the yaw angle coupling error. e c Calculate the hydraulic cylinder pressure compensation amount ΔP to eliminate tracking error.

[0048] The dynamic coupling compensator calculates the pressure compensation amount ΔP according to the following formula: Coupling error e c = ψ leader ψ follower ; ψ_leader represents the desired yaw angle calculated by the drive unit (5) based on the current curvature of the pipeline or the planned path; ψ_follower represents the actual yaw angle measured in real time by the attitude sensor (such as a gyroscope) on the steering device (3); Where Kp, Kd, ​​and Ki are control gains; c p， c d This is the learning rate.

[0049] The model predictive control algorithm in step S31 adopts a multimodal adaptive model predictive control scheme, including: A model library containing dynamic models of various typical pipeline operating conditions is constructed; the high-level decision-maker activates the prediction model that best matches the current operating condition from the model library based on real-time environmental perception data; and the cost function weights of the model prediction control algorithm are fine-tuned online using reinforcement learning.

[0050] In step S31, fuzzy adaptive decision-making is adopted: The pipe curvature κ, hydraulic cylinder pressure deviation ΔP, and turning yaw angle error e are considered. ψ As input variables; The hydraulic cylinder cooperative gain K is calculated and output using fuzzy inference rules. c Along with the damping compensation coefficient β, it is used to adjust the control parameters in real time.

[0051] Fuzzy yaw error e ψ and curvature k Using Gaussian membership functions: Defuzzification output cooperative gain: w j For rule weights, M For the number of rules.

[0052] This invention divides a complex pipeline inspection and repair intelligent device into a steering unit and a drive unit. The drive unit generates the desired trajectory based on the pipeline curvature, while the steering unit tracks the drive unit's posture via a hydraulic servo system. A multimodal adaptive model predictive control (MA-MPC) scheme is employed to construct a dynamic model library encompassing various typical pipeline operating conditions. A high-level decision-maker activates the best-matching predictive model based on real-time environmental perception data and utilizes reinforcement learning to fine-tune the cost function online. This enables online autonomous evolution and optimization of the control strategy, forming a closed-loop control architecture with "cognition-decision-prediction-optimization" capabilities, reducing lateral oscillations during steering. Simultaneously, online fine-tuning of the MA-MPC cost function ensures that the optimization objective not only meets current transient performance requirements but also leads to long-term comprehensive optimality, achieving autonomous evolution of the control strategy and dynamic compensation. This significantly enhances the system's robustness in complex pipeline environments.

Claims

1. An intelligent device for inspecting and repairing pipelines with complex multi-diameter structures, characterized in that, It includes a welding torch repair device (1), a scanning device (2), a steering device (3), a cleaning torch (4), a driving device (5), and a signal transceiver device (6). The welding torch repair device (1) is connected to one end of the scanning device (2), and the other end of the scanning device (2) is connected to the steering device (3); the steering device (3) is connected to the cleaning torch (4); the cleaning torch (4) is hollow inside and has a through groove on its periphery, the driving device (5) is set inside the cleaning torch (4) and connected to the steering device (3), the driving device (5) extends through the through groove to the outside of the periphery of the cleaning torch (4) and contacts the pipe, and the device position is adjusted by the driving device (5); the welding torch repair device (1), the scanning device (2), the steering device (3), the cleaning torch (4), the driving device (5) are connected to the signal transceiver device (6) for communication. The welding torch repair device (1) includes a welding torch (11), a rotating gear device (12), and an axial rotating seat (13); the welding torch (11) is fixed on the rotating gear device (12); the end of the rotating gear device (12) is connected to the rotating seat (13) through a universal joint. The steering device (3) includes several hydraulic cylinders, which are equidistantly arranged along the circumferential edges of the scanning device (2) and the driving device (5). The hydraulic cylinders are connected to telescopic rods. The steering device (3) and the driving device (5) are connected by a hinge. The steering device (3) is equipped with an attitude sensor and a pressure sensor. The drive device (5) includes several front-wheel drive devices (51) and rear-wheel drive devices (52); both the front-wheel drive devices (51) and the rear-wheel drive devices (52) are provided with drive wheels (521) and rotating arms, and a telescopic structure (53) is connected in the middle of the rotating arm. The signal transceiver (6) has a built-in algorithm module; the scanning device (2) acquires real-time images and curvature data of the pipeline environment, and acquires motion state data of the device through attitude sensors and pressure sensors; the algorithm module, based on the real-time images, curvature data and motion state data, controls the high-level decision-maker in the system through the algorithm module, and activates the prediction model that best matches the current pipeline working condition from the pre-stored dynamic model library; the algorithm module, based on the activated prediction model, uses the model prediction control algorithm to perform online rolling optimization, and generates control commands for coordinating the extension and pressure of each hydraulic cylinder of the steering device (3), the speed of each drive wheel (521) of the drive device (5), the welding torch repair device (1) and the cleaning torch (4); the signal transceiver (6) sends the control commands to the hydraulic servo system of the steering device (3) and the motor of the drive device (5), so that the device performs adaptive movement and attitude adjustment in the pipeline.

2. The intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 1, characterized in that, The scanning device (2) includes a scanning fixing device (21), a scanning protective cover (22), and a scanning device instrument (23); the scanning protective cover (22) is a transparent cover, and the scanning device instrument (23) is located inside the scanning protective cover (22); the scanning fixing device (21) is located outside the scanning protective cover (22), and one end passes through the scanning protective cover (22) and is fixedly connected to the scanning device instrument (23). The other end of the scanning fixture (21) is connected to the axial rotating seat (13).

3. The intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 1, characterized in that, The drive wheel (521) is a universal wheel device and uses rubber domed tires.

4. The intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 1, characterized in that, The cleaning mechanism (4) includes a slide (41), a cleaning head (42), a water tank (43), and a water pipe ring (44); the cleaning mechanism (4) is provided with a base connected to the steering device (3), and several slides (41) are vertically provided on the base; the water tank (43) and the water pipe ring (44) are set on the slide (41) and move along the slide (41); the water pipe ring (44) is provided with several cleaning heads (42) on the outside; the cleaning head (42), the water tank (43) and the water pipe ring (44) are connected and water is supplied by the water tank (43).

5. A control method for an intelligent device for inspecting and repairing multi-diameter complex pipelines as described in any one of claims 1 to 4, characterized in that, Includes the following steps: S1. Task decision and data acquisition steps: Receive the inspection task instruction and acquire real-time images and defect data of the inner wall of the pipe through the scanning device (2); at the same time, acquire the motion status data of the device itself; S2, Multimodal decision-making step: Based on the real-time images, defect data and motion state data, the control system makes task decisions to determine whether to perform inspection, cleaning or repair work, and activates the corresponding control model and operation parameters. S3. Steps for generating cooperative control instructions: S31. Motion Coordination Command Generation: Based on the motion state data and pipeline geometry data, a model predictive control algorithm is used to generate coordinated motion control commands for the steering device (3) and the drive device (5) so that the device moves and turns along the pipeline. S32, Operation control instruction generation: In response to the task decision, a corresponding operation control instruction is generated; when the decision is a cleaning operation, an instruction is generated to control the extension, rotation and water spraying of the cleaning head (42) of the cleaning mechanism (4); when the decision is a repair operation, an instruction is generated to control the positioning, angle adjustment and start / stop of the welding torch (11) of the welding torch repair device (1). S4. Command execution and closed-loop feedback steps: The coordinated motion control command and the operation control command are respectively sent to the corresponding actuators to drive the device to complete the movement, cleaning or repair actions, and the control command is adjusted in real time based on sensor feedback.

6. The control method of the intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 5, characterized in that, The generation of cooperative motion control commands in step S31 adopts a master-slave dynamic coupling strategy, specifically including: The desired trajectory is generated by the drive unit (5), and the steering unit (3) tracks this attitude through its hydraulic servo system and uses a dynamic coupling compensator based on the yaw angle coupling error. e c Calculate the hydraulic cylinder pressure compensation amount ΔP to eliminate tracking error.

7. The control method of the intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 6, characterized in that, The dynamic coupling compensator calculates the pressure compensation amount ΔP according to the following formula: Coupling error e c = ψ leader ψ follower ; ψ_leader represents the desired yaw angle calculated by the drive unit (5) based on the current curvature of the pipeline or the planned path; ψ_follower represents the actual yaw angle measured in real time by the attitude sensor (such as a gyroscope) on the steering device (3); Where Kp, Kd, ​​and Ki are control gains; γ p， γ d This is the learning rate.

8. The control method of the intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 5, characterized in that, The model predictive control algorithm in step S31 adopts a multimodal adaptive model predictive control scheme, including: A model library containing dynamic models of various typical pipeline operating conditions is constructed; the high-level decision-maker activates the prediction model that best matches the current operating condition from the model library based on real-time environmental perception data; and the cost function weights of the model prediction control algorithm are fine-tuned online using reinforcement learning.

9. The control method of the intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 8, characterized in that, In step S31, fuzzy adaptive decision-making is adopted: The pipe curvature κ, hydraulic cylinder pressure deviation ΔP, and turning yaw angle error e are considered. ψ As input variables; The hydraulic cylinder cooperative gain K is calculated and output using fuzzy inference rules. c Along with the damping compensation coefficient β, it is used to adjust the control parameters in real time.

10. The control method of the intelligent device for inspecting and repairing multi-diameter complex pipelines according to claim 9, characterized in that, Fuzzy yaw error e ψ and curvature κ Using Gaussian membership functions: Defuzzification output cooperative gain: w j For rule weights, M For the number of rules.

Images