Wheel-foot infrared inspection robot

Abstract

The utility model discloses a wheel-foot infrared inspection robot which comprises a moving platform, a bearing platform, a servo electric cylinder, a double-mechanical-arm assembly, a space sensing system and a tail end executing mechanism. The double-mechanical-arm assembly, the servo electric cylinder and the bearing platform are sequentially arranged on the top of the moving platform from top to bottom, and the space sensing system is arranged on the bearing platform and the main body; a flash lamp and a thermal infrared imager are independently controlled through a double-mechanical-arm assembly, dynamic adjustment of the excitation angle of the flash lamp and the visual angle of infrared thermal imaging is achieved, mechanical arm shielding is avoided, the thermal imaging signal-to-noise ratio and detection precision of defects are remarkably improved, and flash lamp excitation-thermal imaging collaborative optimization is achieved; the moving platform of a four-wheel-foot type structure can adapt to complex terrains such as a grating ground and a dense equipment area, the vibration suppression technology is matched, it is ensured that imaging at the tail end of the mechanical arm is stable, and tiny damage in a composite material is effectively recognized.

Description

Technical Field

[0001] This utility model relates to the field of defect and damage inspection technology, and in particular to a wheeled infrared inspection robot. Background Technology

[0002] In the field of defect detection for complex industrial equipment, high-precision identification of surface cracks and internal damage is a core requirement for ensuring the safe operation of equipment.

[0003] Firstly, traditional flash-excited infrared detection solutions employ fixed detection devices or single-arm robotic inspection systems, which, due to structural limitations, struggle to achieve dynamic coordination between the excitation source and the detection equipment. Fixed installation leads to a limited detection angle, failing to adapt to complex equipment surface morphology. Furthermore, single-arm structures, when simultaneously manipulating the excitation light source and imaging device, are prone to creating detection blind spots due to mechanical interference, and cannot optimize the spatiotemporal matching of thermal excitation and thermal imaging through coordinated motion, directly impacting the signal-to-noise ratio during heat conduction.

[0004] Secondly, conventional robotic arm systems rely heavily on multi-joint compound motion to expand their vertical working range. Due to the lack of an active base height adjustment mechanism, the robotic arm needs to frequently perform large-scale joint movements to compensate for height differences. This not only generates redundant energy consumption but also causes the complexity of the end effector's motion trajectory to increase exponentially, severely restricting the system's response speed and motion stability in continuous detection scenarios.

[0005] Third, traditional wheeled or tracked mobile platforms have significant limitations in complex industrial scenarios. Rigid motion mechanisms are difficult to adapt to discontinuous support surfaces such as grids, and are prone to motion interference in densely populated areas. Although legged robots have the advantage of terrain adaptability, there is an inherent contradiction between mobility efficiency and energy consumption ratio, and their movement can easily cause high-frequency micro-amplitude vibrations in the imaging unit at the end of the robotic arm, which seriously affects the accuracy of identifying defects such as micro-cracks between material layers.

[0006] In summary, current technologies still suffer from insufficient flexibility in equipment damage detection, lack of vertical adjustment capability, and poor dynamic adaptability. There is an urgent need for an integrated inspection solution that combines robotic arm collaborative control, multi-dimensional sensor fusion, and terrain-adaptive movement.

[0007] To address this, a wheeled infrared inspection robot is proposed. Summary of the Invention

[0008] The purpose of this invention is to provide a wheeled infrared inspection robot to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the main technical solutions adopted by this utility model include:

[0010] A wheeled infrared inspection robot, comprising:

[0011] Mobile platform, load-bearing platform, servo electric cylinder, dual robotic arm assembly, spatial perception system, and end effector;

[0012] The dual robotic arm assembly, servo electric cylinder, and carrying platform are arranged sequentially from top to bottom on the top of the mobile platform. The spatial perception system is located on the carrying platform and the main body. The end effector is located on the dual robotic arm assembly.

[0013] The dual robotic arm assembly includes a main body fixedly connected to the top of the servo electric cylinder, with a left robotic arm and a right robotic arm fixed to each side of the main body.

[0014] The spatial perception system includes a lidar fixed on the upper surface of the support platform along the front of the direction of travel, and a binocular vision camera is provided on the top of the main body;

[0015] The end effector includes a flashlight fixedly connected to the end of the left robotic arm and a thermal infrared imager fixedly connected to the end of the right robotic arm.

[0016] As a preferred technical solution, the mobile platform adopts a four-wheeled composite mobile mechanism. The two rear foot ends of the mobile platform along the direction of travel are respectively equipped with rear drive wheels A and B, and the two front foot ends of the mobile platform along the direction of travel are respectively equipped with front drive wheels A and B.

[0017] As a preferred technical solution, the supporting platform is a rounded rectangular table structure. The bottom surface of the supporting platform is rigidly fixed to the top of the mobile platform through distributed connectors. One side of the supporting platform extends along the direction of travel and forms a lidar installation area. The lidar installation area extends beyond the front contour of the mobile platform and forms a suspended section. The lidar is located on the suspended section.

[0018] As a preferred technical solution, the servo electric cylinder is a compact piston rod electric cylinder, which is mounted on the upper surface of the support platform at the rear along the travel direction.

[0019] As a preferred technical solution, both the left and right robotic arms are robotic arms with six degrees of freedom or more, and the left and right robotic arms are respectively fixed to both sides of the main body by four bolts.

[0020] As a preferred technical solution, the lidar is a multi-line lidar, the scanning axis of the lidar coincides with the longitudinal center line of the support platform, and the binocular vision camera adopts an active binocular imaging system, which includes an adjustable focus lens group and an adaptive filter.

[0021] As a preferred technical solution, the flash lamp is composed of an array of high-brightness LED modules, and the thermal infrared imager uses an uncooled focal plane detector.

[0022] This utility model has at least the following beneficial effects:

[0023] I. This application achieves dynamic adjustment of the flash excitation angle and the infrared thermal imaging field of view by independently controlling the flash lamp and the thermal imager through a dual robotic arm assembly, avoiding obstruction by the robotic arm, significantly improving the thermal imaging signal-to-noise ratio and detection accuracy of defects, and realizing the collaborative optimization of flash excitation and thermal imaging.

[0024] Second, this application, by adopting a four-wheeled foot structure mobile platform, can adapt to complex terrains such as grid ground and dense equipment areas. Combined with vibration suppression technology, it ensures stable imaging at the end of the robotic arm and effectively identifies minute damage inside composite materials.

[0025] Third, this application uses a servo electric cylinder to drive the lifting and lowering of a dual robotic arm assembly, expanding the vertical detection range. It also combines environmental perception data to adjust the robotic arm's working reference plane in real time, optimizes the joint motion trajectory of the robotic arm, significantly reduces redundant motion energy consumption, and improves detection efficiency.

[0026] Fourth, this application integrates multimodal sensing with dual robotic arm collaborative path planning to support adaptive detection of irregular surfaces such as metals and composite materials, and simultaneously collects multi-dimensional data in visible light and thermal infrared, thereby improving the reliability of defect identification in interference environments. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the overall structure of this utility model;

[0028] Figure 2 This is a left view of the overall structure of this utility model;

[0029] Figure 3 This is a schematic diagram of the mobile platform cabin structure in this utility model;

[0030] Figure 4 This is a top view of the mobile platform cabin structure in this utility model;

[0031] Figure 5 This is a top view of the load-bearing platform layout in this utility model;

[0032] Figure 6 This is a diagram of the main body structure of this utility model;

[0033] Figure 7 This is a schematic diagram of the structure of the left robotic arm in this utility model;

[0034] Figure 8This is a schematic diagram of the right robotic arm in this utility model;

[0035] Figure 9 This is a schematic diagram of a simplified pitch adjustment structure for the L-shaped bracket in this utility model;

[0036] Figure 10 This is a left view of the simplified pitch adjustment structure of the L-shaped bracket in this utility model.

[0037] In the diagram: 1. Mobile platform; 2. Load-bearing platform; 3. Servo electric cylinder; 4. Main body; 41-42. Flange A; 5. Left robotic arm; 51-56. Left joint; 57. Flange B; 6. Right robotic arm; 61-66. Right joint; 67. Flange C; 7. LiDAR; 8. Binocular vision camera; 81. L-shaped bracket; 811. Pitch adjustment shaft hole; 812. Clamping device; 9. Flash light; 10. Thermal infrared imager; 11. Rear drive wheel A; 12. Rear drive wheel B; 13. Front drive wheel A; 14. Front drive wheel B; 15. Base plate; 151. Hollowed-out partition plate; 16. Main control module; 17. Power module; 171. Power supply anti-vibration bracket. Detailed Implementation

[0038] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0039] Please see Figures 1-10 This utility model provides a wheeled infrared inspection robot, including a mobile platform 1, a support platform 2, a servo electric cylinder 3, a dual-arm assembly, a spatial perception system, and an end effector. The dual-arm assembly, the servo electric cylinder 3, and the support platform 2 are arranged sequentially from top to bottom on the top of the mobile platform 1. The spatial perception system is located on the support platform 2 and the main body 4, and the end effector is located on the dual-arm assembly. The dual-arm assembly is connected by bolts to the main body 4, with a left arm 5 and a right arm 6 fixed to the sides of the main body 4, respectively. The left arm 5 and the right arm 6 are symmetrically arranged about the central axis of the main body 4. The spatial perception system includes a laser radar 7 directly bolted to the upper surface of the support platform 2 along the front of the direction of travel through three mounting holes at the bottom. A binocular vision camera 8 is located on the top of the main body 4. The end effector includes a flash 9 fixedly connected to the end of the left arm 5 and a thermal infrared imager 10 fixedly connected to the end of the right arm 6.

[0040] The mobile platform 1 houses a main control module 16 and a power supply module 17. It employs a four-wheeled composite mobility mechanism. Rear drive wheels A11 and B12 are mounted on the two rear legs of the platform along the direction of travel, while front drive wheels A13 and B14 are mounted on the two front legs. All drive wheels 11-14 are all-terrain rubber tires with deep groove treads, combined with flexible tire materials, to achieve flexible contact with the ground, effectively absorbing vibrations and enhancing obstacle-crossing capabilities. Motion control is based on the fusion data of LiDAR 7 and binocular vision cameras 8. A dynamic weight allocation algorithm generates the four-wheel drive torque. Combining terrain slope, obstacle distribution, and the stability requirements of the robotic arm's end effector, the rotational speed and torque output of each drive wheel 11-14 are adjusted in real time to ensure the stability of the mobile platform 1 in rugged terrain, while simultaneously suppressing imaging jitter caused by platform vibrations at the ends of the left and right robotic arms 5 and 6.

[0041] The mobile platform 1 has a base plate 15, a perforated partition plate 151, a main control module 16, a power module 17, and a power anti-vibration bracket 171 in its belly compartment. The perforated partition plate 151 is fixedly connected to the middle of the base plate 15, dividing the mobile platform 1 into a front control area and a rear energy area. The main control module 16 is located in the front area and is fixed to the base plate 15 through distributed heat sinks. The power module 17 is located in the rear area and is fixed to the base plate 15 through the perforated power anti-vibration bracket 171, which takes into account both heat dissipation and stability, and maintains stable power output in a mechanical vibration environment. Among them, the base plate 15 serves as the basic load-bearing unit, and its surface is precision-machined to form the positioning reference for the main control module 16 and the power module 17. The upright hollow partition plate 151 adopts a topology optimization design and forms a vertical cross support system with the base plate through laser welding process. Its umbrella-shaped hollow structure not only physically isolates the control area and the energy area, but also forms an airflow channel that runs through the cabin, so that the structural load-bearing capacity and heat dissipation efficiency produce a synergistic gain effect, achieving dual-dimensional optimization of mechanical stress transmission and thermal field distribution.

[0042] The main control module 16 integrates an inertial navigation unit, a data acquisition board, a charge amplifier board, a power amplifier, and a flash excitation signal generator. The inertial navigation unit is mounted inside the main control module via vibration-damping pads, acquiring platform attitude angle, acceleration, and angular velocity data in real time. The data acquisition board enables high-precision synchronous acquisition and anti-interference processing of multi-channel sensor signals, ensuring the timing consistency of data from thermal imaging, robotic arm encoders, etc. The flash excitation signal generator and the thermal infrared imager achieve timing matching through a high-precision synchronization protocol.

[0043] The supporting platform 2 is a rounded rectangular table structure. The bottom surface of the supporting platform 2 is rigidly fixed to the top of the mobile platform 1 through distributed connectors. One side of the supporting platform 2 extends along the direction of travel and forms a lidar installation area. The lidar installation area extends beyond the front contour of the mobile platform 1 and forms a suspended section. The lidar 7 is fixedly connected to the top of the suspended section. The design of the suspended section effectively expands the horizontal scanning range of the lidar 7 and avoids the occlusion of point cloud data by the main structure of the mobile platform 1.

[0044] Among them, the servo electric cylinder 3 is a compact piston rod electric cylinder. The servo electric cylinder 3 is installed on the upper surface of the support platform 2 at the rear along the travel direction. The output end of the servo electric cylinder 3 lifts and lowers to drive the movement of the dual robotic arm assembly, thereby adjusting the height of the base roots of the left robotic arm 5 and the right robotic arm 6, and realizing the height adjustment of the working reference plane of the dual robotic arms. The height adjustment strategy of the servo electric cylinder 3 is based on the fusion data of the lidar 7 and the binocular vision camera 8. The lidar 7 provides global height information of the target device, and the binocular vision camera 8 calculates the local height deviation of the target surface through a stereo matching algorithm. The control layer generates the target displacement of the servo electric cylinder 3 accordingly, realizing the adaptive matching of the working reference plane of the dual robotic arms, so that the initial working height of the end effector maintains a preset safe distance from the target surface.

[0045] The left robotic arm 5 and the right robotic arm 6 are both serial robotic arms with six degrees of freedom or greater, and are fixed to both sides of the main body 4 by four bolts. The left robotic arm 5 includes left joints 51-56, and the right robotic arm 6 includes right joints 61-66. Each joint uses an integrated drive unit of harmonic reducer and servo motor. The end of the left robotic arm 5 is installed with a flash lamp 9 via flange B57, and the end of the right robotic arm 6 is installed with a thermal infrared imager 10 via flange C67. The kinematic model of the dual robotic arms is constructed based on DH parameters. The control layer generates joint angle trajectories through inverse kinematics algorithm and combines obstacle point cloud data from lidar 7 to correct the robotic arm movement path in real time to avoid collisions with target equipment or the environment.

[0046] The main body 4 is a box-shaped welded structure with internal cable channels and heat dissipation ducts. Flanges A41 and A42 are symmetrically fixed on both sides. The base of the left robotic arm 5 and the right robotic arm 6 are rigidly connected to the corresponding flange A by four bolt groups. The bolt groups are applied with torque in an alternating pre-tightening manner to avoid deformation of the flange surface.

[0047] The LiDAR 7 is a multi-line LiDAR, with its scanning axis coinciding with the longitudinal centerline of the support platform 2. This ensures that the horizontal rotation scanning plane is perpendicular to the robot's travel direction, thus eliminating point cloud distortion caused by installation misalignment. LiDAR 7 employs a multi-line scanning mode, generating 3D point cloud data covering a 360° horizontal field of view and a preset vertical field of view by emitting laser beams and receiving reflected signals. This data is used to construct a real-time environmental geometric model. The binocular vision camera 8 utilizes an active binocular imaging system, including an adjustable focus lens group and adaptive filters, covering the visible and near-infrared spectra. During scanning, the timing pulse signal of LiDAR 7 and the imaging frame rate of the binocular vision camera 8 are aligned through a hardware synchronization module, achieving timestamp matching between point cloud data and visual images. This provides spatiotemporally consistent raw input for multi-scale fusion mapping. LiDAR 7 provides large-scale, high-precision geometric point cloud data for establishing a global coordinate system for the environment and obstacle contours. The binocular vision camera 8 extracts visible light texture details from local areas using a stereo matching algorithm. Based on hardware synchronization signals, the scanned point cloud of LiDAR 7 and the RGB-D data of binocular vision camera 8 are aligned in coordinate system. A geometric-texture correlation mapping is established through feature point matching such as edges and corners, and key areas are modeled using active imaging technology of binocular vision. During robot movement, LiDAR 7 continuously updates the global map, while binocular vision camera 8 performs localized fine-tuning modeling of the robotic arm's working area. Ultimately, a multi-scale fusion map with both navigation accuracy and detection detail is generated, providing robust environmental model support for robotic arm collaborative control and defect localization.

[0048] The binocular vision camera 8 is mounted on the top of the main body 4 via an L-shaped bracket 81. Its clamping device 812 adopts a quick-release buckle structure, allowing for quick loading and unloading of the binocular vision camera 8. The vertical arm of the L-shaped bracket 81 is provided with a pitch adjustment shaft hole 811, which, in conjunction with the pitch fastening bolt, enables stepless adjustment of the pitch angle of the binocular vision camera 8, ensuring that its field of view covers the working area of ​​the dual robotic arm assembly and the end effector. The active imaging system of the binocular vision camera 8 dynamically adjusts the focal length through an adjustable focus lens group and switches the visible light band in combination with an adaptive filter. Its working spectrum range avoids signal interference with the infrared band of the thermal infrared imager 10.

[0049] The flash lamp 9 consists of 16 high-brightness LED modules arranged in an array, supporting fine adjustment over a wide range of pitch and horizontal angles. These 16 high-brightness LED modules are arranged in a 4×4 rectangular array, achieving full coverage illumination of the target area through a precisely designed distributed optical path structure. The LED modules are arranged in an equally spaced matrix, with each module controlled by an independent pulse drive circuit, supporting regional triggering or synchronized flashing. Its spectral range covers the visible to near-infrared band, used to excite the transient thermal response of the inspected surface and internal defects. The rectangular array layout, through orthogonal symmetry, adapts to the multi-angle movement characteristics of the robotic arm, ensuring uniform light projection onto the target surface under different excitation modes such as oblique and vertical incidence, avoiding detection errors caused by local blind spots or uneven light intensity distribution.

[0050] The thermal infrared imager 10 employs an uncooled focal plane array detector. It is rigidly connected to the end of the right robotic arm 6 via flange C67. The uncooled focal plane array detector works in conjunction with the joint degrees of freedom of the right robotic arm 6 to achieve dynamic adjustment of the pitch and horizontal angles. Based on the 3D model of the target surface fed back by the binocular vision camera 8, and combined with a preset incident angle tolerance range, the right robotic arm 6 adjusts the observation attitude of the thermal infrared imager 10 in real time. This ensures that the detector's optical axis maintains the optimal angle with the normal to the inspected surface, suppressing environmental reflection interference and maximizing the signal-to-noise ratio of the thermal radiation signal.

[0051] The working principle of this utility model is as follows: During use, a four-wheeled mobile platform 1 equipped with a LiDAR 7 and a binocular vision camera 8 is used. Based on the 3D point cloud data from the LiDAR 7 and the stereoscopic vision information from the binocular vision camera 8, an environmental map is constructed. A global path is generated for the mobile platform 1 according to the target coordinates. The torque distribution of the four-wheeled drive wheels 11-14 is synchronously adjusted by a distributed motion controller to achieve adaptive movement across complex terrain, moving the platform to the work area. The movement of the servo electric cylinder 3 is controlled to move the dual robotic arm assembly, achieving height adjustment of the working reference plane of the dual robotic arm assembly. Hand-eye calibration is performed through the binocular vision camera 8 to establish a mapping relationship between the coordinate system of the robotic arm assembly and the coordinate system of the target equipment, generating the initial pose matrix of the robotic arm end effector. The left robotic arm 5 and the right robotic arm 6 respectively control the flashlight 9 and the heat lamp according to a preset collaborative strategy. The infrared imager 10 and the flash lamp 9 perform multi-angle flash excitation according to a preset pulse sequence, synchronously triggering the exposure acquisition of the thermal infrared imager 10. The right robotic arm 6 dynamically adjusts the shooting distance and incident angle according to the real-time temperature field data of the thermal infrared imager 10 to optimize the thermal imaging signal-to-noise ratio and obtain high-contrast damage images. The dual robotic arm components, guided by the LiDAR environment modeling and visual positioning, perform collision avoidance trajectory optimization and multi-angle fine re-inspection based on the collaborative path planning algorithm, and simultaneously acquire multimodal data of visible light texture and thermal infrared features. Based on the thermo-mechanical coupling model, the correlation between surface defects and internal damage is analyzed, the defect location is automatically marked, and a visual diagnostic report containing defect geometric parameters, temperature gradient, and damage level is generated. After the inspection is completed, the robot stores the report, realizing a closed loop of the entire process from environmental perception and dynamic inspection to intelligent decision-making.

[0052] In summary, this utility model solves the technical problems of low efficiency and poor dynamic adaptability of equipment damage detection in complex industrial scenarios by using the coordinated control of dual robotic arm components, the servo electric cylinder 3 to adjust the working height of the dual robotic arm components, and combining laser radar navigation, binocular vision and flash lamp-excited infrared damage detection technology. It also has the advantages of compact structure and high multi-dimensional detection accuracy.

[0053] All parts not described in this utility model are the same as or can be implemented using existing technology. Although embodiments of this utility model have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this utility model, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A wheel-foot infrared inspection robot, characterized in that, The application relates to a mobile platform (1), a bearing platform (2), a servo electric cylinder (3), a double-mechanical-arm assembly, a space perception system and an end execution mechanism. The double-mechanical-arm assembly, the servo electric cylinder (3) and the bearing platform (2) are sequentially arranged on the top of the mobile platform (1) from top to bottom, the space perception system is arranged on the bearing platform (2) and a main trunk (4), and the end execution mechanism is arranged on the double-mechanical-arm assembly. The double-mechanical-arm assembly is connected and comprises the main trunk (4) fixedly connected to the top of the servo electric cylinder (3), and left and right mechanical arms (5) and (6) are respectively fixed to the two sides of the main trunk (4). The space perception system comprises a laser radar (7) fixed to the front part of the upper surface of the bearing platform (2) along the running direction, and a binocular vision camera (8) arranged on the top of the main trunk (4). The end execution mechanism comprises a flash lamp (9) fixedly connected to the end of the left mechanical arm (5) and a thermal infrared imager (10) fixedly connected to the end of the right mechanical arm (6). The mobile platform (1) adopts a four-wheel foot composite moving mechanism, two rear driving wheels A (11) and B (12) are respectively arranged at the two foot ends of the rear part of the mobile platform (1) along the running direction, and two front driving wheels A (13) and B (14) are respectively arranged at the two foot ends of the front part of the mobile platform (1) along the running direction.

2. The wheel-foot infrared inspection robot according to claim 1, characterized in that: The bearing platform (2) is a round-corner rectangular table structure, the bottom surface of the bearing platform (2) is rigidly fixed to the top of the mobile platform (1) through a distributed connecting piece, one side of the bearing platform (2) extends along the running direction and forms a laser radar mounting area, the laser radar mounting area exceeds the front profile of the mobile platform (1) and forms a suspended section, and the laser radar (7) is located on the suspended section.

3. The wheel-foot infrared inspection robot according to claim 2, characterized in that: The servo electric cylinder (3) is a compact piston rod electric cylinder, and the servo electric cylinder (3) is arranged on the upper surface of the bearing platform (2) along the rear part of the running direction.

4. The wheel-foot infrared inspection robot according to claim 3, characterized in that: The left and right mechanical arms (5) and (6) are mechanical arms with six or more degrees of freedom, and the left and right mechanical arms (5) and (6) are respectively fixed to the two sides of the main trunk (4) through four bolts.

5. The wheel-foot infrared inspection robot according to claim 4, characterized in that: The laser radar (7) is a multi-line laser radar, the scanning axis of the laser radar (7) coincides with the longitudinal center line of the bearing platform (2), the binocular vision camera (8) adopts an active binocular imaging system and comprises an adjustable focusing lens group and an adaptive filter.

6. The wheel-foot infrared inspection robot according to claim 5, characterized in that: The flash lamp (9) is composed of 16 groups of high-brightness LED modules arranged in an array, and the thermal infrared imager (10) adopts a non-refrigeration focal plane detector.

7. The wheel-legged infrared inspection robot according to claim 6, characterized in that: ​

Images