CN113757540B

CN113757540B - Oiling system for tracking robot by machine vision guide

Info

Publication number: CN113757540B
Application number: CN202110938084.3A
Authority: CN
Inventors: 孟瑞; 卢宗远; 王璐; 谢能刚; 朱群锋; 赵如新
Original assignee: Anhui University of Technology AHUT
Current assignee: Anhui University of Technology AHUT
Priority date: 2021-08-16
Filing date: 2021-08-16
Publication date: 2023-03-24
Anticipated expiration: 2041-08-16
Also published as: CN113757540A

Abstract

The invention discloses an oiling system for guiding robot tracking by machine vision, which belongs to the technical field of lubricating devices and comprises an oiling module, a vision identification module and a speed measurement module; the oiling module comprises a multi-axis robot, an oiling unit and a control unit, the oiling unit is connected with the multi-axis robot, the multi-axis robot moves the oiling unit, the vision recognition module comprises a camera, and the camera continuously recognizes whether the wheels of the sintering machine trolley need oiling. The manual oiling device can completely replace the traditional manual oiling mode, and well solves the problems of low efficiency, high strength, incapability of ensuring oiling amount and the like existing in the conventional manual oiling mode; the intermittent speed measurement mode is adopted, so that the influence of plane errors among all vehicle sections is avoided, the abrasion of a speed measurement module is reduced, and the service life of the whole oil injection system is prolonged; the oiling process of each wheel is mutually independent, and meanwhile, the oiling work cannot be influenced by the position error of the wheels.

Description

Oiling system for tracking robot by machine vision guide

Technical Field

The invention relates to the technical field of lubricating devices, in particular to an oil injection system for guiding robot tracking through machine vision.

Background

The sintering machine trolley is necessary equipment in sintering production operation, and in the sintering operation, a bearing of a wheel of the sintering machine trolley is subjected to severe working conditions such as high temperature, heavy load, impact vibration, dust and water showering, so that the wheel of the sintering machine trolley needs to be frequently supplemented with lubricating oil to ensure the normal operation of the wheel.

The number of wheels of the sintering trolley used at present is large, and the manual oiling operation is heavy. The current lubrication mode is divided into three parts: the sealing screw is taken down, the oil nozzle is used for injecting oil and the sealing screw is screwed down, the operation steps are multiple, the online manual oil injection operation difficulty of the sintering machine trolley is high, potential safety hazards exist, and the oil injection amount cannot be guaranteed to be uniform and sufficient. Therefore, an oiling system for tracking of a robot guided by machine vision is provided.

Disclosure of Invention

The technical problem to be solved by the invention is as follows: how to solve the problem that the efficiency is low, intensity is big, the oiling volume can't guarantee that present artifical oiling exists, provide an oiling system that machine vision guide robot trailed, can replace traditional artifical oiling mode completely.

The invention solves the technical problems through the following technical scheme, and the invention comprises an oiling module, a visual identification module and a speed measurement module; the oiling module comprises a multi-axis robot, an oiling unit and a control unit (MS all-in-one machine), the oiling unit is connected with the multi-axis robot, the oiling unit of the multi-axis robot moves, the vision recognition module comprises a camera, and the camera continuously recognizes whether wheels of the sintering machine trolley need oiling or not; the speed measuring module comprises an encoder and an encoder wheel which are connected, the encoder is driven by a sintering machine trolley to rotate to acquire speed information, and the multi-axis robot, the oiling unit, the camera and the encoder are all in communication connection with the control unit.

Furthermore, the oiling system for guiding the robot to track by the machine vision further comprises a basic frame unit, the basic frame unit is an installation frame, and the oiling module, the vision recognition module and the speed measurement module are all arranged on the installation frame.

Still further, the vision recognition module further comprises a camera frame through which the camera is connected with the mounting frame.

Furthermore, the oil injection unit comprises an oil injection head and an oil pump, wherein the oil injection head is arranged at the tail end of the multi-axis robot, and the oil injection head is communicated with the oil pump.

Furthermore, the oil injection head and the camera are sequentially arranged along the advancing direction of the sintering pallet, the oil injection head is arranged in front, and the camera is arranged behind.

Furthermore, the oil injection unit further comprises a fixing block, and the oil injection head is connected with the tail end of the multi-axis robot through the fixing block.

Furthermore, the module of testing the speed still includes cylinder, mounting panel, encoder mount pad, the cylinder passes through the mounting panel with the mounting bracket is connected, encoder wheel all are connected with the encoder mount pad, the encoder mount pad with be provided with tension spring between the pivot between encoder, the encoder wheel, operating condition is down the encoder wheel passes through tension spring hugs closely together with sintering machine platform truck side plane.

Furthermore, the control unit comprises a servo driver and a controller, the servo driver is in communication connection with the multi-axis robot, and the oiling unit, the camera and the encoder are in communication connection with the controller.

Further, the process of target positioning of the wheel oil hole by the vision recognition module comprises the following steps:

s1: preprocessing a wheel image of a sintering machine trolley;

s2: identifying and positioning the oil injection hole;

s3: and carrying out image calibration and coordinate conversion.

Further, in step S3, the specific process of image calibration and coordinate transformation includes: the world coordinate system and the camera coordinate system can be coincided through two conversion modes of rotation and translation, the selected SCARA robot (multi-axis robot) and the camera are arranged on the same mounting frame, the mounting heights of the SCARA robot and the camera are different, and the mounting planes are parallel to each other; because the camera and the robot are fixed on the mounting frame, the relative positions of the camera and the robot are unchanged and are constant; the distance between the wheels and the lens of the sintering pallet is constant and is also a fixed value; the conversion between world coordinates and pixel coordinates of the robot can be determined by parameters of a space vector, namely three parameters of the space vector are obtained by calibrating two groups of coordinates at the beginning and the end of three points;

the calibration steps are as follows:

step 1: installing a calibration rod on a flange at the tail end of the robot, and ensuring that the center of a conical workpiece (calibration rod) is the center of the robot flange;

and 2, step: uniformly sticking solid dot calibration paper on a calibration plate, wherein the solid dot calibration paper is shown in figure 1;

and step 3: adjusting parameters of a camera to enable the captured image to be clear;

and 4, step 4: clicking a learning template module of a program on the MS all-in-one machine, and recording pixel coordinates of centers of 3 non-collinear solid dots, namely coordinates of 11, 12 and 13 points in the figure 1;

and 5: and operating the SCARA robot demonstrator, aligning the tip of the tail end of the calibration rod to the centers of the

dots

11, 12 and 13 in the step 4 in sequence, and recording the world coordinates of each dot position respectively.

Compared with the prior art, the invention has the following advantages: the oil injection system for guiding the robot to track by machine vision can completely replace the traditional manual oil injection mode, and well solves the problems of low efficiency, high strength, incapability of ensuring the oil injection amount and the like existing in the conventional manual oil injection; the intermittent speed measurement mode is adopted, so that the influence of plane errors among all vehicle sections is avoided, the abrasion of a speed measurement module is reduced, and the service life of the whole oil injection system is prolonged; the oiling process of each wheel is mutually independent, and the oiling work can not be influenced by the position error of the wheels, so that the device is worthy of being popularized and used.

Drawings

FIG. 1 is a schematic view of a solid dot calibration paper;

FIG. 2 is a schematic view of the overall structure of an oil injection system in the second embodiment of the present invention;

FIG. 3 is a schematic side view of an oil injection system according to a second embodiment of the present invention;

fig. 4 is a schematic structural diagram of a speed measurement module in the second embodiment of the present invention.

In FIGS. 2 to 4: 1. an oil injection head; 2. an oil injection head fixing block; 3. a SCARA robot; 4. a mounting frame; 5. an oil pump; 6. a shielding plate; 7. a camera frame; 8. a camera; 9. a cylinder mounting plate; 10. a cylinder fixing plate; 11. a cylinder; 12. a cylinder piston; 13. an encoder mounting base; 14. a tension spring; 15. an encoder; 16. an encoder wheel; 17. a wheel; 18. sintering machine trolley.

Detailed Description

The following examples are given for the detailed implementation and specific operation of the present invention, but the scope of the present invention is not limited to the following examples.

Example one

The embodiment provides a technical scheme: an oiling system for guiding robot tracking by machine vision comprises a vision identification module, a speed measurement module and a SCARA robot. The oil injection work is completed by the SCARA robot carrying an oil injection head and guided by a visual identification module and a speed measurement module. The SCARA robot also comprises a matched MS all-in-one machine (a Taida drive control all-in-one machine), the MS all-in-one machine combines a control module and a drive module together through a bus, the system has strong operation capability and good real-time performance, can quickly and accurately process position instructions, can perform simple servo control by embedding a soft PLC in the system, and can perform serial port communication or Ethernet communication with surrounding vision to perform data exchange.

In this embodiment, the MS all-in-one machine is an industrial robot controller integrating a controller and a 4-axis servo driver, and integrates functions of oil injection, identification, speed measurement and the like in the same control core.

In this embodiment, the SCARA robot is provided with an oil injection unit, and the oil injection unit mainly includes an oil injection head, an oil pump and other components, and is used for completing the work of injecting oil to the oil injection hole under the driving of the SCARA robot.

In this embodiment, the vision recognition module includes a camera and a camera frame, and the camera can continuously recognize the wheels of the sintering pallet, and after recognizing the wheel to be oiled, the camera sends the position information of the wheel to the MS all-in-one machine of the SCARA robot. Therefore, compared with the traditional oiling mechanism, the oiling process of each wheel is independent, and meanwhile, the oiling work cannot be influenced by the position error of the wheels.

The visual target positioning method for the wheel oil filling hole in the embodiment mainly comprises the following steps of: 1. preprocessing a wheel image of a trolley of the sintering machine; 2. identifying and positioning the oil injection hole; 3. and (5) image calibration and coordinate conversion.

The preprocessing of images in machine vision is an indispensable link in a vision system, and even if the best hardware facilities are used in the acquisition process of the images, noise is formed, and the noise affects the processing, identification and other aspects of the images. The preprocessing of the wheel image of the sintering pallet comprises the following steps:

1.1 image Filtering

In the process of image acquisition and transmission by using a camera, various noises are contained in an image due to many factors such as heating, luminescence and the like, the quality of the image is reduced, and the image cannot be identified under the condition of serious interference. Therefore, it is necessary to perform filtering processing on the image with noise. The median filtering method can eliminate noise and ensure target edge information, and the method needs to detect a round hole (oil hole) and has higher requirement on the edge information, so the median filtering is selected.

1.2 binarization of image

The purpose of binarization is to extract features from the background in a grayscale image. In the invention, in order to reduce the data amount of the subsequent processed images and improve the timeliness of identifying the oil hole of the wheel of the sintering machine, the target image needs to be subjected to binarization processing. The image binaryzation is classified according to the histogram characteristics, wherein a fixed threshold value method is used for effectively dividing a background and a target by a manually established threshold value, the method is high in operability and can be manually adjusted according to the field environment, and therefore the method can be adapted to the current situation. Because the color difference between the oil injection hole of the sintering wheel and the periphery is obvious and the difference between the target and the background is small, the binaryzation treatment of a fixed threshold value is selected.

The identification and positioning of the wheel oil holes of the sintering machine trolley comprises the following contents:

2.1 edge detection of images

The Canny operator has better effect than other methods in the edge processing, so the Canny operator is selected for edge detection.

The steps for edge detection using the Canny algorithm are as follows:

1) Calculating the gradient amplitude and the direction to respectively obtain the strength of the edge and the direction angle of the edge;

2) Non-maximum suppression, namely, preserving the maximum point of the gradient amplitude obtained in the gradient direction in the local gradient range, or replacing the gradient amplitude with 0, thereby achieving the purpose of removing non-edge points;

3) And detecting high and low thresholds and connecting image edges, and selecting reasonable high and low thresholds to perform edge detection and connection on the processed images.

2.2 Hough transform circle detection

For the robot oiling system, the sintering machine wheel oiling holes need to be identified through vision, and coordinates of the oiling holes are converted into robot tail end coordinates and sent to an MS all-in-one machine of the SCARA robot. The external shape of the oil injection hole is approximately circular, so that the Hough transformation method is adopted to detect the circular hole, and the principle of the method is to convert pixel points of an image subjected to edge processing into the peak value of the midpoint of a spatial conical surface.

In the original point image, any coordinate pixel point of an edge pixel is mapped into a Hough space, multiple points intersected with a circular arc in a plane are obtained, and the space circular arc of each point is intersected at one point through a mapping relation. The method is characterized in that parameters are accumulated by setting an accumulator, the accumulator is initialized to be 0, 1 is added to the accumulator corresponding to the parameter equation which is satisfied for each point of the image space, and the point corresponding to the peak value is the required information.

The image calibration and coordinate conversion comprises the following contents:

the camera and the SCARA robot are fixed on the robot mounting frame, so that the relative positions of the camera and the SCARA robot are constant values. The position relation of the sintering machine trolley, the visual identification module and the SCARA robot is calibrated to obtain a transformation matrix of a coordinate system of the sintering machine trolley and a world coordinate system. The specific operation is as follows: firstly, wheels are stopped in a range which can be reached by the robot, the tail end of the robot is operated to reach the wheel oil injection hole, a first coordinate position is recorded, then the sintering machine trolley is started to run for a certain distance, the sintering machine trolley is stopped to run, the tail end of the robot is ensured to reach, the tail end is also operated to reach the wheel oil injection hole, a second coordinate position is recorded, and a transformation matrix of a coordinate system of the sintering machine trolley and a world coordinate system can be obtained. The acquired pictures are preprocessed and identified by an algorithm in static vision through transformation matrix and parameter adjustment, and finally obtained position information is transmitted to a control system (MS all-in-one machine) to realize robot tracking and oiling.

The speed measuring module comprises an installation plate, a cylinder, an encoder wheel and the like, wherein the encoder drives the encoder wheel to rotate through the sintering machine trolley, and the obtained speed information is sent to the MS all-in-one machine of the SCARA robot. Cylinder, encoder and encoder wheel are fixed by the mounting panel and are being close to sintering machine platform truck outer wall, when the camera discerned treating the oil filler point (discerning the oil filler point and showing current wheel promptly and need the oiling, and SCARA robot begins the oiling with the oiling head stretches into in the oil filler point), the solenoid valve effect of control cylinder stretches out encoder and encoder wheel, makes sintering machine platform truck drive the encoder wheel. The cylinder withdraws after the oiling work is finished, and the next oiling is waited. Therefore, compared with the traditional speed measurement, the intermittent speed measurement in the invention is not influenced by plane errors among all vehicle sections, meanwhile, the abrasion of the speed measurement module is reduced, and the service life of the whole oil injection system is prolonged.

Example two

As shown in fig. 2 to 4, the oiling system in this embodiment includes a SCARA robot, a visual recognition module, and a speed measurement module;

the SCARA robot 3 is fixed on a cross beam of the mounting frame 4; the tail end of the SCARA robot 3 is provided with an oil injection head fixing block 2; the front end of the oil injection head fixing block 2 is provided with an oil injection head 1; the oil injection head 1 is connected with an oil pump 5 through an oil injection head fixing block 2 by a grease pipe.

The vision recognition module comprises a camera 8 and the like; the camera 8 is fixed on the left side of the mounting frame 4 through a camera frame 7; the upper, lower, left and rear sides of the camera frame 7 are provided with shielding plates 6.

The speed measuring module comprises an air cylinder mounting plate 9 and the like; the cylinder mounting plate 9 is arranged at the leftmost end of the cross beam of the mounting frame 4, and the cylinder 11 is arranged at the foremost end of the cylinder mounting plate; the encoder mounting base 13 is mounted at the tail end of a piston rod of the air cylinder 11; the encoder wheel 16 is fixed to the encoder mounting base 13, and the encoder wheel 16 is in close contact with the sintering machine pallet 18 in an operating state.

The working principle is as follows: the system is arranged at one side of the running track of wheels 17 of a sintering machine trolley 18, so that the cross beam of the mounting frame 4 is parallel to the running track of the sintering machine trolley 18; when the sintering pallet 18 is running, the wheels 17 move on the corresponding rails; meanwhile, the MS all-in-one machine controls the oil injection equipment (namely an oil injection unit) on the SCARA robot to start, controls the electromagnetic valve of the cylinder 11 to open, and extends the speed measurement module out, so that the encoder wheel 16 is tightly attached to the side plane of the sintering machine trolley 18 by virtue of the pulling force of the tension spring 14 and obtains the moving speed of the sintering machine trolley 18; when the wheel 17 passes through the visual recognition module, a camera of the camera 8 recognizes and acquires the pixel coordinate of the oil injection hole by photographing, and the pixel coordinate is converted into an actual coordinate (world coordinate) by an MS all-in-one machine of the SCARA robot 3; after a certain time delay, the wheel 17 reaches the front of the oil injection head 1 (namely the initial position of the SCARA robot 3), the oil injection head 1 rapidly moves to the obtained coordinate point through the SCARA robot 3 and extends the oil injection head 1 into an oil injection port on the wheel 17, and linear tracking is kept at the speed obtained by the speed measurement module; when the oil injection head 1 extends into the oil injection port for a short time delay, the relay coil of the oil pump 5 is controlled by the MS integrated machine of the SCARA robot 3 to enable the oil pump 5 to start pumping oil, the oil pumping is stopped after a certain time, the oil injection head 1 is retracted after the short time delay, and then the SCARA robot 3 returns to the initial position. And the next wheel 17 to be lubricated moves to the right position, so that the reciprocating motion is realized, and finally, the automatic oiling and lubrication of all the wheels 17 are realized.

In the present embodiment, the process of converting the pixel coordinates of the oil hole to the world coordinates mainly includes the following two steps:

firstly, converting pixel coordinates into a camera coordinate system;

and secondly, transforming from the camera coordinate system to a world coordinate system.

The photographing of the camera is a process of pinhole imaging, wherein 4 coordinate systems are involved:

world coordinate system: the coordinate origin of the space can be determined according to the situation, and can represent the object in the space;

camera coordinate system: the optical center of the camera is taken as the origin (in the pinhole model, namely, the pinhole is taken as the center), the z axis is coincided with the optical axis, namely, the z axis points to the front of the camera (vertical to an imaging plane), the positive directions of the x axis and the y axis are parallel to a world coordinate system, and the unit is a length unit;

image physical coordinate system (also called imaging plane coordinate system): the position of a pixel is expressed by a physical length unit, and the origin of coordinates is the intersection point position of the optical axis of the camera and a physical coordinate system of the image;

pixel coordinate system: the origin of coordinates is at the upper left corner, and the pixel is taken as a unit, so that obvious range limitation exists, namely the pixel length and the pixel length width for representing a full picture;

the position of a certain pixel point in the pixel coordinate system corresponding to the world coordinate system is required to be obtained, the internal reference and the external reference of the camera need to be known, the internal reference of the camera can be obtained through calibration, and the external reference can be set manually.

In summary, the oil injection system for guiding the robot to track by machine vision of the embodiment can completely replace the traditional manual oil injection mode, and well solves the problems of low efficiency, high strength, incapability of ensuring the oil injection amount and the like existing in the conventional manual oil injection; the intermittent speed measurement mode is adopted, so that the influence of plane errors among all vehicle sections is avoided, the abrasion of a speed measurement module is reduced, and the service life of the whole oil injection system is prolonged; the oiling process of each wheel is mutually independent, and the oiling work can not be influenced by the position error of the wheels, so that the device is worthy of being popularized and used.

Although embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are exemplary and not to be construed as limiting the present invention, and that changes, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.

Claims

1. The utility model provides an oiling system that machine vision guide robot tracked which characterized in that: the device comprises an oiling module, a visual recognition module and a speed measurement module; the oiling module comprises a multi-axis robot, an oiling unit and a control unit, the oiling unit is connected with the multi-axis robot, the multi-axis robot carries the oiling unit to move, the vision recognition module comprises a camera, and the camera continuously recognizes whether the wheels of the sintering machine trolley need oiling or not; the speed measuring module comprises an encoder and an encoder wheel which are connected, the encoder drives the encoder wheel to rotate through a sintering machine trolley to acquire speed information, and the multi-axis robot, the oiling unit, the camera and the encoder are all in communication connection with the control unit;

the oil injection unit comprises an oil injection head and an oil pump, the oil injection head is arranged at the tail end of the multi-axis robot, and the oil injection head is communicated with the oil pump;

the oiling head and the camera are sequentially arranged along the advancing direction of the sintering pallet, the oiling head is arranged in front of the camera, and the camera is arranged behind the camera;

the oil injection unit further comprises a fixed block, and the oil injection head is connected with the tail end of the multi-axis robot through the fixed block;

the oiling system for guiding the tracking of the robot by the machine vision also comprises a basic frame unit, wherein the basic frame unit is an installation frame, and the oiling module, the vision identification module and the speed measurement module are all arranged on the installation frame;

the process of positioning the target of the wheel oil filling hole by the visual identification module comprises the following steps: firstly, stopping wheels in the working range of a robot, operating the tail end of the robot to a wheel oil hole, recording a first coordinate position, starting a sintering machine trolley to run for a certain distance, stopping running the sintering machine trolley, ensuring that the tail end of the robot reaches the wheel oil hole, recording a second coordinate position, generating a transformation matrix of a sintering machine trolley coordinate system and a world coordinate system, and transmitting position information obtained by preprocessing and algorithm recognition of collected pictures in static vision to a control system for realizing robot tracking oil injection through the transformation matrix and parameter adjustment;

the visual recognition module further comprises a camera frame, and the camera is connected with the mounting frame through the camera frame;

the process of the visual identification module for positioning the target of the wheel oil filling hole comprises the following steps:

s1: preprocessing a wheel image of a sintering machine trolley, wherein the preprocessing is used for performing image filtering and image binarization processing on the wheel image of the sintering machine trolley;

s2: identifying and positioning the oil injection hole, wherein the identifying and positioning comprises the steps of detecting a circular hole through a Hough transformation method after edge detection is carried out on the preprocessed trolley wheel image of the sintering machine through a Canny algorithm, and converting pixel points of the edge-processed image into peak values of the middle points of the spatial conical surface;

s3: carrying out image calibration and coordinate conversion, wherein three parameters of a space vector are obtained by calibrating two groups of coordinates at the beginning and the end of three points, and the conversion between world coordinates and pixel coordinates of the robot is obtained according to the parameters of the space vector;

the speed measuring module further comprises an air cylinder, a mounting plate and an encoder mounting seat, the air cylinder is connected with the mounting frame through the mounting plate, the encoder and an encoder wheel are both connected with the encoder mounting seat, a tension spring is arranged between the encoder mounting seat and a rotating shaft between the encoder and the encoder wheel, and the encoder wheel is tightly attached to the lateral plane of the sintering machine trolley through the tension spring in a working state;

when the camera discerned treating the oil filler point, the solenoid valve effect of control cylinder stretched out encoder and encoder wheel, made sintering machine platform truck drive the encoder wheel, and the encoder drives the encoder wheel through the sintering machine platform truck and rotatory and then in the speed information who obtains sending to the control unit, the cylinder withdraws after the oiling work.

2. The oil injection system for machine vision guided robot tracking of claim 1, wherein: the control unit comprises a servo driver and a controller, the servo driver is in communication connection with the multi-axis robot, and the oil injection unit, the camera and the encoder are in communication connection with the controller.