An intelligent two-wheeled inspection robot and its monocular vision recognition control system

Abstract

The present invention relates to the field of robot technology, and specifically to an intelligent two-wheeled inspection robot and a monocular vision recognition control system thereof; by making an arc-shaped baffle block the annular airbag, the annular airbag expands downward when it is inflated, so that when the wheel hub rotates, the annular airbag has a better shock-absorbing effect on obstacles, and as the wheel hub rotates, the inflated annular airbag passes through the arc-shaped baffle, and the arc-shaped baffle squeezes the rotating annular airbag. At this time, the annular airbag is squeezed, and the annular airbag squeezes the internal gas downward, which also makes the shock-absorbing effect of the part of the annular airbag that contacts the ground better, and by arranging rollers inside and at both ends of the arc-shaped baffle, when the inflated arc-shaped airbag is squeezed by the arc-shaped baffle, the rollers come into contact with the annular airbag, and the rollers rotate, and the rollers reduce the friction between the annular airbag and the arc-shaped baffle, thereby improving the service life of the annular airbag.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, in particular to an intelligent two-wheel inspection robot and a monocular vision recognition control system thereof. Background Art

[0002] In factory environments with high temperatures, strong corrosion, and other high-risk environments, the use of inspection robots can significantly improve the shortcomings of manual inspections. Traditional manual inspections are not only costly and pose significant safety risks, but are also prone to serious accidents caused by misjudgments and omissions, and the traceability of inspection data is also poor.

[0003] Currently, existing wheeled inspection robots offer features such as autonomous locomotion, obstacle avoidance, and voice interaction. Equipped with a gimbal, they offer practical functions such as high-definition photography, rapid snapshots, night vision photography, precise temperature measurement, and alarms for abnormalities. These robots can meet the needs of daily inspections. However, in practice, the robots are too cumbersome to navigate, making them difficult to inspect in complex and confined areas. They can also easily tilt or even fall over on uneven surfaces.

[0004] In short, it is necessary to study a self-balancing, highly adaptable two-wheeled intelligent two-wheeled inspection robot to solve the above problems.

[0005] In summary, in order to solve the technical problems raised in this article, the present invention proposes an intelligent two-wheeled inspection robot and a monocular vision recognition control system thereof. Summary of the Invention

[0006] In order to make up for the deficiencies of the prior art, the present invention proposes an intelligent two-wheeled inspection robot and a monocular vision recognition control system thereof to solve the above technical problems.

[0007] The technical solution adopted by the present invention to solve the technical problem is: an intelligent two-wheeled inspection robot, comprising: a housing, a wheel hub, a pan / tilt platform, a camera, a control system board and a wireless remote controller; the control system board is provided with a motor drive board, a router, a three-axis gyroscope, a motor drive board and an ultrasonic sensor; and further comprising:

[0008] An annular groove, wherein the annular groove is provided on the wheel hub, and the edge between the annular groove and the wheel hub is of a chamfered design;

[0009] An annular airbag, wherein the inner ring of the annular airbag is fixedly connected to the annular groove, and a one-way valve is provided on the annular airbag;

[0010] A micro fan is embedded in the inner wall of the annular groove, is connected to the annular airbag, and is electrically connected to the control system board; and a ventilation hole is opened on the wheel hub, and is connected to the air inlet of the micro motor;

[0011] An arc-shaped baffle is arranged on the shell, and rollers are rotatably connected to both ends and the inner wall of the arc-shaped baffle.

[0012] A monocular vision recognition control system for an intelligent two-wheeled inspection robot, which is suitable for the above-mentioned intelligent inspection robot; it is characterized in that an electronic control system is provided on the control board system card, and the electronic control system includes a wireless transmission module, a wireless starting module, an image acquisition module and a start-stop module.

[0013] Specifically, the wireless transmission module adopts ASO-MLOIDP as the wireless transmission module, and the wireless transmission module is electrically connected to the control system board.

[0014] Specifically, the image acquisition module is electrically connected to the camera, and the algorithm adopted by the image acquisition module is a mathematical model driven method.

[0015] Specifically, the electronic control system further includes a visual recognition module, and the visual recognition module is connected to the wireless start-stop module.

[0016] Specifically, the vision module adopts a single vision recognition system.

[0017] The beneficial effects of the present invention are as follows.

[0018] 1. The intelligent two-wheeled inspection robot and its monocular vision recognition control system described in the present invention have an arc-shaped baffle on the shell. When the micro-blower is working, the annular airbag expands, and when the annular airbag expands, the diameter of the annular airbag gradually becomes larger than the diameter of the wheel hub. At this time, the arc-shaped baffle blocks the annular airbag, so that when the annular airbag expands, the annular airbag expands downward, so that when the wheel hub rotates, the annular airbag has a better cushioning effect on obstacles, and as the wheel hub rotates, the expansion When the inflated annular airbag passes through the arc-shaped baffle, the arc-shaped baffle squeezes the rotating annular airbag. At this time, the annular airbag is squeezed, and the annular airbag squeezes the internal gas downward, which also makes the shock-absorbing effect of the part of the annular airbag in contact with the ground better. In addition, by arranging rollers inside and at both ends of the arc-shaped baffle, when the inflated arc-shaped airbag is squeezed by the arc-shaped baffle, the rollers come into contact with the annular airbag, and the rollers rotate. The rollers reduce the friction between the annular airbag and the arc-shaped baffle, thereby increasing the service life of the annular airbag.

[0019] 2. The intelligent two-wheeled inspection robot and its monocular vision recognition control system described in the present invention have an annular groove on the wheel hub. When the annular groove is formed, the contact area between the wheel hub and the ground is reduced, and some impurity particles can pass through the annular groove, thereby improving the stability of the inspection robot to a certain extent; and if larger impurity particles are encountered, the staff can control the micro-blower to work through a wireless remote control. The micro-blower draws gas to the outside through the ventilation holes, and then the blower inflates the inside of the annular airbag, and the annular airbag expands. When the annular airbag expands, the diameter of the annular airbag is larger than the diameter of the wheel hub, and the annular airbag is in a soft state, and the annular airbag has not expanded to a hard surface. At this time, when the inspection robot passes through particulate impurities and pits, the impurity particles will be embedded in the annular airbag, and the annular airbag will be squeezed and depressed, so that the inspection robot will not vibrate significantly, thereby ensuring the stability of the inspection robot during movement. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of the monocular vision recognition control system of the inspection robot in the present invention.

[0021] Figure 2 It is a system flow chart of the visual recognition module in the present invention.

[0022] Figure 3 It is a three-dimensional diagram of the inspection robot in the present invention.

[0023] Figure 4 yes Figure 3 Structural view of the control system board.

[0024] Figure 5 yes Figure 4 Partial cross-sectional view of the middle hub.

[0025] In the figure: housing 1, wheel hub 11, pan / tilt head 12, camera 13, control system board 14, annular groove 2, annular airbag 21, micro fan 22, ventilation hole 23, arc baffle 24, roller 25. DETAILED DESCRIPTION

[0026] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0027] Example 1

[0028] An intelligent two-wheeled inspection robot, comprising: a housing 1, a wheel hub 11, a pan / tilt head 12, a camera 13, a control system board 14, and a wireless remote controller; the control system board 14 is provided with a motor drive board, a router, a three-axis gyroscope, a motor drive board, and an ultrasonic sensor; and is characterized in that: it also includes;

[0029] An annular groove 2, the annular groove 2 is provided on the wheel hub 11, and the edge between the annular groove 2 and the wheel hub 11 is a rounded design;

[0030] An annular airbag 21, the inner ring of which is fixedly connected to the annular groove 2, and a one-way valve is provided on the annular airbag 21;

[0031] A micro fan 22 is embedded in the inner wall of the annular groove 2 and is connected to the annular airbag 21. The micro fan 22 is electrically connected to the control system board 14. A ventilation hole 23 is provided on the hub 11, and the ventilation hole 23 is connected to the air inlet of the micro motor.

[0032] An arc-shaped baffle 24 is provided on the housing 1 , and rollers 25 are rotatably connected to both ends and the inner wall of the arc-shaped baffle 24 ;

[0033] First, before the robot is used, the staff needs to conduct a performance test to ensure that the robot's various performances can operate normally when it is used. At this time, the staff needs to turn on the inspection robot through a handheld wireless remote control and turn it on. After testing the normal operation of the robot's various indicators, the robot can be used to inspect the factory area.

[0034] And the patrol robot's balancing system adopts the "dynamic stability" principle (such as the principle of a balancing car) to maintain the stable operation of the robot;

[0035] When the staff conducts an inspection by holding a wireless remote control, the motor inside the robot drives the wheel hub 11 to rotate, and the rotation of the wheel hub 11 drives the robot as a whole to conduct an inspection inside the factory. At this time, the camera 13 installed on the inspection robot takes pictures and videos throughout the process, and the pictures taken by the camera 13 are transmitted to the display screen in front of the staff in real time through the Internet. When there are smaller impurity particles in front of the robot, the control personnel can control the inspection robot to turn and take a detour, and by opening an annular groove 2 on the wheel hub 11, the contact area between the wheel hub 11 and the ground is reduced, and some impurity particles can pass through the annular groove 2, thereby improving the stability of the inspection robot to a certain extent; and if When encountering larger impurity particles, the staff can control the micro-blower 22 to work through the wireless remote control. The micro-blower 22 extracts gas to the outside through the ventilation hole 23, and then the blower inflates the inside of the annular airbag 21, and the annular airbag 21 expands. When the annular airbag 21 expands, the diameter of the annular airbag 21 is larger than the diameter of the hub 11, and the annular airbag 21 is in a soft state at this time. The annular airbag 21 has not expanded to a state where the surface is hard. At this time, when the inspection robot passes through particulate impurities and pits, the impurity particles will be embedded in the annular airbag 21, and the annular airbag 21 will be squeezed and concave, so that the inspection robot will not vibrate greatly, thereby ensuring the stability of the inspection robot during movement.

[0036] Furthermore, a one-way valve is provided on the annular airbag 21, which is an electric one-way valve. When the inspection robot passes over an obstacle, the staff controls the one-way valve to open, and the gas inside the annular airbag 21 is discharged to the outside through the one-way valve. As the inspection robot walks, the gas is ejected from the annular groove 2. The ejected gas can pneumatically clean the inside of the annular groove 2 to a certain extent, and then when the wheel hub 11 rotates, the dust and impurities in the annular groove 2 can be cleaned by the gas blown out of the airbag, thereby improving the cleanliness of the annular groove 2 inside the wheel hub 11.

[0037] And by providing the arc-shaped baffle 24 on the housing 1, when the micro-blower 22 is working, the annular airbag 21 is expanded, and when the annular airbag 21 is expanded, the diameter of the annular airbag 21 gradually becomes larger than the diameter of the hub 11. At this time, the arc-shaped baffle 24 blocks the annular airbag 21, so that when the annular airbag 21 is expanded, the annular airbag 21 expands downward, so that when the hub 11 rotates, the annular airbag 21 has a better shock-absorbing effect on obstacles, and as the hub 11 rotates, the expanded annular airbag 21 passes through the arc-shaped baffle 24. The arc-shaped baffle 24 squeezes the rotating annular airbag 21. At this time, the annular airbag 21 is squeezed, and the annular airbag 21 squeezes the internal gas downward, which also makes the shock-absorbing effect of the part of the annular airbag 21 in contact with the ground better. In addition, by arranging rollers 25 inside and at both ends of the arc-shaped baffle 24, when the inflated arc-shaped airbag is squeezed by the arc-shaped baffle 24, the rollers 25 come into contact with the annular airbag 21. At this time, the rollers 25 rotate, and the rollers 25 reduce the friction between the annular airbag 21 and the arc-shaped baffle 24, thereby improving the service life of the annular airbag 21.

[0038] Example 2

[0039] A monocular vision recognition control system for an intelligent two-wheeled inspection robot, the control system being applicable to the intelligent inspection robot of claim 1; characterized in that an electronic control system is provided on the control board system card, the electronic control system comprising a wireless transmission module, a wireless starting module, an image acquisition module, and a start-stop module;

[0040] By arranging an electric control system on the control system board, the electric control system includes a wireless transmission module, a wireless starting module, an image acquisition module and a start-stop module;

[0041] When the inspection robot is in use, the staff controls the inspection robot to start and stop through the wireless remote control. At this time, the inspection robot is online, and the image acquisition module inside the inspection robot works. At this time, the inspection robot is working, and the image acquisition module takes pictures of the surrounding environment in real time. When the inspection robot is working, the image acquired by the image acquisition module is transmitted to the inspection robot and placed on the control board system card. When the image acquisition system determines that there is an obstacle in front, the image acquisition system determines the type of obstacle in front; it is divided into the following situations;

[0042] 1) When the inspection robot cannot cross the obstacle, the image acquisition module transmits the captured image to the control board system board. At this time, the control system board 14 internal control start-stop module stops the robot and waits for the staff to determine the specific direction, or the inspection robot returns to the original route;

[0043] 2) When the inspection robot determines that the obstacle ahead is insurmountable but that it can bypass it, the image acquisition module transmits the captured image to the control system board. The inspection robot then rotates through the wheel hub 11 to bypass the obstacle.

[0044] 3) When the inspection robot determines that the obstacle ahead is surmountable, the control system board 14 controls the micro fan 22 inside the wheel hub 11 to operate. The micro fan 22 inflates the annular airbag 21, increasing the diameter of the annular airbag 21. When the annular airbag 21 surmounts the obstacle, the obstacle sinks into the annular airbag 21, thereby achieving shock absorption for the inspection robot by the annular airbag 21, thereby stabilizing the inspection robot during inspection.

[0045] Example 3:

[0046] The wireless transmission module adopts ASO1-MLOIDP5 as the wireless transmission module, and the wireless transmission module is electrically connected to the control system board 14;

[0047] The ASO1-MLOIDP5 wireless transmission module is used. When the inspection robot is performing inspections, it uses the ASO1-MLOIDP5 as a wireless transmission module to deal with various situations. This module is an industrial-grade wireless transceiver integrated data transmission module that uses the original NORDIC Nrf24LO1P RF chip and RFX2401C power amplifier chip, has a built-in LNA, and is equipped with a metal shield for strong anti-interference. The module has sufficient transmission power, good spectrum characteristics, low harmonics, low channel crosstalk, and a small size, making it suitable for inspection robot applications. The performance parameters of the module are shown in the table below.

[0048] Performance parameters

[0049]

[0050] Example 4:

[0051] The image acquisition module is electrically connected to the camera 13, and the algorithm adopted by the image acquisition module is a digital model driven method;

[0052] The image acquisition module is connected to the camera 13, and the algorithm of the image acquisition module is driven by a mathematical model. The mathematical model-driven method proposes a series of state hypotheses based on the target model, and achieves target tracking by estimating and correcting these image observation hypotheses. It mainly consists of four parts: target representation, observation representation, hypothesis generation, and hypothesis estimation. This method uses a priori estimation of the known target state and continuously solves the maximum a posteriori estimate of the target state after obtaining new measurements. First, the target state vector {X} is determined, k = 0, 1, 2..., where k represents a discrete time series and the state is the output of the tracking system. The corresponding state equation is as follows:

[0053] Xk+1=EkXk+Vk

[0054] Then determine the target observation vector {100 million}, k = 0, 1, 2... Observation refers to the various features extracted from the image, and the corresponding observation equation is as follows:

[0055] Z k =F k X k +W k

[0056] Formula Xk+1=EkXk+Vk and Z k =F k X k +W k V in k and W k represents noise sequences, which are usually assumed to be independent and identically distributed.

[0057] When the system noise is Gaussian distributed and E k and F k When the relationship is linear, Kalman filtering can be used to solve the posterior estimate; if E and F are nonlinear, extended Kalman filtering can be used to solve the posterior estimate; and then through the numerical model plugging method, the inspection robot can detect, extract, identify and track obstacles when photographing the surrounding environment, obtain the parameters of the obstacle, such as shape, size, etc., and determine whether the inspection robot can cross the obstacle.

[0058] Example 5

[0059] The electronic control system further includes a visual recognition module, and the visual recognition module is connected to the wireless start-stop module;

[0060] The visual recognition module adopts a single visual recognition system;

[0061] By setting a visual recognition module inside the electronic control system and connecting the recognition visual module with the wireless starting module; when the inspection robot is moving, the visual recognition module works, and the visual recognition module judges various conditions of the environment in front of the inspection robot through the visual recognition module;

[0062] Monocular vision systems generally use the corresponding point calibration method to obtain the depth information of the image, and the monocular vision recognition system can be used with the inspection robot to measure the gap between the inspection robot and the obstacle during the inspection, so that the staff can use the data transmitted by the inspection robot to control the start and stop plan of the inspection robot, and through the monocular vision recognition algorithm, the ranging process can be regarded as its corresponding inverse process; the distance measurement process is based on prior knowledge, mapping the data information of the image space W={u,v} to a certain plane S={(x,y,o)eW} process; the vertical coordinate of the bottom midpoint of the target object in the image plane is u, and the position of the camera 13 in the three-dimensional space is represented by l, d, and h; the vertical field of view of the camera 13 is 2a; the clarity of the camera 13 is n×n; finally, the relationship between the coordinates (u, v) of the image space I and the data of the real space plane S={(x, y, o)eW} is established based on the ray angle; therefore, the inspection robot can calculate the actual position of the object in the S plane by obtaining the pixel coordinates of the target object in the captured image through the image acquisition module; and the distance between the target object and the robot is calculated according to the following formula;

[0063]

[0064] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent two-wheeled inspection robot, comprising: a housing (1), a wheel hub (11), a pan / tilt platform (12), a camera (13), a control system board (14), and a wireless remote controller; The control system board (14) is provided with a router, a three-axis gyroscope, a motor drive board and an ultrasonic sensor; and is characterized in that: Also includes: An annular groove (2), wherein the annular groove (2) is provided on the wheel hub (11), and the edge between the annular groove (2) and the wheel hub (11) is designed to be chamfered; An annular airbag (21), the inner ring of which is fixedly connected to the annular groove (2), and a one-way valve is provided on the annular airbag (21); gas inside the annular airbag (21) is discharged to the outside through the one-way valve; A micro fan (22), the micro fan (22) is embedded in the inner wall of the annular groove (2), the micro fan (22) is connected to the annular air bag (21), and the micro fan (22) is electrically connected to the control system board (14); and a ventilation hole (23) is opened on the wheel hub (11), and the ventilation hole (23) is connected to the air inlet of the micro motor; An arc-shaped baffle (24) is provided on the housing (1), and rollers (25) are rotatably connected to both ends and the inner wall of the arc-shaped baffle (24); the arc-shaped baffle (24) blocks the annular airbag (21), so that the annular airbag (21) expands downward when the annular airbag (21) expands.

2. The intelligent two-wheeled inspection robot according to claim 1, characterized in that: The control system board (14) is provided with an electric control system, which includes a wireless transmission module, an image acquisition module and a wireless start-stop module.

3. The intelligent two-wheeled inspection robot according to claim 2, characterized in that: The wireless transmission module adopts AS01-MLOIDP5 as the wireless transmission module, and the wireless transmission module is electrically connected to the control system board (14).

4. The intelligent two-wheeled inspection robot according to claim 2, characterized in that: The image acquisition module is electrically connected to the camera (13), and the algorithm adopted by the image acquisition module is a numerical model driven method.

5. The intelligent two-wheeled inspection robot according to claim 2, characterized in that: The electronic control system further includes a visual recognition module, which is connected to the wireless start-stop module.

6. The intelligent two-wheeled inspection robot according to claim 5, characterized in that: The visual recognition module adopts a monocular visual recognition system.

Images