AI image detection elevator maintenance robot

Abstract

The invention relates to the technical field of robots, in particular to an AI image detection elevator maintenance robot. According to the technical scheme, the system comprises a moving module, an image collecting module, an image enhancing module, an image generating module, an image processing and abnormity recognizing module and an AI module, and the image collecting module is used for collecting images in an elevator and synchronously obtaining position information of the moving module; the image enhancement module is connected with the image acquisition module and is used for enhancing the image acquired by the image acquisition module; the image generation module is connected with the image enhancement module and used for generating data processed and enhanced by the image enhancement module; and the image processing and anomaly recognition module is used for realizing anomaly detection through a pre-trained AI model based on the enhanced image and the generated auxiliary data. According to the invention, the position data of the mobile module is synchronously bound during image acquisition, and specific coordinates and abnormity types can be directly marked after abnormity identification, so that secondary troubleshooting and positioning of maintenance personnel are avoided.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of robots, in particular to an AI image detection elevator maintenance robot. BACKGROUND

[0002] A robot is a machine device that automatically performs work. It can either accept human command or run a pre-programmed program. It mainly assists or replaces human work. At present, manual detection and cleaning are still the main methods in the field of elevator maintenance. Maintenance personnel need to carry a flashlight, an endoscope, a brush and other tools to work in the elevator shaft. The elevator shaft is narrow, and there are complex environments such as motor noise, dim light, humidity and dust. When manually detecting, it is easy to miss hidden dangers such as guide rail rust and door machine screw loosening due to visual blind spots, and there are safety risks such as falling and mechanical extrusion.

[0003] Some existing semi-automatic maintenance devices have obvious limitations. They need manual operation and cannot automatically identify defects. Some robots can identify defects, but cannot obtain the corresponding positions. Therefore, the technical personnel in the field provide an AI image detection elevator maintenance robot to solve the problems in the background art. SUMMARY

[0004] The purpose of the present application is to provide an AI image detection elevator maintenance robot to solve the problems in the background art.

[0005] To achieve the above purpose, the present application provides the following technical scheme: an AI image detection elevator maintenance robot, comprising a moving module, an image acquisition module, an image enhancement module, an image generation module, an image processing and anomaly recognition module, and an AI module,

[0006] The image acquisition module is used for acquiring images in the elevator and synchronously obtaining position information of the moving module;

[0007] The image enhancement module is connected with the image acquisition module and enhances the images collected by the image acquisition module;

[0008] The image generation module is connected with the image enhancement module and generates data processed and enhanced by the image enhancement module;

[0009] The image processing and anomaly recognition module realizes anomaly detection through a pre-trained AI model based on the enhanced images and generated auxiliary data, extracts features from the images, judges whether the features are abnormal, and associates the position information;

[0010] The AI module is used for judging the type of the anomaly;

[0011] The maintenance module is used for maintaining the anomaly.

[0012] An AI image detection elevator maintenance robot also includes a main body structure, a lifting structure, a brushing structure, an adjusting structure and a cleaning structure,

[0013] The main body structure includes a base, a movement module inside the lower end of the base, a voice module at one end of the base, and a control box with an AI module assembled inside the base;

[0014] The image acquisition module includes a camera located on the outer wall of the base;

[0015] The lifting structure includes a mounting arm on one side of the upper end of the base, and a stroke arm on one side of the mounting arm;

[0016] The brushing structure includes bearing seat one symmetrically distributed on the inner wall of the stroke arm, support pipe rotatably mounted inside bearing seat one, cleaning roller in communication with support pipe, cleaning hole in the form of a ring array distributed inside the cleaning roller, and brush in the form of a ring array distributed on the outer wall of the cleaning roller;

[0017] The cleaning structure includes a box on the upper end of the base, a shunt pipe one rotatably mounted inside the support pipe in the form of an arch, a partition plate inside the box, a water pump on the upper end of the partition plate, a delivery pipe one on the output end of the water pump and in communication with the shunt pipe, a hot air blower on one end of the stroke arm, a shunt pipe two penetrating the inner wall of the stroke arm on both sides, and a delivery pipe two on one end of the hot air blower.

[0018] Preferably, the inside of the box is provided with a water chamber one above the partition plate for containing cleaning liquid, and the inside of the box is provided with a water chamber two below the partition plate for containing cleaning water.

[0019] Preferably, a stroke hole is formed in the inside of the upper end of the box, the delivery pipe one penetrates the stroke hole, and a counterweight is sleeved on the outer wall of the delivery pipe one.

[0020] Preferably, the front end of the box is respectively provided with a liquid inlet pipe penetrating the water chamber one and the water chamber two, a sealing plug is arranged at the opening of the liquid inlet pipe, a hydraulic rod one is arranged on one side of the upper end of the base, and a Z-shaped plate connected with the mounting arm is arranged at the output end of the hydraulic rod one.

[0021] Preferably, the output end of the water pump is provided with a suction pipe one penetrating the water chamber one, one end of the suction pipe one is provided with a suction pipe two penetrating the water chamber two, an adjusting valve one is arranged at the connection between the suction pipe one and the suction pipe two, liquid level sensors are arranged on the inner wall of the upper end of the water chamber one and the inner wall of the upper end of the partition plate, and an adjusting valve two is arranged at the communication between the delivery pipe two and the shunt pipe two.

[0022] Preferably, a motor is provided on the upper inner wall, a screw is provided at the output end of the motor, a symmetrically distributed sliding sleeve is provided at one end of the stroke arm, a guide rod connected to the inner wall of the mounting arm is slidably installed inside the sliding sleeve, and a nut connected to the stroke arm is threadedly installed on the outer wall of the screw.

[0023] Preferably, a second motor is provided on the inner wall of one side of the conveying stroke arm, and a rotating shaft is provided at the output end of the second motor, one end of which is rotatably mounted to the inner wall of one side of the stroke arm. A bevel gear is sleeved on the outer wall of the rotating shaft.

[0024] Preferably, the inner wall of the travel arm is provided with symmetrically distributed bearing seats two, and an installation shaft is rotatably installed inside the bearing seats two. Both ends of the installation shaft are fitted with bevel gears three, which mesh with bevel gear one. The outer wall of the support tube is fitted with bevel gear two, which meshes with bevel gear three.

[0025] Preferably, the adjustment structure includes a second hydraulic rod located on the inner wall of the upper end of the base, the telescopic end of the second hydraulic rod is provided with a stroke plate, the lower end of the stroke plate is rotatably mounted with a support shaft, and the lower end of the support shaft is provided with a support plate.

[0026] Preferably, a motor is provided at the lower end of the travel plate, a drive gear is provided at the output end of the motor, and a driven gear that meshes with the drive gear is sleeved on the outer wall of the support shaft.

[0027] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0028] The location data of the mobile module is synchronously bound during image acquisition. After an anomaly is identified, the specific coordinates and anomaly type can be directly marked, avoiding secondary troubleshooting and location by maintenance personnel.

[0029] The high-strength base of this invention provides stable support for each component. The 360° wide-angle night vision camera can comprehensively capture images of components inside the elevator shaft. Combined with the AI ​​module, it can automatically identify component abnormalities without the need for manual visual inspection, reducing the omission of hidden dangers caused by blind spots. The mobile module supports flexible movement and precise positioning, adapting to different ground surfaces and complex paths inside the shaft. The voice module enables real-time two-way interaction between maintenance personnel and the robot, allowing them to obtain information and issue commands without getting close, improving operational convenience.

[0030] Meanwhile, the lifting structure can drive the brushing structure to precisely adjust its height and lateral position, ensuring that the brushing structure fits precisely with the parts to be cleaned. The brushing structure drives the cleaning roller to rotate through a motor and gear transmission, and with the liquid supply through the cleaning holes, it can achieve efficient brushing of elevator parts. The dual water chamber design of the cleaning structure can store cleaning liquid and cleaning water separately, and switch the supply as needed. The hot air blower can also quickly dry the cleaned parts, avoiding liquid residue from affecting the performance of the parts. The adjustment structure lifts the equipment through a hydraulic rod and rotates it with gear transmission, allowing the robot to flexibly adjust its direction in the narrow space of the shaft without manual handling, greatly improving maintenance efficiency and safety, and reducing the intensity of manual labor. Attached Figure Description

[0031] Figure 1 This is a side-view perspective view of the three-dimensional structure of the present invention; Figure 2 This is a bottom-view three-dimensional structural diagram of the present invention; Figure 3 This is a front-view three-dimensional structural schematic diagram of the present invention; Figure 4 This is a schematic diagram of the main cross-sectional three-dimensional structure of the present invention; Figure 5 This is a front-view perspective three-dimensional structural diagram of the box body of the present invention; Figure 6 This is a three-dimensional structural schematic diagram of the main cross-section of the box body of the present invention; Figure 7 This is a three-dimensional structural diagram of the motor side view of the present invention; Figure 8 This is a side-view perspective view of the three-dimensional structure of the cleaning roller of the present invention; Figure 9 This is a top-view perspective view of the three-dimensional structure of the cleaning roller of the present invention; Figure 10 This is a side sectional view of the cleaning roller of the present invention from a first angle, representing a three-dimensional structure. Figure 11 This is a side sectional view of the cleaning roller of the present invention from a second angle, representing a three-dimensional structure. Figure 12 This is a bottom-view three-dimensional structural diagram of the adjustment structure of the present invention; Figure 13 This is a schematic diagram of the present invention.

[0032] Reference numerals: 100, main structure; 101, base; 102, camera; 103, voice module; 104, liquid level sensor; 105, control box;

[0033] 200. Lifting structure; 201. Hydraulic rod one; 202. Z-shaped plate; 203. Mounting arm; 204. Motor one; 205. Screw; 206. Guide rod; 207. Sliding sleeve; 208. Nut; 209. Travel arm;

[0034] 300. Brushing structure; 301. Cleaning roller; 302. Brush; 303. Bevel gear one; 304. Cleaning hole; 305. Bearing housing one; 306. Support tube; 307. Mounting shaft; 308. Bearing housing two; 309. Bevel gear two; 310. Motor two; 311. Rotating shaft; 312. Bevel gear three;

[0035] 400. Adjustment structure; 401. Hydraulic rod two; 402. Stroke plate; 403. Support shaft; 404. Motor three; 405. Drive gear; 406. Driven gear; 407. Support plate;

[0036] 500. Cleaning structure; 501. Box body; 502. Liquid inlet pipe; 503. Sealing plug; 504. Water chamber one; 505. Partition plate; 506. Water chamber two; 507. Water pump; 508. Suction pipe one; 509. Suction pipe two; 510. Regulating valve one; 511. Stroke orifice; 512. Counterweight; 513. Diverter pipe one; 514. Delivery pipe one; 515. Hot air blower; 516. Regulating valve two; 517. Diverter pipe two; 518. Delivery pipe two;

[0037] 600. Image acquisition module; 700. Image enhancement module; 800. Image generation module; 900. Image processing and anomaly recognition module. Detailed Implementation

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

[0039] Please see Figures 1 to 13 The present invention provides four embodiments:

[0040] Example 1: An AI image detection elevator maintenance robot, comprising a mobile module, an image acquisition module 600, an image enhancement module 700, an image generation module 800, an image processing and anomaly recognition module 900, and an AI module.

[0041] The image acquisition module 600 is used to acquire images inside the elevator and simultaneously obtain the position information of the moving module. During image acquisition, the position data of the moving module is bound simultaneously. After an anomaly is identified, the specific coordinates and anomaly type can be directly marked, avoiding secondary troubleshooting and positioning by maintenance personnel.

[0042] The image enhancement module 700 is connected to the image acquisition module 600 and enhances the image acquired by the image acquisition module 600.

[0043] The image generation module 800 is connected to the image enhancement module 700 and generates the enhanced data processed by the image enhancement module 700.

[0044] The image processing and anomaly recognition module 900, based on the enhanced image and generated auxiliary data, uses a pre-trained AI model to achieve anomaly detection, extracts features from the image, determines whether it is abnormal, and associates location information;

[0045] The AI ​​module is used to determine the type of anomaly;

[0046] The maintenance module is used to maintain the system in case of abnormal situations.

[0047] Image enhancement module 700 includes noise reduction, contrast enhancement, reflection suppression, and distortion correction;

[0048] Noise reduction: For image noise in night vision mode, non-local mean filtering or BM3D algorithm is used;

[0049] Contrast enhancement: For images with uneven lighting (such as a bright top and dark bottom in the car), adaptive histogram equalization is used to highlight the edges of components;

[0050] Reflection suppression: To address overexposure in areas caused by reflections from stainless steel walls, a multi-scale Retinex algorithm is used to separate the illumination and reflection components of the image, suppressing highlight areas and restoring information about stains or rust under reflections.

[0051] Distortion correction: Wide-angle lenses have a fisheye effect. Through camera intrinsic parameter calibration, real-time correction is performed after acquisition to ensure that the straight line characteristics of the door operator and car wall are not distorted.

[0052] An AI image detection elevator maintenance robot also includes a main structure 100, a lifting structure 200, a brushing structure 300, an adjusting structure 400, and a cleaning structure 500.

[0053] The main structure 100 includes a base 101, a moving module located inside the lower end of the base 101, a voice module 103 located at one end of the base 101, and a control box 105 equipped with an AI module located inside the base 101.

[0054] The image acquisition module 600 includes cameras 102 located on the four outer walls of the base 101; the cameras 102 are directly connected to the image signal processor inside the control box 105 via a MIPI-CSI interface;

[0055] In this embodiment, the base 101 is integrally die-cast from 6061 high-strength aluminum alloy. Wide-angle high-definition night vision cameras 102 are embedded in the outer walls of the four sides of the base 101. The camera 102 has a field of view covering 120°. The four-sided layout can achieve 360° no blind spot shooting, avoiding detection omissions caused by obstruction of the elevator car structure. The mobile module is integrated inside the lower end of the base 101 and adopts a four-wheel layout with two drive wheels and two universal wheels. The voice module 103 is installed in the groove at one end of the base 101. The control box 105 is embedded in the center of the base 101 and integrates a main control board with an AI module, which can receive signals from the camera 102, sensors and other devices and output control commands.

[0056] The four cameras 102 on the base 101 capture images of the elevator car walls, door operators and other components in real time. The servo motor encoder of the mobile module synchronously collects the robot's position data. If an anomaly is detected, the AI ​​module outputs a signal to the voice module 103 through the control box 105, and broadcasts the anomaly information to the maintenance personnel through the speaker. The maintenance personnel can also use voice commands to transmit the information to the voice module 103 through the microphone. The AI ​​module then analyzes the commands and controls the mobile module to adjust its position. The power module in the control box 105 provides power to the cameras 102, the mobile module, the voice module 103 and other structures.

[0057] Traditional elevator maintenance relies on manual inspection using tools such as flashlights and endoscopes, which is not only time-consuming but also prone to overlooking potential hazards such as rust and loose door operator screws due to structural obstructions. This main structure 100 achieves blind-spot-free, highly accurate automated inspection through a 360° camera 102 and AI recognition. Traditional robots require remote control or data cable connection for operation, and maintenance personnel must follow the robot, resulting in low collaboration efficiency. The voice module 103 of this main structure 100 enables remote two-way interaction, allowing maintenance personnel to obtain information and issue commands in real time outside the shaft, improving human-machine collaboration efficiency.

[0058] Example 2:

[0059] The brushing structure 300 includes a bearing seat 305 symmetrically distributed on the inner wall of the stroke arm 209, a support tube 306 rotatably installed inside the bearing seat 305, a cleaning roller 301 communicating with the support tube 306, cleaning holes 304 arranged in a ring array inside the cleaning roller 301, and brushes 302 arranged in a ring array on the outer wall of the cleaning roller 301.

[0060] A second motor 310 is provided on the inner wall of one side of the stroke arm 209. A rotating shaft 311 is provided at the output end of the second motor 310 and is rotatably mounted on the inner wall of one side of the stroke arm 209. A bevel gear 303 is sleeved on the outer wall of the rotating shaft 311.

[0061] The inner wall of the stroke arm 209 is provided with symmetrically distributed bearing seats 308. The mounting shaft 307 is rotatably installed inside the bearing seat 308. The outer walls of both ends of the mounting shaft 307 are fitted with bevel gears 312. The bevel gears 312 mesh with bevel gears 303. The outer wall of the support tube 306 is fitted with bevel gears 309 that mesh with bevel gears 312.

[0062] In this embodiment, bearing housing 305 is symmetrically installed on both sides of the inner wall of stroke arm 209. It is made of ZG270-500 cast steel and is fitted with deep groove ball bearings. The steel material can withstand the radial load when the cleaning roller 301 rotates. The groove ball bearings reduce the friction coefficient between the support tube 306 and the bearing housing 305, ensuring the long-term stable rotation of the support tube 306. The support tube 306 passes through the bearing housing 305 and is connected to the cleaning roller 301. It is made of 304 stainless steel seamless tube and the inner wall is polished. The polished inner wall reduces the flow resistance of the cleaning liquid and ensures that the liquid enters the cleaning roller 301 smoothly. The cleaning roller 301 and the support tube 306 are integrally formed. The roller body has cleaning holes 304 distributed in a ring array inside. The outer wall is bonded with a brush 302, which can conform to the curved surface of the component for cleaning. Motor 310 is fixed on one side of the inner wall of stroke arm 209 and is a DC servo motor.

[0063] When cleaning elevator components is required, control box 105 outputs a start signal to motor 2 310. Motor 2 310 drives shaft 311 to rotate, and bevel gear 1 303 on shaft 311 rotates synchronously. Bevel gear 1 303 meshes with bevel gear 312 at both ends of mounting shaft 307, transmitting power to mounting shaft 307, causing mounting shaft 307 to rotate around bearing seat 2 308. Bevel gear 312 on mounting shaft 307 then meshes with bevel gear 2 309 on the outer wall of support tube 306, ultimately driving support tube 306 and cleaning roller 301 to rotate synchronously. The rotation direction is the same as the direction of rotation of motor 2 310, and the speed can be adjusted by motor 2 310. Delivery pipe 1 514 in cleaning structure 500 delivers cleaning fluid to the inside of support tube 306. The cleaning fluid is connected to the cleaning roller 301. The cleaning fluid flows along the inner wall of the support tube 306 and finally passes through the annular array of cleaning holes 304 on the cleaning roller 301. It is then evenly sprayed onto the brush 302 on the outer wall of the cleaning roller 301, keeping the brush 302 moist. The rotating cleaning roller 301 drives the brush 302 to contact the surface of the elevator component. Through mechanical friction, the brush 302 removes oil, dust, rust and other stains from the surface of the component. At the same time, the cleaning fluid penetrates through the brush 302 to the contact surface between the stains and the component surface, reducing the adhesion of the stains and improving the cleaning effect. During the cleaning process, the rotation speed of the cleaning roller 301 can be adjusted according to the severity of the stains. Light stains use a low speed of 100-200 rpm, and heavy stains use a high speed of 300-500 rpm.

[0064] Driven by a motor and a bevel gear, the cleaning roller 301 rotates, and in conjunction with the wet nylon brush 302, it can perform all-round brushing on the flat and curved surfaces of elevator guide rails, car walls, and other components. The cleaning efficiency is far higher than that of a handheld brush 302. The connection design between the support tube 306 and the cleaning roller 301, combined with the annular array distribution of the cleaning holes 304, ensures that the cleaning liquid can evenly cover the brush 302, avoiding waste of cleaning liquid, while ensuring that the brush 302 always stays wet, avoiding scratches on the surface of the components caused by dry brushing. Based on the precise speed regulation characteristics of the servo motor, the rotation speed of the cleaning roller 301 can be adjusted according to the type and severity of the stains, which can ensure the cleaning effect of heavy stains and avoid over-brushing in the case of light stains. The brush 302 has a fine filament diameter and high density, which can penetrate into the small gaps on the surface of the components to achieve thorough cleaning of stains in the gaps.

[0065] Traditional elevator component cleaning relies on manual hand-held brushes 302 dipped in cleaning solution for wiping, which is not only time-consuming but also difficult to clean the gaps and curved areas of the components, easily leading to stains and causing component wear. This brush cleaning structure 300 shortens the cleaning time of a single elevator and improves the cleaning rate of gaps through automated rotating brushing and precise liquid supply.

[0066] Example 3:

[0067] The lifting structure 200 includes a mounting arm 203 located on one side of the upper end of the base 101 and a travel arm 209 located on one side of the mounting arm 203;

[0068] The upper inner wall is provided with a motor 204, the output end of the motor 204 is provided with a screw 205, one end of the stroke arm 209 is provided with symmetrically distributed sliding sleeves 207, the inside of the sliding sleeve 207 is slidably installed with a guide rod 206 connected to the inner wall of the mounting arm 203, and the outer wall of the screw 205 is threadedly fitted with a nut 208 one end connected to the stroke arm 209;

[0069] A hydraulic rod 201 is provided on one side of the upper end of the base 101, and a Z-shaped plate 202 connected to the mounting arm 203 is provided at the output end of the hydraulic rod 201.

[0070] In this embodiment, hydraulic rod 201 is a small single-rod double-acting hydraulic rod; hydraulic rod 201 can provide stable thrust to realize the position adjustment of mounting arm 203 from 0-100mm. Z-shaped plate 202 is rigidly connected to mounting arm 203 by bolts to ensure effective force transmission and avoid deformation of mounting arm 203 during adjustment. When the AI ​​module of the main structure 100 identifies the position of the part to be cleaned through camera 102, it sends a control signal to hydraulic rod 201. The piston rod of hydraulic rod 201 extends or retracts, pushing Z-shaped plate 202 to drive mounting arm 203 to move longitudinally. Through the stroke feedback of displacement sensor built into hydraulic rod 201, AI module adjusts the piston rod extension and retraction in real time.

[0071] Control box 105 sends a start signal to motor 204. Motor 204, based on the target height and determined by the AI ​​module using position data from camera 102, rotates clockwise or counterclockwise, driving screw 205 to rotate synchronously. Because the sliding sleeve 207 and guide rod 206 restrict the rotational freedom of nut 208, the rotational motion of screw 205 is converted into linear motion of nut 208. Nut 208 drives stroke arm 209 to vertically rise and fall along guide rod 206. The encoder built into motor 204 feeds back speed and position data to control box 105 in real time. The AI ​​module precisely controls the start and stop of motor 204 by comparing the feedback data with the target height. When the height is 2m, the control motor 204 rotates forward a specific number of times and then stops. After reaching the target height, the motor 204 is powered off and the brake function is activated. The servo motor has an electromagnetic brake to prevent the travel arm 209 from falling due to gravity. At the same time, the tight fit between the sliding sleeve 207 and the guide rod 206 limits the radial sway of the travel arm 209, ensuring that the brushing structure 300 always maintains a stable fit with the part to be cleaned during the cleaning process. After cleaning is completed, the control box 105 controls the motor 204 to reverse, driving the travel arm 209 to descend to the initial position. The piston rod of the hydraulic rod 201 retracts, resetting the mounting arm 203 to the initial position, waiting for the next lifting command.

[0072] Mounting arm 203 serves as a fixed support frame for lifting structure 200, while travel arm 209 is a movable frame connected to brushing structure 300.

[0073] It can achieve vertical lifting within a range of 0-500mm, covering the height range of a single section of guide rail in the elevator shaft. The hydraulic rod 201 provides a stroke of 0-100mm, which can adapt to the guide rail spacing of different elevators. It can achieve precise contact between the brushing structure 300 and the guide rail without replacing the lifting components, improving the equipment's versatility. It can receive AI module instructions from the control box 105 of the main structure 100, eliminating the need for manual operation. It adopts a transmission method with servo motor, screw 205, nut 208, and encoder feedback, which is far superior to the cylinder lifting or belt drive of traditional elevator maintenance equipment. This ensures that the brushing structure 300 is accurately aligned with the stains, reducing cleaning dead corners.

[0074] Motor 204 uses a DC servo motor. Screw 205 and nut 208 are the core components for achieving linear transmission. Existing lifting structures 200 are mostly of fixed size and cannot adapt to the guide rail spacing of different elevators. For example, the guide rail spacing of old elevators is 80mm, while that of new elevators is 120mm. It is necessary to customize lifting frames with different spacings for each elevator, which increases equipment costs and inventory pressure. This structure can achieve single-structure multi-spacing adaptation through fine adjustment of hydraulic rod 201.

[0075] Example 4:

[0076] The cleaning structure 500 includes a housing 501 located at the upper end of the base 101, a first branch pipe 513 rotatably installed inside the support pipe 306 and arranged in an arch shape, a partition 505 located inside the housing 501, a water pump 507 located at the upper end of the partition 505, a first delivery pipe 514 located at the output end of the water pump 507 and connected to the branch pipe, a hot air blower 515 located at one end of the stroke arm 209, a second branch pipe 517 penetrating the inner walls of both sides of the first branch pipe, and a second delivery pipe 518 located at one end of the hot air blower 515.

[0077] The cabinet 501 has a water chamber 504 located above the partition 505 for holding cleaning liquid and a water chamber 506 located below the partition 505 for holding cleaning water.

[0078] A stroke hole 511 is opened inside the upper end of the box 501, and a first conveying pipe 514 passes through the stroke hole 511. A counterweight 512 is sleeved on the outer wall of the first conveying pipe 514.

[0079] The output end of the water pump 507 is provided with a suction pipe 508 that passes through the first water chamber 504. One end of the suction pipe 508 is provided with a suction pipe 509 that passes through the second water chamber 506. A regulating valve 510 is provided at the connection between the first suction pipe 508 and the second suction pipe 509. Liquid level sensors 104 are provided on the upper inner wall of the first water chamber 504 and the upper inner wall of the partition 505. A regulating valve 516 is provided at the connection between the second delivery pipe 518 and the second diversion pipe 517.

[0080] The front end of the housing 501 is provided with liquid inlet pipes 502 that penetrate water chamber one 504 and water chamber two 506 respectively, and a sealing plug 503 is provided at the opening of the liquid inlet pipe 502.

[0081] The adjustment structure 400 includes a hydraulic rod 401 located on the inner wall of the upper end of the base 101. The extension end of the hydraulic rod 401 is provided with a stroke plate 402. A support shaft 403 is rotatably mounted on the lower end of the stroke plate 402. A support plate 407 is provided on the lower end of the support shaft 403.

[0082] A motor 404 is installed at the lower end of the travel plate 402. A drive gear 405 is installed at the output end of the motor 404. A driven gear 406 that meshes with the drive gear 405 is sleeved on the outer wall of the support shaft 403.

[0083] In this embodiment, the housing 501 serves as a storage container for the cleaning medium. A PPH partition 505 is welded inside the housing 501, dividing the housing 501 into an upper water chamber 504 and a lower water chamber 506 to prevent the cleaning liquid from mixing and contaminating the cleaning water. The water pump 507 is a miniature corrosion-resistant centrifugal pump. A brass regulating valve 510 is installed at the intersection of the suction pipe 508 and the suction pipe 509. The brass regulating valve 510 has good sealing performance. The delivery pipe 514 is a PU hose with a cast iron counterweight 512 sleeved on the outer wall. The diversion pipe 513 is a 304 stainless steel pipe. The PU hose has good flexibility and can be bent without breaking as the brush structure 300 rises and falls. The counterweight 512 can prevent the delivery pipe 514 from getting tangled when the stroke arm 209 rises and falls. The arched diversion pipe 513 and the liquid outlet design can make the cleaning medium evenly sprayed onto the inner wall of the cleaning roller 301, and together with the cleaning hole 304, the brush 302 is fully wetted.

[0084] The hot air blower 515 uses a small centrifugal hot air generator. The delivery pipe 2 518 is a silicone tube that runs through the inner walls of both sides of the diversion pipe 1 513. A capacitive liquid level sensor 104 is installed on the upper inner wall of water chamber 1 504 and the upper inner wall of partition 505. The sensors are connected to the control box 105 via wires to monitor the liquid level in water chamber 1 504 and water chamber 2 506 in real time. When the liquid level is lower than a set threshold, an alarm signal is sent to the control box 105 to prevent the water pump 507 from running dry and being damaged. The main structure 100AI module identifies the type of stain through the camera 102. If it is stubborn oil or rust, the control regulating valve 1 510 opens the suction pipe 1 508, and the water pump 507 starts to draw water from water chamber 1 506. 04 Cleaning fluid; If it is ordinary dust, the suction pipe 2 509 draws the cleaning water and cleaning medium from the water chamber 2 506, which enters the arched diversion pipe 1 513 through the delivery pipe 1 514. It is then evenly sprayed onto the inner wall of the support pipe 306 through the liquid outlet hole, and then penetrates into the brush 302 through the cleaning hole 304 of the cleaning roller 301, keeping the brush 302 moist. The cleaning medium comes into contact with the surface of the component through the brush 302, reducing the adhesion of the stains. Combined with the mechanical friction of the rotating cleaning roller 301, the stains are removed more quickly. During the process, the liquid level sensor 104 provides real-time feedback on the remaining water in the water chamber. If it is lower than the threshold, the control box 105 broadcasts to the maintenance personnel through the voice module 103 that the cleaning fluid and cleaning water are insufficient, reminding them to replenish.

[0085] After scrubbing, the AI ​​module controls the water pump 507 to stop and the regulating valve 1 510 to close. Simultaneously, the hot air blower 515 starts and the regulating valve 2 516 opens. Hot air enters the distribution pipe 2 517 through the delivery pipe 2 518 and is blown onto the cleaned component surface through the air outlet. The temperature of the hot air blower 515 is adjusted according to the component material. The hot air acts directly on the wet surface, accelerating moisture evaporation and preventing guide rail corrosion or electrical component short circuits caused by liquid residue. It can automatically switch between cleaning liquid and cleaning water according to the type of stain, eliminating the need for manual container changes. It is suitable for use in elevator shafts. The cleaning needs of different types of stains can be precisely adjusted by regulating valve 510 and water pump 507 to control the flow rate. The supply of cleaning medium can be precisely adjusted, and the uniform liquid output of diversion pipe 513 ensures that the brush 302 is moderately wetted. Liquid level sensor 104 monitors the remaining liquid in the dual water chambers in real time and automatically alarms when the liquid level is low to prevent water pump 507 from running dry. Dry running can easily lead to overheating and damage to the pump body. At the same time, it reminds maintenance personnel to replenish the liquid in time to avoid interruption of cleaning operations. Adjustable hot air can quickly dry the surface of the parts to avoid corrosion or electrical failure caused by liquid residue after cleaning.

[0086] This structure achieves targeted cleaning and improves stain removal rate through dual water chambers and intelligent switching. Manual cleaning is prone to waste of cleaning solution due to excessive application or container tipping, and leaked liquid may seep into elevator electrical components and cause malfunctions. This structure reduces waste rate through quantitative supply and sealing design. The hot air drying of this structure can quickly dry the solution and solve the rust problem.

[0087] When the AI ​​module of the main structure 100 detects an obstacle in front, such as a compensation chain or speed limiter, and needs to turn to avoid it or adjust the washing direction, it sends a start signal to the hydraulic rod 401. The piston rod of the hydraulic rod 401 extends, pushing the stroke plate 402, support shaft 403 and support plate 407 to descend synchronously until the rubber pad of the support plate 407 contacts the ground. The piston rod continues to extend, lifting the entire equipment, so that the drive wheel and caster of the moving module are completely off the ground, avoiding friction between the wheels and the ground during rotation and hindering steering. After the equipment is lifted, the control box 105 sends a rotation command to the motor 404. The motor 404 drives the drive gear 405 to rotate. The drive gear 405 meshes with the driven gear 406, driving the support shaft 403 to rotate around the bearing. The support shaft 403 synchronously drives the main body of the equipment to rotate. The encoder built into the motor 404 feeds back the rotation angle data to the control box 105 in real time. The AI ​​module compares the feedback angle with the target angle. When the error exceeds the threshold, it controls the motor 404 to stop, completing the precise adjustment.

[0088] After the reversal is completed, the control box 105 controls the piston rod of hydraulic rod 401 to retract, driving the support plate 407 to rise until the drive wheel and caster of the moving module re-contact the ground. After the support plate 407 is completely off the ground, hydraulic rod 401 stops retracting to prevent the support plate 407 from rubbing against the ground when the equipment moves. At this time, the equipment can travel in the new direction through the moving module to enter the next maintenance station. The entire equipment can be lifted to lift the moving module off the ground, providing frictionless conditions for rotation and reversal. This is suitable for the narrow spaces of elevator shafts or cars, and is driven by a servo motor and gears. It can rotate at any angle from 0 to 360°, precisely adjusting the direction of the equipment to ensure that the brushing structure 300 is aligned with the part to be cleaned. It receives instructions from the main structure 100AI module and automatically completes the entire process of lifting, rotating, and lowering without manual operation, improving the efficiency of operation in confined spaces. In elevator shafts, components such as compensation chains and speed governors often obstruct the view. This structure completely solves the space limitation problem by lifting and rotating. Traditional manual pushing and steering has a large error in the steering angle. After adjusting the direction, the brushing structure 300 cannot be accurately aligned with the guide rail and requires repeated adjustments. One adjustment is enough to align it, eliminating the need for repeated adjustments.

[0089] The above specific embodiments are merely several preferred embodiments of the present invention. Based on the technical solutions of the present invention and the relevant teachings of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

[0090] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. An AI image detection elevator maintenance robot, characterized in that: It includes a mobile module, an image acquisition module (600), an image enhancement module (700), an image generation module (800), an image processing and anomaly recognition module (900), and an AI module. The image acquisition module (600) is used to acquire images inside the elevator and simultaneously obtain the position information of the moving module; An image enhancement module (700) is connected to an image acquisition module (600) to enhance the image acquired by the image acquisition module (600); An image generation module (800) is connected to an image enhancement module (700) to generate the enhanced data processed by the image enhancement module (700); The image processing and anomaly detection module (900) uses a pre-trained AI model to detect anomalies based on the enhanced image and generated auxiliary data. It extracts features from the image, determines whether it is abnormal, and associates location information. The AI ​​module is used to determine the type of anomaly; The maintenance module is used to maintain the system in case of abnormal situations.

2. The AI ​​image detection elevator maintenance robot according to claim 1, characterized in that: It includes a main structure (100), a lifting structure (200), a brushing structure (300), an adjusting structure (400), and a cleaning structure (500). The main structure (100) includes a base (101), a moving module located inside the lower end of the base (101), a voice module (103) located at one end of the base (101), and a control box (105) equipped with an AI module located inside the base (101). The image acquisition module (600) includes cameras (102) located on the four outer walls of the base (101); The lifting structure (200) includes a mounting arm (203) located on one side of the upper end of the base (101) and a travel arm (209) located on one side of the mounting arm (203). The brushing structure (300) includes a bearing seat (305) symmetrically distributed on the inner wall of the stroke arm (209), a support tube (306) rotatably installed inside the bearing seat (305), a cleaning roller (301) communicating with the support tube (306), cleaning holes (304) arranged in a ring array inside the cleaning roller (301), and brushes (302) arranged in a ring array on the outer wall of the cleaning roller (301). The cleaning structure (500) includes a housing (501) located at the upper end of the base (101), a first branch pipe (513) rotatably installed inside the support pipe (306) and arranged in an arch shape, a partition (505) located inside the housing (501), a water pump (507) located at the upper end of the partition (505), a first delivery pipe (514) located at the output end of the water pump (507) and connected to the branch pipe, a hot air blower (515) located at one end of the stroke arm (209), a second branch pipe (517) penetrating the inner walls of both sides of the first air flow pipe, and a second delivery pipe (518) located at one end of the hot air blower (515). The box (501) is provided with a water chamber 1 (504) located above the partition (505) for holding cleaning liquid, and a water chamber 2 (506) located below the partition (505) for holding cleaning water.

3. The AI ​​image detection elevator maintenance robot according to claim 2, characterized in that: The upper end of the box (501) has a stroke hole (511), the first conveying pipe (514) passes through the stroke hole (511), and a counterweight (512) is sleeved on the outer wall of the first conveying pipe (514).

4. The AI ​​image detection elevator maintenance robot according to claim 2, characterized in that: The front end of the housing (501) is provided with an inlet pipe (502) that passes through water chamber one (504) and water chamber two (506). A sealing plug (503) is provided at the opening of the inlet pipe (502). A hydraulic rod one (201) is provided on one side of the upper end of the base (101). A Z-shaped plate (202) connected to the mounting arm (203) is provided at the output end of the hydraulic rod one (201).

5. The AI ​​image detection elevator maintenance robot according to claim 2, characterized in that: The output end of the water pump (507) is provided with a suction pipe (508) that passes through the first water chamber (504). One end of the suction pipe (508) is provided with a suction pipe (509) that passes through the second water chamber (506). A regulating valve (510) is provided at the connection between the first suction pipe (508) and the second suction pipe (509). A liquid level sensor (104) is provided on the upper inner wall of the first water chamber (504) and the upper inner wall of the partition (505). A regulating valve (516) is provided at the connection between the second delivery pipe (518) and the second diversion pipe (517).

6. The AI ​​image detection elevator maintenance robot according to claim 2, characterized in that: The upper inner wall is provided with a motor (204), the output end of the motor (204) is provided with a screw (205), one end of the stroke arm (209) is provided with symmetrically distributed sliding sleeves (207), the sliding sleeve (207) is slidably installed with a guide rod (206) connected to the inner wall of the mounting arm (203), and the outer wall of the screw (205) is threaded with a nut (208) connected to the stroke arm (209) at one end.

7. The AI ​​image detection elevator maintenance robot according to claim 2, characterized in that: A second motor (310) is provided on the inner wall of one side of the conveying stroke arm (209). The output end of the second motor (310) is provided with a rotating shaft (311) that is rotatably installed on the inner wall of one side of the stroke arm (209). A bevel gear (303) is sleeved on the outer wall of the rotating shaft (311).

8. The AI ​​image detection elevator maintenance robot according to claim 7, characterized in that: The inner wall of the stroke arm (209) is provided with symmetrically distributed bearing seats two (308). The bearing seat two (308) is rotatably mounted with a mounting shaft (307). Both ends of the mounting shaft (307) are sleeved with bevel gear three (312). The bevel gear three (312) meshes with bevel gear one (303). The outer wall of the support tube (306) is sleeved with bevel gear two (309) that meshes with bevel gear three (312).

9. The AI ​​image detection elevator maintenance robot according to claim 2, characterized in that: The adjustment structure (400) includes a hydraulic rod two (401) located on the inner wall of the upper end of the base (101). The extension end of the hydraulic rod two (401) is provided with a stroke plate (402). The lower end of the stroke plate (402) is rotatably mounted with a support shaft (403). The lower end of the support shaft (403) is provided with a support plate (407).

10. The AI ​​image detection elevator maintenance robot according to claim 9, characterized in that: The lower end of the stroke plate (402) is provided with a motor three (404), the output end of the motor three (404) is provided with a drive gear (405), and the outer wall of the support shaft (403) is sleeved with a driven gear (406) that meshes with the drive gear (405).

Images