WO2021208230A1

WO2021208230A1 - Intelligent assembly control system

Info

Publication number: WO2021208230A1
Application number: PCT/CN2020/097229
Authority: WO
Inventors: 杨皓; 张伯强; 方宇; 陶翰中; 周志峰; 吴明晖
Original assignee: 上海工程技术大学
Priority date: 2020-04-15
Filing date: 2020-06-19
Publication date: 2021-10-21
Also published as: CN111331367B; CN111331367A

Abstract

Disclosed is an intelligent assembly control system, in particular, a fully-automatic precise measurement and intelligent assembly integrated system. The intelligent assembling control system comprises a mobile robot (8), a precise dual-frequency damping vibration-isolation optical platform (4), a five-degree-of-freedom electric sliding platform (5), a line laser scanner (3), a first deep learning camera (7), a second deep learning camera, a ball screw (2), a first mechanical arm (1), a second mechanical arm (8-1) and an upper computer. The system can automatically mount a mobile phone backplate wiring plug-in, such that a fully automatic workflow of dispatching, assembling and then measuring a component is realized, labor costs are saved on, the production efficiency is increased, and a drag teaching operation is realized.

Description

Intelligent assembly control system

Technical field

The invention relates to the technical field of intelligent manufacturing, in particular to an intelligent assembly control system.

Background technique

With the continuous development of the robotics field, intelligent assembly and high-precision testing equipment are gradually popularized. People save a lot of labor costs in production and greatly improve production efficiency.

However, due to the widespread application of complex shapes in 3C and other industries, traditional robot perception and intelligent control cannot meet the increasingly complex operating requirements, such as the detection and assembly of curved glass, the assembly and testing of OLED flexible screens, and the precision measurement of large-format screens. As with assembly, the degree of automation is still not high enough, and many links require manual intervention. In this way, the assembly and inspection accuracy is insufficient, and the safety of the working environment cannot be guaranteed.

For operating objects in the same operating space with different fields of view, robot environment perception and modeling methods based on information fusion such as vision and force have the problems of low efficiency, difficulty in cross-scale, and weak anti-interference ability; at the level of robot assembly motion planning, How to realize the autonomous planning and assembly of "machine-environment" with force position servo sensing to adapt to the narrow and brittle environment in the assembly of precision devices is an urgent problem to be solved. Therefore, the study of key technologies such as precision sensing and compliance control of high-reflective device assembly robots is of great significance to solving technical problems in industry applications.

Summary of the invention

The purpose of the present invention is to provide an intelligent assembly control system to solve the problem that the existing robot environment perception and modeling method cannot operate the operating objects in the same operating space and different field of view ranges.

In order to solve the above technical problems, the present invention provides an intelligent assembly control system. The intelligent assembly control system includes a mobile robot, a precision dual-frequency damped vibration isolation optical platform, a five-degree-of-freedom electric sliding table, a line laser scanner, and a first depth Learning camera, second deep learning camera, ball screw, first robotic arm, second robotic arm and host computer, among them:

Build the precision dual-frequency damped vibration isolation optical platform, the five-degree-of-freedom electric sliding table, the line laser scanner, the ball screw, the first deep learning camera and the first mechanical arm are installed in the On the precision dual-frequency damping vibration isolation optical platform;

The second robot arm is installed on the mobile robot, and the second deep learning camera is installed at the end of the second robot arm; the mobile robot performs omni-directional scanning in the working area to construct map information, and set Determine the target area, working area, target position and working position of the workpiece to be tested;

After the mobile robot reaches the target area, the second robot arm is moved to the target position by teaching, and the second deep learning camera takes pictures of the workpiece to be measured, and extracts the features of the workpiece to be measured Point information to identify the workpiece to be tested, and transmit the identification data to the host computer for processing to obtain the relative position relationship between the second deep learning camera and the workpiece to be tested;

The positional relationship between the second robot arm and the workpiece to be measured is calculated through the robot motion processing matrix, and the converted coordinates of the workpiece to be measured are sent to the second robot arm, and the second robot arm performs Posture adjustment and grasping the workpiece to be measured;

The mobile robot moves to the working area, and the first deep learning camera above the five-degree-of-freedom electric sliding table takes pictures of the precision dual-frequency damping vibration isolation optical platform, and sends the platform information to the For a mobile robot, the second mechanical arm places and fixes the workpiece to be tested on a precision dual-frequency damping and vibration isolation optical platform according to the platform information;

The first robotic arm and the first deep learning camera perform hand-eye calibration, and the second deep learning camera takes pictures of the fixed workpiece to be tested, and converts the fixed data into the position information of the assembly point. Perform hand-eye calibration, and send the position information to the first robotic arm, and the vacuum chuck at the end of the first robotic arm installs the parts under the visual guidance of the first deep learning camera;

The five-degree-of-freedom electric sliding table transports the installed workpiece under the line laser scanner to perform omnidirectional and multi-angle detection, and generate point cloud data of each part of the workpiece to be tested. The line laser scanner The point cloud data is transmitted to the host computer for processing, and the processed point cloud data model is compared with the CAD model of the workpiece to analyze specific dimensional errors and defects of the workpiece to be tested, and judge whether the workpiece to be tested is qualified or not;

After the detection is completed, the upper computer sends a detection completion signal to the mobile robot, and the mobile robot moves to a new target area, grabs the workpiece, and places it in a designated position.

Optionally, in the intelligent assembly control system, the five-degree-of-freedom electric sliding table is installed in the central area of the precision dual-frequency damping and vibration isolation optical platform, and is integral to the precision dual-frequency damping and vibration isolation optical platform. Placed in a vertical state, and connected with the precision dual-frequency damping vibration isolation optical platform through a thread;

The intelligent assembly control system also includes a motion control card, the motion control card is located under the precision dual-frequency damping vibration isolation optical platform, the motion control card is connected to the joint cable of the five-degree-of-freedom electric sliding table, and is connected The upper computer, the upper computer sends instructions to the motion control card to control the movement of the five-degree-of-freedom electric sliding table; the motion control card adopts the pulse output type of pulse combined direction to realize multi-axis independent movement, with acceleration and deceleration, point position, and trajectory Motion planning function;

The line laser scanner is located above the central axis of the five-degree-of-freedom electric sliding table, and is parallel to the precision dual-frequency damping vibration isolation optical platform, and there is a fixed frame above the line laser scanner;

The ball screw is installed on a profile placed perpendicular to the precision dual-frequency damped vibration isolation optical platform, and is located on the right side of the line laser scanner;

The host computer is located on the precision dual-frequency damped vibration isolation optical platform, and is connected to the motion control card, the cable of the line laser scanner, the first mechanical arm, and the first deep learning camera;

The first mechanical arm is installed on the precision dual-frequency damping vibration isolation optical platform, and is located directly behind the guide rail of the five-degree-of-freedom electric sliding table;

The first deep learning camera is located directly above the five-degree-of-freedom electric sliding table, parallel to the precision dual-frequency damping vibration isolation optical platform, and the first deep learning camera is fixed on the camera frame.

Optionally, in the intelligent assembly control system, the mobile robot includes an AGV vehicle and a motion control cabinet, wherein:

The quota load of the AGV vehicle is 60-80 kilograms, the AGV vehicle has map construction and autonomous navigation functions, the AGV vehicle has a safety laser sensor, and is positioned by a safety laser sensor based on environment mapping, and autonomously selects a route;

The motion control cabinet is located above the AGV vehicle, and the second mechanical arm is located above the motion control cabinet;

The motion control cabinet is connected to the host computer through a wireless network, and the motion control cabinet receives the mobile robot status information and movement instructions sent by the host computer; the motion control cabinet is connected to the safety laser sensor for real-time Obtain the positioning data of the safety laser sensor.

Optionally, in the intelligent assembly control system, each joint of the first mechanical arm and the second mechanical arm has a force sensor, and the force sensor has a collision detection function; the first mechanical arm and The second mechanical arm can realize dragging and teaching operation; the mechanical arm and the host computer are connected through the TCP/IP protocol, and the script command string sent by the host computer is received, and the received script command is executed to complete the specified action ；

Both the first deep learning camera and the second deep learning camera have an intelligent random sorting module, which integrates 3D structured light, image analysis, and robot arm motion control. It can quickly identify different objects in three dimensions through 3D structured light. The position and posture of the spatial placement can accurately guide the first robotic arm and the second robotic arm to pick and place.

Optionally, in the intelligent assembly control system, the precision dual-frequency damping vibration isolation optical platform includes a table, a bracket, a dual-frequency damping vibration isolation mechanism, a height adjustment mechanism, and brake silent casters;

The table top is located above the support, and is used to carry the first mechanical arm, the five-degree-of-freedom electric sliding table, the line laser scanner, and the first deep learning camera;

The bracket is a supporting structure, and the four brackets are connected by two trusses in pairs;

The dual-frequency damping vibration isolation mechanism is located between the table and the support for vibration isolation;

A height adjustment mechanism and brake silent casters are provided under the bracket.

Optionally, in the intelligent assembly control system, the five-degree-of-freedom electric sliding table includes an X-axis electric linear sliding table, an X-axis electric swing sliding table, a Y-axis electric linear sliding table, and a Y-axis electric swing sliding table. And electric rotary sliding table;

The X-axis electric linear sliding table and the Y-axis electric linear sliding table are both composed of a dust cover, a sliding table ball screw, a linear slider guide rail, a coupling, a U-shaped base plate, and a servo motor; the sliding table ball wire The bar is installed in the middle of the U-shaped bottom plate; the dust cover is installed above the U-shaped bottom plate, which is made of stainless steel; the linear slider guide rail moves on the ball screw of the sliding table under the control of the servo motor;

The X-axis electric swing slide table and the Y-axis electric swing slide table are both composed of a first base, a first worm gear, an arc-shaped V-shaped guide rail, and a stepping motor; the arc-shaped V-shaped guide rail is located above the first base, Swinging around the first worm gear driven by the stepping motor;

The electric rotary sliding table is composed of a second base, a second worm gear, a backlash adjustment structure, and a cross ball collar; the cross ball collar can be used horizontally, vertically, and upside down; the backlash adjustment structure is used to reduce backlash .

Optionally, in the intelligent assembly control system, the line laser scanner is composed of a blue laser sensor, a cylindrical objective lens, a CMOS sensor, and a connector; the blue laser sensor emits a blue laser and passes through the column The surface objective lens is processed to realize the uniform distribution of light; the CMOS sensor is located inside the line laser scanner; the connector is located at the tail of the line laser scanner, which is used to connect the line laser scanner and the cable.

The ball screw is composed of a spiral dial, a lead screw, a nut, a slider, and a slide rail; the spiral dial has a calibrated size range, which can visually display the rotating distance; the lead screw adopts an external circulation transmission, The lead screw is the active body, and the nut converts the rotary motion into linear motion with the rotation angle of the lead screw. The slider moves repeatedly on the slide rail to realize the vertical movement of the line laser scanner; ball screw and line laser scanner It is fixed by the fixing frame.

Optionally, in the intelligent assembly control system, the host computer is a computer, which is used to display the real-time status of the mobile robot, including the position information display of the mobile robot during the navigation process and the remote setting of the target;

The host computer is also used to control the trajectory planning of the five-degree-of-freedom electric sliding table;

The host computer is also used to receive and process the three-dimensional point cloud data transmitted by the line laser scanner to obtain the test results of qualified or unqualified parts, and transmit the test results to the first robotic arm;

The host computer is also used to receive feedback information from the infrared sensor, and feed back the operating status and location of the five-degree-of-freedom electric sliding table;

The host computer is also used to process photos taken by the first deep learning camera, extract point information, and send the processed information to the first robotic arm.

Optionally, in the intelligent assembly control system, the hand-eye calibration method is to perform hand-eye calibration on the first robotic arm and the first deep learning camera through the Tsai-Lenz algorithm, so that the first The end effector of the robotic arm fixes the calibration board and keeps the calibration board in the picture of the first deep learning camera. The first deep learning camera shoots the calibration board to collect a series of pictures, and at the same time records the first robotic arm’s The pose information is one-to-one correspondence, and the spatial relationship between the first deep learning camera and the first robot arm is solved through the correspondence between the visual information and the first robot arm's pose.

Optionally, in the intelligent assembly control system, the odometer information is provided by the control system of the mobile robot. After the control system of the second manipulator receives it, the odometer data is transformed into the first Second, the displacement amount of the base coordinate system of the robot arm, and then according to the determined end pose provided by the workpiece to be tested and the displacement amount of the second robot arm base, the equivalent planning target after the corresponding update of the second robot arm is obtained, and the equivalent planning target is obtained. An updated end pose;

By solving the inverse kinematics of the second manipulator, the angle control target refined to each joint is obtained, and finally the joint angle is controlled according to the PID algorithm.

In the intelligent assembly control system provided by the present invention, a system for automatically installing the backplane cable plug-in of a mobile phone is proposed. It can realize the full automation of the parts from the factory to the assembly and inspection process, so as to achieve the purpose of saving labor costs, improving production efficiency and verifying multiple technical means in a true sense. The integrated system of full-automatic detection and intelligent assembly proposed by the present invention can realize full-automatic in the true sense. After the initial teaching operation, manual intervention is no longer required in the entire work flow, and at the same time, when someone enters the work area Later, the sensors equipped on the robotic arms and mobile robots will react to the approaching human body and stop moving, so that safety can also be guaranteed. An integrated system of fully automatic inspection and intelligent assembly integrates the automatic assembly and inspection of workpieces, which improves production efficiency while ensuring production quality. Each joint of the seven-degree-of-freedom manipulator is equipped with a force sensor, which can realize precise force control. At the same time, it can realize safe human-computer interaction, can accurately detect collisions, and ensure safety; it can realize dragging and teaching operation and assist programming software to improve programming efficiency.

Description of the drawings

Figure 1 is a schematic diagram of the overall structure of a fully automatic detection and intelligent assembly control system provided by a specific embodiment of the present invention;

2 is a schematic structural diagram of a precision dual-frequency damping vibration isolation optical platform provided by a specific embodiment of the present invention;

Figure 3 is a schematic structural diagram of a five-degree-of-freedom electric sliding table provided by a specific embodiment of the present invention;

4 is a schematic structural diagram of a line laser scanner provided by a specific embodiment of the present invention;

FIG. 5 is a schematic diagram of the structure of a ball screw and a profile provided by a specific embodiment of the present invention;

6 is a schematic diagram of the structure of a first mechanical arm and a second mechanical arm (seven degrees of freedom mechanical arm) provided by a specific embodiment of the present invention;

Fig. 7 is a schematic diagram of a mobile robot provided by a specific embodiment of the present invention;

FIG. 8 is a block diagram of the overall system structure provided by a specific embodiment of the present invention;

Figure 9 is a schematic diagram of a mobile robot control system provided by a specific embodiment of the present invention;

FIG. 10 is a technical road map of parts identification and positioning provided by a specific embodiment of the present invention;

FIG. 11 is a motion planning route map of a mobile robot provided by a specific embodiment of the present invention;

Shown in the picture: 1 first mechanical arm; 2 ball screw; 3 line laser scanner; 4 precision dual-frequency damping vibration isolation optical platform; 5 five-degree-of-freedom electric sliding table; 6 camera frame; 7 first deep learning camera 8 mobile robot; 1-1 the first (two) main structure of the robotic arm; 1-2 the first (two) front end structure of the robotic arm; 2-1 profile; 2-2 slider guide rail; 2-3 fixed base; 2 -4 Adjustable dial; 2-5 slider; 2-6 fixed frame; 2-7 screw; 4-1 precision dual-frequency damping vibration isolation optical platform table; 4-2 dual-frequency damping vibration isolation mechanism; 4- 3 bracket; 4-4 brake silent casters; 4-5 height adjustment mechanism; 5-1X axis swing direction sliding table structure; 5-2X axis linear direction sliding table structure; 5-3 rotary sliding table structure; 5-4Y axis linear Direction and swing direction sliding table structure; 8-1 second mechanical arm; 8-2 profile frame; 8-3 control cabinet; 8-4 AGV car.

Detailed ways

The intelligent assembly control system proposed by the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. According to the following description and claims, the advantages and features of the present invention will be clearer. It should be noted that the drawings all adopt a very simplified form and all use imprecise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention.

In the description of the present invention, it needs to be understood that the orientation or positional relationship indicated by the terms "upper" and "lower" are based on the orientation or positional relationship shown in the drawings, or are habitually placed when the product of the present invention is used. The orientation or position relationship, or the orientation or position relationship commonly understood by those skilled in the art, is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation or a specific orientation. The azimuth structure and operation cannot be understood as a limitation of the present invention. The "first", "second", etc. in the present invention are only used for distinguishing in description, and have no special meaning. In the description of the present invention, it should also be noted that, unless otherwise clearly defined and limited, the terms “set” and “installation” should be understood in a broad sense, for example, it may be a fixed connection or a detachable connection, or Integrally connected; it can be directly connected or indirectly connected through an intermediary. For those of ordinary skill in the art, the specific meanings of the above-mentioned terms in the present invention can be understood in specific situations.

The core idea of the present invention is to provide an intelligent assembly control system to solve the problem that the existing robot environment perception and modeling method cannot operate the operating objects in the same operating space and different field of view ranges.

In order to realize the above ideas, the present invention provides an intelligent assembly control system. The intelligent assembly control system includes a mobile robot, a precision dual-frequency damped vibration isolation optical platform, a five-degree-of-freedom electric sliding table, a line laser scanner, and a first depth Learning camera, second deep learning camera, ball screw, first manipulator, second manipulator and host computer, in which: the precision dual-frequency damping vibration isolation optical platform is built, the five-degree-of-freedom electric sliding table, The line laser scanner, the ball screw, the first deep learning camera, and the first mechanical arm are installed on the precision dual-frequency damping vibration isolation optical platform; the second mechanical arm is installed on the mobile On the robot, the second deep learning camera is installed at the end of the second mechanical arm; the mobile robot performs omni-directional scanning in the work area, constructs map information, and sets the target area, work area, and work area of the workpiece to be tested. Target position and working position; after the mobile robot reaches the target area, the second robot arm is moved to the target position through teaching, and the second deep learning camera takes pictures of the workpiece to be measured and extracts all The feature point information of the workpiece to be tested is used to identify the workpiece, and the identification data is transmitted to the host computer for processing to obtain the relative positional relationship between the second deep learning camera and the workpiece to be tested; The positional relationship between the second robot arm and the workpiece to be measured is calculated through the robot motion processing matrix, and the converted coordinates of the workpiece to be measured are sent to the second robot arm, and the second robot arm performs The posture is adjusted and the workpiece to be measured is grasped; the mobile robot moves to the working area, and the precision dual-frequency damping is damped by the first deep learning camera above the five-degree-of-freedom electric sliding table The vibration isolation optical platform takes pictures, and the platform information is sent to the mobile robot. The second mechanical arm places the workpiece to be tested on the precision dual-frequency damping vibration isolation optical platform and fixes it according to the platform information; A robotic arm performs hand-eye calibration with a first deep learning camera. The second deep learning camera takes pictures of a fixed workpiece to be tested, converts the fixed data into position information of assembly points, and performs hand-eye calibration on the position information , And send the position information to the first robotic arm, and the vacuum suction cup at the end of the first robotic arm will install the parts under the visual guidance of the first deep learning camera; the five-degree-of-freedom electric sliding table will be installed The workpiece to be tested is transported to the line laser scanner for omnidirectional and multi-angle detection to generate point cloud data of each part of the workpiece to be tested, and the line laser scanner transmits the point cloud data to the host computer After processing, the point cloud data model formed after processing is compared with the CAD model of the workpiece to analyze specific dimensional errors and defects of the workpiece to be tested, and to judge whether the workpiece to be tested is qualified or not; after the detection is completed, the upper computer sends a detection completion signal to the mobile The robot moves to a new target area by the mobile robot, grabs the workpiece, and places it in the designated position.

This embodiment provides an integrated system of fully automatic precision inspection and intelligent assembly, including a first mechanical arm 1, a ball screw 2, a line laser scanner 3, a precision dual-frequency damped vibration isolation optical platform 4, and a five-degree-of-freedom electric Slide table 5, camera stand 6, first deep learning camera 7, mobile robot 8, motion control card, upper computer. After the platform is built, the mobile robot 8 scans the work area in all directions while constructing map information, and sets the target position and work position of the workpiece to be tested. After reaching the position of the workpiece to be measured, the second robot arm 8-1 is taught to move to the target area where the workpiece is located. The second deep learning camera mounted on the end of the robot arm on the mobile robot takes pictures of the workpiece to be measured and extracts features Point information, identify the workpiece to be tested, transmit the data to the host computer for processing, and obtain the relative positional relationship between the second deep learning camera and the workpiece, and then calculate the relationship between the second robot arm and the workpiece through the mobile robot motion processing matrix Send the converted workpiece coordinates to the second manipulator 8-1, and the second manipulator 8-1 adjusts the position and posture and grabs the workpiece. The mobile robot 8 continues to move to the working position, the first deep learning camera 7 above the five-degree-of-freedom electric sliding table 5 takes pictures of the stage, and sends the position information to the mobile robot 8, and the second robotic arm 8-1 will be tested The workpiece is placed on the table and fixed. The first robotic arm 1 and the first deep learning camera 7 perform hand-eye calibration. After completion, the first deep learning camera 7 takes a picture of the fixed workpiece, and converts the data into position information of the assembly point, and then calibrates this through the hand-eye relationship. Position, so that the first robot arm 1 can accurately formulate actions, and send the position information to the first robot arm 1. The first robot arm 1 is guided by the first deep learning camera 7 through the vacuum suction cup at the end to perform the wiring Install. The five-degree-of-freedom electric sliding table 5 transports the installed workpiece to the line laser scanner 3 to perform omni-directional and multi-angle detection to generate point cloud data of each part of the part, and the line laser scanner 3 transmits the point cloud data to the host computer Process, compare the processed point cloud data model with the workpiece CAD model, analyze the specific size error and defect of the workpiece, and judge whether the workpiece is qualified or not. After the detection is completed, the upper computer sends a signal to the mobile robot 8, and the mobile robot 8 moves to a new working position, grabs the workpiece, and places it in a designated position.

The five-degree-of-freedom electric sliding table 5 is installed on a precision dual-frequency damping vibration isolation optical platform, placed in a vertical state with the precision dual-frequency damping vibration isolation optical platform 4, located in the middle area of the platform 4, and the precision dual-frequency damping vibration isolation optical platform The frequency damping and vibration isolation optical platform 4 is connected by threads.

The motion control card is located under the precision dual-frequency damping vibration isolation optical platform 4, and is connected to the sliding table joint cable while being connected to the upper computer.

The line laser scanner 3 is located above the central axis of the five-degree-of-freedom electric sliding table 5, and is parallel to the precision dual-frequency damping vibration isolation optical platform 4, and there is a fixed frame above the line laser scanner 3.

The ball and ball screw 2 is installed on a profile placed perpendicular to the precision dual-frequency damping and vibration isolation optical platform 4, and is located on the right side of the line laser scanner 3.

The host computer is located outside the working area, and is connected to the motion control card, the line laser scanner 3 cable, the first mechanical arm 1 and the first deep learning camera 7.

The first mechanical arm 1 is installed on the precision dual-frequency damping and vibration isolation optical platform 4, and is located directly behind the guide rail of the five-degree-of-freedom electric sliding table 5.

The first deep learning camera 7 is located directly above the five-degree-of-freedom electric sliding table 5, parallel to the precision dual-frequency damping vibration isolation optical platform 4, and the first deep learning camera 7 is fixed on the camera frame.

The deep learning camera at the end of the robotic arm is installed at the end of the second robotic arm 8-1, and is connected to the robotic arm 8-1 through a connecting piece.

The mobile robot 8 includes an AGV vehicle 8-4, a seven-degree-of-freedom manipulator 8-1, and a motion control cabinet 8-3. The AGV vehicle 8-4 has a load of up to 60 kg and is equipped with a safety laser sensor, has map construction and autonomous navigation functions, and uses a safety scanning laser sensor based on environment mapping for positioning and autonomous selection of a route. The motion control cabinet 8-3 is located above the AGV vehicle 8-4, the seven-degree-of-freedom manipulator 8-1 is located above the motion control cabinet 8-3, and the three are connected by a profile connecting frame 8-2. Among them, the controller and the upper computer are connected through a wireless network as shown in Figure 9 to receive the status information and movement instructions of the mobile robot sent by the upper computer; Ethernet is connected with the laser sensor module to obtain the data of the laser sensor in real time.

Each joint of the seven-degree-of-freedom manipulator arms 8-1 and 1 is equipped with a force sensor, which can realize precise force control. At the same time, it can realize safe human-computer interaction, can accurately detect collisions, and ensure safety; it can realize dragging and teaching operation and assist programming software to improve programming efficiency. The robot arm 1 and the host computer are connected through the TCP/IP protocol, and the script command string sent by the host computer is received through a specific programming interface, and the received script command is executed to complete the specified action.

The precision dual-frequency damping vibration isolation optical platform 4 as a whole is a three-layer sandwich honeycomb structure, which is divided into a table 4-1, a bracket 4-3, a dual-frequency damping vibration isolation mechanism 4-2, a height adjustment mechanism 4-5, and a belt The brake is composed of 4-4 silent casters. The table top 4-1 is located at the top of the platform, and has a three-layer sandwich honeycomb structure inside, which uses ferromagnetic stainless steel, which has good corrosion resistance. The dual-frequency damping vibration isolation mechanism 4-2 is located under the table 4-1, in the middle position of the table 4-1 and the bracket 4-3, and has the effect of vibration isolation. The bracket 4-3 adopts an integral welding process. The shape of the long hair body is a four-support structure. The four supports are connected by two trusses, which have good rigidity and stability. Below the supports are equipped with a height adjustment mechanism 4-5 and silent casters 4-4 with brakes. The height adjustment mechanism 4-5 is under each support leg, and the bottom is oblate, which increases the contact area with the ground. By adjusting the upper and lower distance of the height adjustment mechanism 4-5, the problem of bracket distortion and deformation caused by uneven ground can be solved. The silent caster 4-4 is located below the lower truss, and is connected to the truss by four bolts, which facilitates movement and handling. The precision dual-frequency damped vibration isolation optical platform provides good rigidity and Vibration isolation performance.

The five-degree-of-freedom electric sliding table 5 consists of four parts, namely the electric linear sliding table 5-2 in the X-axis direction, the electric swing sliding table 5-1 in the X-axis swing direction, and the electric linear sliding table in the Y-axis direction. 5-4, Y-axis swing direction electric swing slide 5-4, and electric rotary slide 5-5. The X-axis electric linear sliding table 5-2 is composed of a dust cover, a sliding table ball screw, a linear slider guide rail, a coupling, a U-shaped bottom plate, and a servo motor. The U-shaped bottom plate is made of aluminum alloy and the surface is oxidized. Processing, good wear resistance; the ball screw of the sliding table is installed in the middle of the U-shaped bottom plate, and the positioning accuracy is high; the dust cover is installed above the U-shaped bottom plate, which is made of stainless steel; the linear slider guide rail is driven by the servo motor to slide Move on the ball screw. The X-axis electric swing sliding table 5-1 and the Y-axis electric swing sliding table 5-4 are composed of a base, a worm gear, an arc-shaped V-shaped guide rail, and a stepping motor. The base is made of aluminum alloy and the surface has been oxidized; the arc V-shaped guide rail is located above the base and has a strong load capacity. It swings around the worm wheel and worm driven by a stepping motor with high positioning accuracy. The electric rotary sliding table 5-5 is composed of a base, a worm gear, a clearance adjustment structure, and a cross ball collar. The base is made of hard aluminum alloy, which has good wear resistance. The worm wheel is made of tin bronze material, and the worm is made of steel material, which has high hardness and good rigidity; the guiding mechanism adopts a cross ball collar, which can be used horizontally, vertically, and upside down; the clearance structure effectively reduces the backlash and ensures smooth operation.

The five-degree-of-freedom electric sliding table 5 can perform real-time scanning and detection of parts from multiple angles and all directions, improving the molding efficiency and scanning accuracy of the parts, making the overall outline of the parts clearer, and facilitating the judgment and analysis of part errors.

The motion control card adopts the pulse output type of pulse + direction (PUL+DIR), can realize multi-axis independent motion, and has the functions of acceleration and deceleration, point position, and trajectory motion planning.

The line laser scanner 3 is composed of a blue laser sensor, a cylindrical objective lens, a CMOS sensor, and a connector. The blue laser sensor emits blue laser light, which is processed by the cylindrical objective lens to realize the uniform distribution of light; the CMOS sensor is located inside the line laser scanner, which has high speed and high dynamic range; the connector is located at the tail of the line laser scanner for Connect the line laser scanner and cable.

The rolling ball screw 2 is composed of a spiral dial 2-4, a screw 2-7, a nut, a ball, a pre-compression piece, a slide rail 2-2, a reverser, and a dustproof device. There is a calibrated size range on the spiral dial 2-4, which can visually display the distance rotated; the screw 2-7 adopts external circulation transmission to ensure the overall transmission technology. The screw is the active body, and the nut is with the screw. The rotation angle of φ converts the rotational motion into linear motion, and the slider 2-5 moves repeatedly on the slider rail 2-2 to realize the vertical movement of the line laser scanner 3 up and down. The ball screw 2 and the line laser scanner 3 are fixed by a fixing frame 2-6.

The first deep learning camera 7 has a smart random sorting module, which integrates 3D structured light, image analysis, and robot arm motion control. It can quickly identify the position and posture of different objects in three-dimensional space through the 3D structured light. , Can accurately guide the robot arm to pick and place.

The host computer is a computer, used to display the real-time status of the robot, including the position information display of the mobile robot 8 during the navigation process, and the remote setting of the target; used to control the trajectory planning of the five-degree-of-freedom electric sliding table 5; Receive and process the three-dimensional point cloud data transmitted by the line laser scanner 3 to obtain the qualified or unqualified test results of the parts, and transmit the test results to the robotic arm 8-1; used to receive the feedback information from the infrared sensor and feed back the electric sliding table 5 The operating status and location of the camera; used to process the photos taken by the first deep learning camera 7, extract point information, and send the processed information to the robotic arm 1.

The hand-eye calibration method is to perform hand-eye calibration of the robotic arm 1 and the first deep learning camera 7 through the Tsai-Lenz algorithm, so that the end effector of the robotic arm 1 is fixed to the calibration board, and the calibration board is kept on the camera screen. In, the first deep learning camera 7 takes a series of pictures taken by the calibration board, and records the pose information of the robotic arm at the same time, and matches them one by one. The camera 7 is compared with the corresponding relationship between the visual information and the pose 1 of the robotic arm. Solve the spatial relationship with robotic arm 1. Figure 10 shows the information transfer process between the deep learning camera and the robotic arm.

The movement plan of the mobile robot 8 is shown in Fig. 11. First, the control system of the mobile robot 8 provides the odometer information. After the control system of the robot arm receives it, the odometer data is transformed into a machine by coordinate transformation. The displacement of the base coordinate system of the arm, and then according to the determined end pose provided by the workpiece to be measured and the displacement of the base of the manipulator, the equivalent planning target after the corresponding update of the manipulator is obtained, and an updated end pose is obtained. . By solving the inverse kinematics of the manipulator, the angle control target refined to each joint is obtained, and finally the joint angle is controlled according to the PID algorithm. The system is refreshed at a certain frequency to ensure the flexibility and accuracy of mobile operations.

Preferably, the clamping jaws at the end of the robotic arm can be replaced according to requirements, and are not limited to one type; the clamps on the electric sliding table can be replaced according to the type of the workpiece to be tested, and are not limited to one type; the mobile machinery The deep learning camera mounted on the end of the arm can be replaced with different models according to needs, and it is not limited to one.

In summary, the above embodiments describe in detail the different configurations of the intelligent assembly control system. Of course, the present invention includes but is not limited to the configurations listed in the above embodiments, and any configuration is performed on the basis of the configurations provided in the above embodiments. The content of the change belongs to the protection scope of the present invention. Those skilled in the art can draw inferences based on the content of the above-mentioned embodiments.

The above description is only a description of the preferred embodiments of the present invention and does not limit the scope of the present invention in any way. Any changes or modifications made by a person of ordinary skill in the field of the present invention based on the above disclosure shall fall within the protection scope of the claims.

Claims

An intelligent assembly control system, characterized in that the intelligent assembly control system includes a mobile robot, a precision dual-frequency damped vibration isolation optical platform, a five-degree-of-freedom electric sliding table, a line laser scanner, a first deep learning camera, and a second Deep learning camera, ball screw, first robot arm, second robot arm and host computer, including:

Build the precision dual-frequency damped vibration isolation optical platform, the five-degree-of-freedom electric sliding table, the line laser scanner, the ball screw, the first deep learning camera, and the first mechanical arm are installed in the On the precision dual-frequency damping vibration isolation optical platform;

The second robot arm is installed on the mobile robot, and the second deep learning camera is installed at the end of the second robot arm; the mobile robot performs omni-directional scanning in the working area to construct map information, and set Determine the target area, working area, target position and working position of the workpiece to be tested;

After the mobile robot reaches the target area, the second robot arm is moved to the target position by teaching, and the second deep learning camera takes pictures of the workpiece to be measured, and extracts the features of the workpiece to be measured Point information to identify the workpiece to be tested, and transmit the identification data to the host computer for processing to obtain the relative position relationship between the second deep learning camera and the workpiece to be tested;

The positional relationship between the second robot arm and the workpiece to be measured is calculated through the robot motion processing matrix, and the converted coordinates of the workpiece to be measured are sent to the second robot arm, and the second robot arm performs Posture adjustment and grasping the workpiece to be measured;

The mobile robot moves to the working area, and the first deep learning camera above the five-degree-of-freedom electric sliding table takes pictures of the precision dual-frequency damping vibration isolation optical platform, and sends the platform information to the For a mobile robot, the second mechanical arm places and fixes the workpiece to be tested on a precision dual-frequency damping and vibration isolation optical platform according to the platform information;

The first robotic arm and the first deep learning camera perform hand-eye calibration, and the second deep learning camera takes pictures of the fixed workpiece to be tested, and converts the fixed data into the position information of the assembly point. Perform hand-eye calibration, and send the position information to the first robotic arm, and the vacuum chuck at the end of the first robotic arm installs the parts under the visual guidance of the first deep learning camera;

The five-degree-of-freedom electric sliding table transports the installed workpiece under the line laser scanner to perform omnidirectional and multi-angle detection, and generate point cloud data of each part of the workpiece to be tested. The line laser scanner The point cloud data is transmitted to the host computer for processing, and the processed point cloud data model is compared with the CAD model of the workpiece to analyze specific dimensional errors and defects of the workpiece to be tested, and judge whether the workpiece to be tested is qualified or not;

After the detection is completed, the upper computer sends a detection completion signal to the mobile robot, and the mobile robot moves to a new target area, grabs the workpiece, and places it in a designated position.
The intelligent assembly control system of claim 1, wherein the five-degree-of-freedom electric sliding table is installed in the central area of the precision dual-frequency damping vibration isolation optical platform, and is connected to the precision dual-frequency damping vibration isolation optical platform. The platform is placed in a vertical state, and is connected with the precision dual-frequency damping and vibration isolation optical platform through threads;

The intelligent assembly control system also includes a motion control card, the motion control card is located under the precision dual-frequency damping vibration isolation optical platform, the motion control card is connected to the joint cable of the five-degree-of-freedom electric sliding table, and is connected The upper computer, the upper computer sends instructions to the motion control card to control the movement of the five-degree-of-freedom electric sliding table; the motion control card adopts the pulse output type of pulse combined direction to realize multi-axis independent movement, with acceleration and deceleration, point position, and trajectory Motion planning function;

The line laser scanner is located above the central axis of the five-degree-of-freedom electric sliding table, and is parallel to the precision dual-frequency damping vibration isolation optical platform, and there is a fixed frame above the line laser scanner;

The ball screw is installed on a profile placed perpendicular to the precision dual-frequency damped vibration isolation optical platform, and is located on the right side of the line laser scanner;

The host computer is located on the precision dual-frequency damped vibration isolation optical platform, and is connected to the motion control card, the cable of the line laser scanner, the first mechanical arm, and the first deep learning camera;

The first mechanical arm is installed on the precision dual-frequency damping vibration isolation optical platform, and is located directly behind the guide rail of the five-degree-of-freedom electric sliding table;

The first deep learning camera is located directly above the five-degree-of-freedom electric sliding table, parallel to the precision dual-frequency damping vibration isolation optical platform, and the first deep learning camera is fixed on the camera frame.
The intelligent assembly control system of claim 1, wherein the mobile robot includes an AGV vehicle and a motion control cabinet, wherein:

The quota load of the AGV vehicle is 60-80 kilograms, the AGV vehicle has map construction and autonomous navigation functions, the AGV vehicle has a safety laser sensor, and is positioned by the safety laser sensor based on environment mapping, and autonomously selects a route;

The motion control cabinet is located above the AGV vehicle, and the second mechanical arm is located above the motion control cabinet;

The motion control cabinet is connected to the host computer through a wireless network, and the motion control cabinet receives the mobile robot status information and movement instructions sent by the host computer; the motion control cabinet is connected to the safety laser sensor for real-time Obtain the positioning data of the safety laser sensor.
The intelligent assembly control system of claim 1, wherein each joint of the first mechanical arm and the second mechanical arm has a force sensor, and the force sensor has a collision detection function; the first mechanical arm The arm and the second mechanical arm can realize the drag teaching operation; the mechanical arm and the upper computer are connected through the TCP/IP protocol, the script command string sent by the upper computer is received, and the received script command is executed to complete Designated action

Both the first deep learning camera and the second deep learning camera have an intelligent random sorting module, which integrates 3D structured light, image analysis, and robot arm motion control. It can quickly identify different objects in three dimensions through 3D structured light. The position and posture of the spatial placement can accurately guide the first robotic arm and the second robotic arm to pick and place.
The intelligent assembly control system of claim 1, wherein the precision dual-frequency damping vibration isolation optical platform includes a table, a bracket, a dual-frequency damping vibration isolation mechanism, a height adjustment mechanism, and a brake silent caster;

The table top is located above the support, and is used to carry the first mechanical arm, the five-degree-of-freedom electric sliding table, the line laser scanner, and the first deep learning camera;

The bracket is a supporting structure, and the four brackets are connected by two trusses in pairs;

The dual-frequency damping vibration isolation mechanism is located between the table and the support for vibration isolation;

A height adjustment mechanism and brake silent casters are provided under the bracket.
The intelligent assembly control system of claim 1, wherein the five-degree-of-freedom electric sliding table includes an X-axis electric linear sliding table, an X-axis electric swing sliding table, a Y-axis electric linear sliding table, and a Y-axis electric swing Slide table and electric rotary slide table;

The X-axis electric linear sliding table and the Y-axis electric linear sliding table are both composed of a dust cover, a sliding table ball screw, a linear slider guide rail, a coupling, a U-shaped base plate, and a servo motor; the sliding table ball wire The bar is installed in the middle of the U-shaped bottom plate; the dust cover is installed above the U-shaped bottom plate, which is made of stainless steel; the linear slider guide rail moves on the ball screw of the sliding table under the control of the servo motor;

The X-axis electric swing slide table and the Y-axis electric swing slide table are both composed of a first base, a first worm gear, an arc-shaped V-shaped guide rail, and a stepping motor; the arc-shaped V-shaped guide rail is located above the first base, Swinging around the first worm gear driven by the stepping motor;

The electric rotary sliding table is composed of a second base, a second worm gear, a backlash adjustment structure, and a cross ball collar; the cross ball collar can be used horizontally, vertically, and upside down; the backlash adjustment structure is used to reduce backlash .
The intelligent assembly control system of claim 1, wherein the line laser scanner is composed of a blue laser sensor, a cylindrical objective lens, a CMOS sensor, and a connector; the blue laser sensor emits a blue laser, After the cylindrical objective lens is processed, the light is evenly distributed; the CMOS sensor is located inside the line laser scanner; the connector is located at the tail of the line laser scanner, and is used to connect the line laser scanner and the cable.

The ball screw is composed of a spiral dial, a lead screw, a nut, a slider, and a slide rail; the spiral dial has a calibrated size range, which can visually display the rotating distance; the lead screw adopts an external circulation transmission, The lead screw is the active body. The nut converts the rotational motion into linear motion with the rotation angle of the lead screw. The slider moves repeatedly on the slide rail to realize the vertical movement of the line laser scanner; ball and ball screw and line laser scanning The instrument is fixed by the fixing frame.
The intelligent assembly control system according to claim 1, wherein the host computer is a computer, which is used to display the real-time status of the mobile robot, including the display of the position information of the mobile robot during the navigation process and the remote setting of the target;

The host computer is also used to control the trajectory planning of the five-degree-of-freedom electric sliding table;

The host computer is also used to receive and process the three-dimensional point cloud data transmitted by the line laser scanner to obtain the test results of qualified or unqualified parts, and transmit the test results to the first robotic arm;

The host computer is also used to receive feedback information from the infrared sensor, and feed back the operating status and location of the five-degree-of-freedom electric sliding table;

The host computer is also used to process photos taken by the first deep learning camera, extract point information, and send the processed information to the first robotic arm.
The intelligent assembly control system of claim 1, wherein the hand-eye calibration method is to perform hand-eye calibration on the first robotic arm and the first deep learning camera through the Tsai-Lenz algorithm, so that the The end effector of the first robotic arm fixes the calibration board and keeps the calibration board in the picture of the first deep learning camera. The first deep learning camera shoots the calibration board to collect a series of pictures, and at the same time records the first machine The posture information of the arms is one-to-one correspondence, and the spatial relationship between the first deep learning camera and the first robotic arm is solved through the correspondence between the visual information and the first robotic arm's posture.
The intelligent assembly control system of claim 1, wherein the odometer information is provided by the control system of the mobile robot, and when the control system of the second manipulator receives it, the odometer data is first transformed into coordinates. Is the displacement of the base coordinate system of the second manipulator, and then according to the determined end pose provided by the workpiece to be tested and the displacement of the second manipulator base, the equivalent planning target after the corresponding update of the second manipulator is obtained , Get an updated end pose;

By solving the inverse kinematics of the second manipulator, the angle control target refined to each joint is obtained, and finally the joint angle is controlled according to the PID algorithm.