A ROS-based telescopic structure robot system

By designing a ROS-based telescopic robot system, using a simple structure and an STM32 controller, the problems of high cost and insufficient flexibility of existing robot systems are solved. This system achieves low-cost, high-efficiency object grasping and three-dimensional spatial movement, suitable for home, medical, and commercial scenarios.

CN119635588BActive Publication Date: 2026-11-10ZHAOQING UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411888670.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-12-20
Publication Date
2026-11-10
Estimated Expiration
2044-12-20

AI Technical Summary

Technical Problem

Existing robot systems are complex in structure and expensive, making them difficult to widely promote in daily application scenarios. They also have low flexibility and cannot meet diverse object grasping needs.

Method used

A ROS-based telescopic robot system was designed, including a mobile chassis, a vertical lift, a telescopic arm, and a gripper. It adopts a simple structure and control method, and combines an STM32 controller and computer vision to achieve flexible three-dimensional spatial movement and object grasping.

Benefits of technology

It has developed a low-cost, easy-to-develop and controllable robot system that can meet the object grasping needs in home, medical and commercial scenarios, and has flexible three-dimensional spatial movement capabilities and automatic grasping functions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119635588B_ABST
    Figure CN119635588B_ABST
Patent Text Reader

Abstract

This invention discloses a ROS-based telescopic robot system, comprising a mobile chassis, a vertical lift mounted on the mobile chassis, a telescopic arm, and a gripper. The telescopic robot system is designed based on ROS2 and includes a vision recognition and control module. The main controller is an STM32 microcontroller, utilizing the OpenCTR B20 control board from Tucker Innovations. Compared to traditional complex and expensive robot systems, this telescopic robot features a simpler structure and control method, making its development easier and less costly. Controlled by ROS combined with an STM32 microcontroller, it can move in any direction and grasp objects within its effective range, easily meeting the object grasping needs of various everyday scenarios. Furthermore, combined with computer vision, it can achieve automatic object grasping, meeting the needs of home, medical, and commercial applications, and is easier to promote and commercialize.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of robotics, specifically to a retractable structure robot system based on ROS. Background Technology

[0002] In recent years, robotics technology has made rapid progress, expanding from the industrial sector to daily life and providing humans with a range of intelligent services, such as cleaning, food delivery, companionship, and home assistance robots. These service robots are now an indispensable part of modern society, assisting humans in completing tasks in various fields such as homes, healthcare, and commerce, significantly reducing the burden on human resources and improving the quality of life.

[0003] Traditional industrial robots mainly refer to large robotic arms, which are typically fixed in specific positions to perform simple grasping and placement tasks. However, with technological advancements and the development of supply chains, smaller robots, such as robotic vacuum cleaners and food delivery robots, have rapidly captured the market, allowing people to directly experience the convenience brought by technology. In recent years, robotics technology has developed rapidly, with robots represented by small, high-precision, and flexible robotic arms becoming a research focus for enterprises and research institutions. Some companies have begun researching and even developing humanoid robots. These robots rely on high-precision, high-torque servo motors, and through multi-axis joints driven by multiple servo motors, they can flexibly adapt to various application scenarios, such as precise object grasping and handling.

[0004] However, existing robots still have some shortcomings: While robots are increasingly used in service scenarios such as home and healthcare, the most traditional examples include sweeping and food delivery robots, whose functions are relatively limited and only applicable to specific situations. More advanced intelligent robots, such as AI-powered smart home robots with robotic arms, offer a wider range of functions. For example, Samsung's Bot Handy is an AI-powered smart home robot with a cylindrical design that allows for vertical lifting and lowering. Equipped with a multi-axis robotic arm, it can perform tasks such as clearing the table, pouring wine, or putting dishes in the dishwasher. However, these robots often cost hundreds of thousands of RMB, partly due to high hardware costs. The robotic arm itself costs tens of thousands of RMB, and the complex control algorithms further increase the overall cost, making widespread adoption time-consuming. Service robots, both domestically and internationally, mostly adopt robotic arm solutions. While robotic arms have many advantages, their high technological cost presents challenges for large-scale deployment. To address this issue, the American company Hello Robot launched Stretch, a collaborative robot with a telescopic structure. The Stretch robot abandons the traditional robotic arm approach, instead employing a simple telescopic arm and lifting rod design, combined with a cart chassis to construct a three-dimensional coordinate system (X, Y, Z). By controlling the movement of the telescopic arm, lifting rod, and cart chassis, the Stretch robot's end effector can reach any position in three-dimensional space. The Stretch robot's three-axis motion is similar to that of a 3D printer, achieving point-to-point movement through extension or retraction. This places low demands on the drive motors and does not require complex motion calculations for control. While the Stretch robot is sufficient for most everyday scenarios, its flexibility is relatively low, and it cannot move at high speeds.

[0005] Therefore, there is an urgent need to research a product that is simple in structure, easy to control, inexpensive to develop, can meet the needs of most daily application scenarios, and can more quickly promote the integration of intelligent robot products into people's lives. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to overcome the defects of the above-mentioned technologies and provide a telescopic structure robot system based on ROS.

[0007] To solve the above-mentioned technical problems, the technical solution provided by the present invention is a telescopic structure robot system based on ROS: including a mobile chassis, a vertical lift installed on the mobile chassis, a telescopic arm, and a gripper;

[0008] The mobile chassis includes a wheel mounting base, multiple support columns fixed above the wheel mounting base, an upper platform fixed to the upper end of the support columns, two drive wheels, and two driven wheels. The two drive wheels are symmetrically arranged at both ends of the wheel mounting base and are independently driven by their respective connected motors. The two driven wheels are symmetrically arranged at the other two ends of the wheel mounting base.

[0009] The vertical lifting platform includes a 360-degree rotating base, two sets of screw lifting mechanisms, and two sets of connecting modules. The 360-degree rotating base relies on a hollow rotary bearing as its support foundation and is equipped with a stepper motor and gear linkage to achieve free rotation in all directions. The screw lifting mechanism is assembled from 2020 aluminum profiles, 8mm screws, support seats, and sliders. The telescopic arm is connected to two sliders through two sets of connecting modules. The function of the connecting modules is to connect the sliders and the telescopic arm. When the two sliders move synchronously, the overall lifting state of the telescopic arm is controlled. When the two sliders move with a height difference, the overall pitch state of the telescopic arm is controlled.

[0010] The telescopic robot system is designed based on ROS2 and includes a vision recognition and control module.

[0011] Furthermore, the telescopic arm is constructed from four square aluminum alloy sleeves, each 180mm in length. The outer diameters of the four sleeves decrease sequentially, being 45mm, 40mm, 35mm, and 30mm respectively, with a uniform thickness of 1.5mm. A small gap of 1mm is maintained between adjacent sleeves. An outer support seat and an inner support seat are installed on each side of the sleeve. Each outer support seat and inner support seat is equipped with eight miniature pulleys with a diameter of 5mm.

[0012] Furthermore, the gripper includes four servo motors, a connecting base, and a mechanical gripper. The mechanical gripper uses a commercially available product with a maximum opening of 92mm and a maximum gripping weight of 500g. It achieves four degrees of freedom through four servo motors: horizontal rotation, vertical pitch, spin, and gripper tension.

[0013] Furthermore, the main controller is an STM32, and the main controller uses the OpenCTR B20 control board from Tucker Innovations.

[0014] The advantages of this invention compared to existing technologies are as follows: Compared to traditional complex and expensive robot systems, the telescopic robot has a simpler structure and control method, making its development simpler and less costly; controlled by ROS combined with an STM32 microcontroller, it can move in any direction and grasp within an effective range, easily meeting the needs of grasping objects in various daily scenarios; at the same time, combined with computer vision, it can achieve automatic object grasping; meeting the needs of home, medical and commercial application scenarios, it is easier to promote and commercialize. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the structure of a ROS-based telescopic robot system according to the present invention. Figure 1 .

[0016] Figure 2 This is a schematic diagram of the structure of a ROS-based telescopic robot system according to the present invention. Figure 2 .

[0017] Figure 3 This is a schematic diagram of the structure of a ROS-based telescopic robot system according to the present invention. Figure 3 .

[0018] Figure 4 This is a partial structural diagram of the mobile chassis portion of a ROS-based telescopic robot system according to the present invention.

[0019] Figure 5 This is a schematic diagram of the mobile chassis of a ROS-based telescopic robot system according to the present invention.

[0020] Figure 6 This is a partial structural diagram of the mobile chassis and vertical lifting platform of a ROS-based telescopic robot system according to the present invention. Figure 1 .

[0021] Figure 7 This is a partial structural diagram of the mobile chassis and vertical lifting platform of a ROS-based telescopic robot system according to the present invention. Figure 2 .

[0022] Figure 8 This is a schematic diagram of the telescopic arm of a ROS-based telescopic robot system according to the present invention. Figure 1 .

[0023] Figure 9 This is a schematic diagram of the telescopic arm of a ROS-based telescopic robot system according to the present invention. Figure 2 .

[0024] Figure 10This is a schematic diagram of the gripper structure of a ROS-based telescopic robot system according to the present invention.

[0025] Figure 11 This is a system design block diagram of a ROS-based telescopic structure robot system according to the present invention.

[0026] Figure 12 This is a functional block diagram of a ROS-based telescopic structure robot system according to the present invention.

[0027] Figure 13 This is a block diagram of the STM32 control software in this invention.

[0028] Figure 14 This is a schematic diagram of the ROS coordinate definition in this invention.

[0029] Figure 15 This is a schematic diagram of the chassis kinematics analysis of a ROS-based telescopic structure robot system according to the present invention.

[0030] Figure 16 This is a flowchart of the speed closed-loop control of a ROS-based telescopic structure robot system according to the present invention.

[0031] Figure 17 This is a schematic diagram illustrating the steps for setting up a ROS2 environment for a ROS-based telescopic structure robot system according to the present invention.

[0032] Figure 18 This is a built-in demo diagram for the test startup of a ROS-based telescopic structure robot system according to the present invention.

[0033] Figure 19 This is a ROS2 node relationship diagram of a ROS-based telescopic structure robot system according to the present invention.

[0034] Figure 20 This invention relates to a control interface for a ROS-based telescopic structure robot system.

[0035] Figure 21 This invention provides a component selection and editing interface for a ROS-based telescopic structure robot system.

[0036] Figure 22 This is a schematic diagram of the connection between the APP and the device of a ROS-based telescopic structure robot system according to the present invention.

[0037] Figure 23 This is a flowchart of the APP-side control program for a ROS-based telescopic structure robot system according to the present invention.

[0038] Figure 24 This is a result image of the image processing process of a ROS-based telescopic structure robot system according to the present invention.

[0039] Figure 25 This is a diagram showing the result of obtaining the coordinates of the center point of a rectangle using a ROS-based telescopic structure robot system according to the present invention.

[0040] Figure 26 This is a schematic diagram of image region division for a ROS-based telescopic structure robot system according to the present invention.

[0041] Figure 27 This is a flowchart of the identification and control process of a ROS-based telescopic structure robot system according to the present invention.

[0042] Figure 28 This is a motion analysis diagram of the grasping mechanism of a ROS-based telescopic structure robot system according to the present invention.

[0043] As shown in the figure:

[0044] 1. Mobile chassis; 101. Wheel mounting base; 102. Support column; 103. Upper platform; 104. Drive wheel; 105. Driven wheel; 2. Vertical lifting ladder; 201. Rotating base; 202. Screw lifting mechanism; 3. Telescopic arm; 4. Gripper. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0046] Example 1, in conjunction with Appendix Figure 1-28 A ROS-based telescopic structure robot system includes a mobile chassis 1, a vertical lift 2 mounted on the mobile chassis 1, a telescopic arm 3, and a gripper 4. The telescopic structure robot system is designed based on ROS2 and includes a vision recognition and control module.

[0047] The mobile chassis 1 includes a wheel mounting base 101, multiple support columns 102 fixed above the wheel mounting base 101, an upper platform 103 fixed to the upper end of the support columns 102, two drive wheels 104, and two driven wheels 105. The two drive wheels 104 are symmetrically arranged at both ends of the wheel mounting base 101 and are independently driven by their respective connected motors. The two driven wheels 105 are symmetrically arranged at the other two ends of the wheel mounting base 101. The mobile chassis 1 adopts a two-wheel differential drive, which is a common type of wheeled chassis, characterized by a relatively simple drive system and flexible movement. The two drive wheels 104 can be controlled independently; by controlling the different speeds of the two drive wheels 104, the chassis can achieve frictionless rotation and zero-radius turning. The driven wheels 105 are typically omnidirectional or caster wheels; omnidirectional wheels are selected in this design. Omnidirectional wheels are characterized by their ability to move in any direction without steering, offering relatively stable control, but their load-bearing capacity is relatively weak. While caster wheels have a relatively strong load-bearing capacity, their control is less stable, exhibiting a fishtailing effect during sharp turns. Considering all factors, using omnidirectional wheels as passive wheels and placing them symmetrically front and rear is more beneficial for control stability and movement flexibility. Eight support columns 102 are provided, distributed around the wheel mounting base 101, and made of copper, to support the upper platform 103. Using copper columns for support not only reduces design complexity and saves internal space but also ensures sufficient rigidity and stability for the entire system.

[0048] The vertical lifting platform 2 includes a 360-degree rotating base 201, two sets of lead screw lifting mechanisms 202, and two sets of connecting modules. The 360-degree rotating base 201 relies on a hollow rotary bearing as its support foundation and is equipped with a stepper motor, achieving free rotation in all directions through gear linkage. The lead screw lifting mechanism 202 is assembled from 2020 aluminum profiles, an 8mm lead screw, a support base, and a slider. The lead screw design significantly reduces torque requirements and possesses self-locking functionality and high-precision control characteristics. Driving the lead screw rotation via the stepper motor allows for precise control of the slider's lifting and lowering. The telescopic arm 3 is connected to two sliders via two sets of connecting modules. These connecting modules connect the sliders to the telescopic arm 3. When the two sliders move synchronously, the overall lifting and lowering state of the telescopic arm 3 is controlled; when the two sliders have a height difference, the overall pitch state of the telescopic arm 3 is controlled. Compared to traditional designs where a single motor controls a single state, this design cleverly achieves simultaneous control of both lifting and pitch states by coordinating the control of two motors. During lifting operations, the two motors work together to provide power, effectively distributing the demand on motor torque and structural inertia; and during the pitch adjustment phase, the lever principle is used, requiring only a small force to achieve precise attitude adjustment.

[0049] The telescopic arm 3 is constructed from four square aluminum alloy sleeves, each 180mm in length. The outer diameters of the four sleeves decrease sequentially: 45mm, 40mm, 35mm, and 30mm, respectively, with a uniform thickness of 1.5mm. A 1mm gap is maintained between adjacent sleeves. An outer support and an inner support are installed on each side of the sleeve. Each outer and inner support is equipped with eight 5mm diameter micro pulleys, effectively reducing frictional resistance between adjacent sleeves and compensating for gaps caused by spacing, thereby minimizing the sag at the end of the telescopic arm when fully extended. The telescopic arm 3 has a total length of 230mm when fully retracted and reaches 640mm when fully extended, with the sag at the end controlled within 30mm. The telescopic boom 3 uses a synchronous rope drive technology, employing a 0.35mm diameter high-wear-resistant PE wire and a specific pulley arrangement to construct a highly efficient synchronous tensioning mechanism. This mechanism can drive two sections of the sleeve to extend and retract synchronously. Since the telescopic boom 3 has three sections of sleeve that extend and retract synchronously, two such synchronous tensioning mechanisms are required. The entire extension and retraction process is powered by a stepper motor. The motor controls the winding and unwinding of the reel to tension or relax the drive wire, thereby achieving the extension and retraction of the telescopic boom 3.

[0050] The gripper 4 comprises four servo motors, a connecting base, and a mechanical gripper. The mechanical gripper uses a commercially available product with a maximum opening of 92mm and a maximum gripping weight of 500g. It achieves four degrees of freedom via the four servo motors: horizontal rotation, vertical pitch, spin, and gripper tension. Each degree of freedom is precisely controlled by an independent servo motor, resulting in flexible and accurate movement. Horizontal rotation allows the gripper to rotate 270 degrees relative to the horizontal direction; vertical pitch allows the gripper to adjust up and down by a certain angle in the vertical direction, adapting to gripping tasks at different heights; spin gives the gripper the ability to rotate on its own axis, enabling it to grasp objects with irregular shapes or in special positions; gripper tension is achieved by controlling the servo motors to drive the mechanical gripper to open and close, thus achieving a firm grip and release of the object.

[0051] The robot is mostly constructed from aluminum profiles, with a few parts manufactured using 3D printing. When fully retracted, the robot measures 70cm in height, 30cm in length, and 30cm in width. The vertical lift has a lifting range of 20cm to 62cm relative to the ground. The telescopic arm extends 25cm when fully retracted and 70cm when fully extended. The gripper consists of four servo actuators, enabling pitch, rotation, left and right turns, and gripping release. Its maximum gripping opening is 92mm, and its maximum weight capacity is 500g.

[0052] In this embodiment, a total of ten motors are used. For the mobile chassis 1, two 520 encoder DC geared motors are selected. These motors have high torque and are equipped with Hall effect encoders, allowing real-time acquisition of motor speed information for closed-loop speed control. The no-load speed of these motors is 178 RPM, the reduction ratio is 56, and the maximum torque is 9.0 kg·cm. The diameter of the drive wheel 104 is 58 mm. Even with the overall weight of the telescopic robot reaching 8 kg, these two encoder motors can still flexibly drive the robot's movement, achieving a maximum speed of 0.5 m / s. For the vertical lift 2 and the telescopic arm 3, 42 stepper motors are used. Stepper motors are characterized by high positioning accuracy, simple control, smooth movement, and extremely low cost. Four stepper motors are used in these two parts: one on the 360-degree rotating base 201, one on each of the two sets of lead screw lifting mechanisms 202, and one on the telescopic arm 3. Each motor has a body length of 34mm and a maximum torque of 2.9Kg.cm. Compared to the joint motors of other similar robots, the power is much less. However, because the vertical lifting platform 2 uses a lead screw combined with a lever and the telescopic arm 3 uses a rope-driven design, using this motor is still more than sufficient to drive the lifting and telescopic movements. However, stepper motors also have some drawbacks. They are prone to step loss under heavy loads, and the numerous controller wirings are not conducive to the overall robot wiring. To solve these problems, a closed-loop servo drive module, model Emm42_V5.0, sourced from ZDT closed-loop servo technology, is added to the bottom of all stepper motors. The Emm42_V5.0 module employs an advanced FOC vector closed-loop control algorithm, featuring a high-precision magnetic encoder and current sensor. It supports speed and position modes, curve acceleration / deceleration, and immediate stopping. Control is via serial communication, supporting multi-module communication. It also features stall protection and smooth, quiet operation. By adding the Emm42_V5.0 module, stepper motors can achieve absolute position control like servo motors, making lifting and telescopic control more precise and simpler. Wiring is also simplified, requiring only two wires to control multiple motors.

[0053] In this embodiment, the gripper 4 uses four 20kg serial bus servo motors to control horizontal rotation, vertical pitch, spin, and gripper tension, respectively. The servo motors are model ZP20D, interconnected via aluminum alloy brackets, with a maximum torque of 20kg·cm and a rotation angle of 270°, meeting the strength and flexibility requirements of the mechanical gripper. The control method is serial communication. The servo motors are connected in series via an interface; by setting different IDs in the commands, multiple servo motors can be controlled by a single data line, greatly simplifying wiring and control.

[0054] In this embodiment, motion control is performed directly through an STM32 control board. The STM32 control includes DC motor drive, encoder reading, PID control, and serial communication.

[0055] Kinematic analysis of mobile chassis

[0056] The mobile chassis of the telescopic robot is a two-wheel differential chassis. By controlling the difference in the rotational speed of the two driving wheels, the chassis will rotate around their common axis, which is differential steering.

[0057] In the ROS system, the coordinate system follows the right-hand rule. For the moving chassis, the X-axis points directly forward, the Y-axis points to the left, and the Z-axis represents counter-clockwise rotation. The moving chassis typically moves within the same plane and does not experience rolling or pitching motions on a horizontal plane; therefore, the motion analysis of the X and Y axes only considers linear velocity, not angular velocity. Similarly, the height of the moving chassis does not change, so the change in linear velocity along the Z-axis is not considered. The analysis is attached. Figure 15 As shown.

[0058] The parameters have the following meanings:

[0059] Robot vector velocity:

[0060] VX: The robot's forward and backward velocity at point O, with forward velocity being positive, unit: m / s.

[0061] VY: The robot's left and right velocities at point O, with leftward movement being positive. Unit: m / s. Two wheels are ignored.

[0062] VW: The robot's rotational speed around the target point O, with counterclockwise rotation being positive, unit: rad / s.

[0063] Chassis parameters:

[0064] W: Wheelbase, the distance between the two driving wheels, centered at point O, unit: m.

[0065] R: Turning radius R generated by the chassis moving forward and rotating simultaneously, unit: m.

[0066] Chassis wheel speed:

[0067] VA: Robot's left wheel speed, positive in the direction of travel, unit: m / s.

[0068] VB: Robot's right wheel speed, positive in the forward direction, unit: / S.

[0069] path

[0070] SL: The path traveled by the revolver within a certain time t.

[0071] SM: The path traversed by the intermediate point O within a certain time t.

[0072] SR: The path traveled by the right wheel within a certain time t.

[0073] θ: The angle of rotation of the robot within a certain time t, in rad.

[0074] The following are the forward and inverse kinematics formulas for solving the chassis. First, we can derive the formula from the integral of velocity over time being equal to distance:

[0075] S L =V A *t

[0076] S M =V X *t

[0077] S R =V B *t

[0078] From the equation that arc length divided by radius equals radians, we get:

[0079]

[0080] Right now:

[0081]

[0082] Integrating over time, we get:

[0083]

[0084] The decomposition yields the following results:

[0085]

[0086] Based on the above equations, the inverse kinematics formula can be obtained, given the target velocities VX and VW.

[0087] The speeds VA and VB of the driving wheels are obtained as follows:

[0088]

[0089] The inverse kinematics formula can be transformed into the forward kinematics formula. Given the velocities VA and VB of the driving wheel, the target velocities VX and VW can be obtained as follows:

[0090]

[0091] Mobile chassis control methods

[0092] The mobile chassis motor is controlled by an OpenCTR B20 control board. The motor connects to two onboard encoder motor interfaces. Each interface occupies four I / O ports on the STM32 microcontroller. Two of these I / O ports are used for PWM output to control the motor speed and forward / reverse rotation, while the other two I / O ports are used to read the voltage levels of the Hall encoder's A and B phases, respectively. The encoder motor control also utilizes a PID algorithm to achieve closed-loop speed control, as shown in Figure 16.

[0093] First, the target speeds of the chassis's X, Y, and Z axes are acquired. These three target speed parameters are defined as global variables. Control commands from the host computer are received via serial communication. The received commands are parsed to extract the target speed values ​​for each axis. After checking for out-of-limit values, the values ​​are assigned to the pre-defined global variables. Then, the target speeds of motors A and B are calculated using an inverse kinematics algorithm. These motor target speeds are then fed into the PID controller to calculate the PWM duty cycle required to control the motors to reach the target speeds. After the motors start running, the encoder provides real-time feedback of the current motor speed to the PID controller, which recalculates the PWM duty cycle value. This cycle repeats, thereby controlling the motors to run smoothly at the target speeds.

[0094] The PID controller described above functions to achieve closed-loop speed control. Its basic principle is incremental PID, which measures the motor speed per unit time and compares it with the target value to obtain the control deviation. Then, by controlling the proportional, integral, and derivative of this deviation, the deviation is brought towards zero. The complete formula for incremental PID is as follows:

[0095] PWM+=Kp[e(k)-e(k-1)]+Ki*e(k)+Kd[e(k)-2e(k-1)+e(k-2)]

[0096] Where e(k) is the current deviation, e(k-1) is the previous deviation, e(k-2) is the deviation from the previous time, Kp is the proportional term parameter, Ki is the integral term parameter, Kd is the derivative term parameter, and PWM represents incremental output. Since only PI control is used in this chassis closed-loop control system, the PID controller can be simplified to the following formula:

[0097] PWM+=Kp[e(k)-e(k-1)]+Ki*e(k)

[0098] The specific implementation is shown below. The input parameters are the encoder measurement value (actual speed) and the target speed, and the return value is the motor PWM. The PID controller code is as follows:

[0099]

[0100] The encoder measurement (actual speed) is achieved by capturing and counting the encoder signal through the input capture of the STM32 timer and external interrupt functions. Based on the motor and wheel parameters, the real-time wheel speed can be calculated, as follows:

[0101]

[0102] Where WHEEL_DIAMETER is the wheel diameter, WHEEL_RESOLUTION is the encoder resolution, Counter is the Hall encoder count value, PID_RATE is the PID frequency, and PI is pi.

[0103] Kinematic analysis of vertical elevator, telescopic arm and gripper

[0104] The vertical lifting platform consists of two identical screw-type lifting mechanisms responsible for controlling the raising, lowering, and tilting of the telescopic boom. First, we analyze a single screw-type lifting mechanism to determine the relationship between the stepper motor pulse count and the slider lifting amount. The specific relationship is as follows:

[0105]

[0106] Where CLK is the pulse count, L is the slider lifting amount, C is the single-turn pulse coefficient, and S is the lead screw. When the telescopic boom needs to move vertically, the two lifting mechanisms operate synchronously. When performing pitching motion, there will be a difference (Dif) in the lifting amount of the two lifting mechanisms. The magnitude and direction of the difference determine the magnitude and direction of the pitch angle, respectively.

[0107] First, convert the target pitch angle (Angle) into the difference in rise and fall:

[0108] θ = Angle; Dif = tanθ * d

[0109] The lifting amounts of lifting mechanisms A and B are calculated as follows:

[0110]

[0111] Where d is the distance between the connection points of the telescopic arm and the two sets of lifting mechanisms, and L a和 L b These are the current positions of the sliders of lifting mechanisms A and B, respectively.

[0112] The analysis of the telescopic boom requires determining the relationship between the motor pulse count (CLK) and the target extension length (Lr) of the telescopic boom. When the telescopic boom is fully retracted, Lr is 0. The specific relationship is as follows:

[0113]

[0114] Where CLK is the number of pulses, Lr is the target length, C is the single-turn pulse coefficient, and Tp is the diameter of the winding reel.

[0115] Each joint of the gripper is directly driven by a servo motor, requiring only one angle parameter, so no analysis or calculation will be performed for now.

[0116] In this section, the stepper motors used in the vertical lifting platform and telescopic boom, and the servo motors used in the grippers, are all controlled via a serial bus and communicate with serial port 3 on the OpenCTR B20 control board. The main control board sends control commands and receives position information from the motors through serial port 3. The stepper motor control communication protocol is shown in Table 1-1:

[0117] Table 1-1 Motor Control Communication Protocol

[0118]

[0119] All motors are connected in series. A specific motor is controlled by setting its address (ID_Addr). The default address is 1, and the configurable range is 1-255. 0 is the broadcast address. Different commands correspond to different function codes and instruction data. The commands used in this system are shown in Table 1-2:

[0120] Table 1-2 Motor Control Command Analysis

[0121]

[0122]

[0123] The specific control function for the vertical lift is as follows. The function takes the movement speed (vel) and absolute movement height (leng) as input parameters. After calculating the motor speed and pulse count based on these two parameters, all data is passed to the `Emm_V5_Pos_Control()` function. The function then integrates the commands according to the stepper motor protocol and sends them to the motor via serial port. The vertical lift control code is as follows:

[0124]

[0125] The specific control function for the telescopic boom is as follows. Its input parameters are the movement speed (vel) and the absolute telescopic length (leng). Similarly, the motor speed and pulse count are calculated based on these two parameters, integrated, and then sent to the motor. The telescopic boom control code is as follows:

[0126]

[0127]

[0128] Servo motors and stepper motors use different protocols. Stepper motor commands are in hexadecimal format, while servo motor commands are in string format. The specific protocols are as follows:

[0129] Table 1-3 Servo Control Protocol

[0130]

[0131] The specific implementation is as follows: the function is passed the servo address and the angle of motion as parameters. The servo control code is as follows:

[0132]

[0133] ROS2 Overview

[0134] ROS2 (Robot Operating System 2) is a standardized software framework designed specifically for robot development. It provides a standardized approach to handling coordination between multiple modules, simplifying communication and task allocation between multi-robot systems, reducing software complexity, and improving development efficiency. In ROS development, different software functions of a robot can be encapsulated into different nodes, each of which executes tasks independently.

[0135] ROS2's communication mechanism is one of its core functionalities, allowing for efficient and reliable data exchange between different nodes. Communication is conducted through topics, which are communication channels used for publishing and subscribing to messages. Nodes can provide data by publishing messages to topics, while other nodes can subscribe to these topics to receive data. In addition to the publish / subscribe model, ROS2 also supports services communication, which allows nodes to interact using a request / response pattern. Services communication is typically used in situations requiring interactive operations between nodes.

[0136] Setting up and using the ROS2 environment

[0137] The host computer hardware used in this solution is the Horizon Robotics X3 ARM development board, running Ubuntu 20.04 Desktop, supporting ROS2 Foxy. Python is used for development, as shown in the attached diagram. Figure 17 Steps to set up a ROS2 environment:

[0138] To test the built-in message sending and receiving demo, first execute `ros2 run demo_nodes_pylistener` in one terminal to start the listener node, and then execute `ros2 run demo_nodes_cpp talker` in another terminal to start the talker node. The two terminals will then communicate continuously, as shown in the attached diagram. Figure 18 As shown, for every message sent by the talker node, the listener node hears it each time.

[0139] ROS2 Node Overview

[0140] In the ROS system, a node is an executable file that independently performs tasks and can communicate with other nodes through the ROSMaster. A node can publish or receive a topic, and can also provide or use a service. This project has a total of 6 nodes, as shown in Table 1-4:

[0141] Table 1-4 Node Introduction

[0142]

[0143] Nodes interact with each other through topics, and the specific relationships are as follows: Figure 19 As shown, the ellipses represent node names, and the boxes represent topics. Topics act as bridges for communication between nodes. The keyboard node publishes control mode information through the mode_selet topic, and the Select_ctl node receives and sets the control mode. The Keyboard_Ctl, Blink_Ctl, and Vision_Ctl nodes publish robot control commands through the mode_cmd topic. The Select_ctl node determines whether to process the received command based on the set control mode. If processing is required, the command is further split into chassis and body parts and parsed into a format corresponding to the lower-level machine's communication protocol. These are then published through the cmd_vel and cmd_body topics, respectively. Finally, the Uart_Ctl node sends the command to the STM32 lower-level machine via serial port.

[0144] Specific functions and implementation of key nodes

[0145] Lower-level communication node

[0146] The main function of this node is to enable communication between the ROS system and the STM32 lower-level slave device, sending control commands. This communication is achieved via serial port. To ensure communication reliability and meet the diverse needs of command functions, a dedicated communication protocol is defined, as shown in Table 1-5:

[0147] Table 1-5 Lower-level machine communication protocols

[0148]

[0149] This communication protocol uses a double frame header and trailer, providing stronger anti-interference capabilities. The data is high-order first, and its length is variable, allowing for free combination of 8-bit, 16-bit, and 32-bit data. The checksum is the lower 8 bits of the sum of the preceding data. The following example frame transmits data as 1000, -1000: AA55|10|03E8FC18|XX. The function code definitions are shown in Table 1-6.

[0150] Table 1-6 Function Code Definitions

[0151]

[0152] Keyboard control node

[0153] Keyboard control is the most basic control method. It allows you to control the robot's movement and the motion of other parts. You can also select control modes using the keyboard; the V and B keys activate visual recognition control and mobile phone control modes respectively. The keyboard control node is implemented using the pynput library to listen for keyboard press and release events. The keyboard control code is as follows:

[0154] event_queue = queue.Queue()

[0155] def on_press(key): # Define a function to handle keyboard events

[0156] try:

[0157] event_queue.put(key.char,block=False)

[0158] except AttributeError:

[0159] event_queue.put(key,block=False)

[0160] def on_release(key): # Define a function to handle the key release event.

[0161] if key == keyboard.Key.esc:

[0162] #Press the Esc key to exit the listener

[0163] return False

[0164] event_queue.put(key,block=False)

[0165] listener = keyboard.Listener(on_press = on_press, on_release = on_release) # Create a keyboard listener listener.start() # Start listening for keyboard events

[0166] A keyboard listener is defined to store keyboard events in a queue for later processing. Whenever a key event occurs, whether a key is pressed or released, the listener adds the event to the queue. If the key is Esc, the listener stops working. Upon detecting a key event, the listener sends a corresponding message based on the key pressed. For example, pressing the W key sends a message to move the chassis forward, and releasing the W key sends a stop message.

[0167] Blinker Wireless Control Node

[0168] The primary function of the Blinker wireless control node is to enable remote control of robot movement via a mobile phone. Blinker is a professional and easy-to-use IoT solution launched by Diandeng Technology, providing server, application, and device SDK support. Using Blinker's simple and convenient application, along with its multi-device support SDK, allows for rapid device access and management.

[0169] Blinker wireless control enables the control of the robot's forward and backward movements, lifting, extension and retraction, and grasping actions. It also supports modifying parameters such as speed, as shown in the attached diagram. Figure 20 The image shows the mobile control interface. The Blinker mobile app provides modular component functionality, with rich and editable components that greatly reduce development difficulty, as shown in the attached image. Figure 21 The image shows the component selection and editing process.

[0170] The Blinker wireless control node is primarily designed using the Blinker Python SDK. The Blinker Python SDK internally implements MQTT communication, data transmission and reception, component creation, and other functions. Simply connecting the device to the network and configuring the device key enables communication between the Blinker mobile app and the device. Communication between the app and the device is via MQTT. The mobile app publishes topics to the Blinker server, and the device subscribes to relevant topics to send and receive data. The specific process is shown in the attached diagram. Figure 22 As shown.

[0171] After the app and device are connected, the relevant control component functions and component names need to be declared in the code, and the corresponding callback functions need to be registered. In the app, control buttons are created manually by dragging and dropping, and bound to the previously declared component names. Clicking a button in the app sends data to the server; the device receives the data and executes the corresponding callback function. The specific program flow is shown in the attached figure. Figure 23 As shown.

[0172] Visual recognition control node

[0173] The main function of the visual recognition control node is to realize color block recognition and following, as well as color block recognition and grasping. Images are captured by a camera, and OpenCV is used to process and analyze the images to obtain the coordinates of the color block relative to the horizontal plane. By calculating the distance and position between the color block and the robot, it determines whether to control the robot to move accordingly, stopping only when the color block is within a specified range, thus achieving the following function. Similarly, color block recognition and grasping are achieved by determining whether the color block is directly facing the gripper and the distance between the color block and the robot. When the color block is not on the robot's central axis, the robot is adjusted by rotating left or right. Then, the distance between the color block and the robot's center point is calculated, and the extension length of the telescopic arm is controlled according to the distance to achieve grasping.

[0174] Design of a control module based on vision recognition

[0175] Color block recognition based on OpenCV

[0176] Color block recognition can be achieved using the OpenCV library. OpenCV is an open-source computer vision and machine learning software library that provides rich image processing and analysis capabilities. The specific steps for color block recognition based on OpenCV are as follows:

[0177] First, capture video frames by opening the camera using `cv2.VideoCapture(0)` and reading one frame using `ret,frame = cap.read()`. Second, perform Gaussian blur and color space conversion on the image. Gaussian blur aims to reduce noise and enhance color consistency, improving subsequent processing. Color space conversion converts the BGR color space image to HSV, facilitating the extraction of individual colors. Third, perform color filtering using `cv2.inRange()` to create a mask based on a predefined color range, filtering out color regions that meet the criteria. Morphological operations are then performed: erosion to eliminate small noise points, followed by dilation to restore shape loss caused by erosion. Finally, extract and draw contours using `cv2.findContours()` to find contours on the binary mask in the image, and `cv2.drawContours()` to draw the found contours back to the original image or onto another empty image. The results of the above image processing are attached. Figure 24 As shown, a is color space conversion, b is color filtering, and c is drawing the outline.

[0178] Visual recognition and following

[0179] Visual recognition tracking refers to the robot moving in the same direction as a color block that is on the same plane as the robot's chassis, thus achieving target tracking. Specifically, when visual recognition tracking mode is enabled, the robot begins to identify the presence of color blocks. When a color block is detected, the coordinates of its center point are obtained. The robot uses the center point of the frame as its target coordinates. That is, if the coordinates of the color block's center point are not equal to or outside the target coordinate range, the next step is to control the robot's movement. How this movement is implemented will be explained in detail below. The color block recognition implemented in the previous section is insufficient; it is necessary to further obtain the pixel area and center point coordinates of the color block. Obtaining the area of ​​the color block is to filter out objects that are too small or too large from being identified as target objects. By setting an area threshold, only targets within the threshold are ultimately identified. The `minAreaRect()` function is used to find the smallest bounding rectangle of the outline in the image, and then the coordinates of the rectangle's center point are obtained. These coordinates represent the center point coordinates of the color block. A specific image is attached. Figure 25 As shown, the green dot is the block center point, the red dot is the frame center point, and the upper left corner represents the coordinates of the two points.

[0180] After obtaining the coordinates of the color block's center point and the frame's center point, the orientation and straight-line distance between the color block's center point and the frame's center point can be calculated. (See attached image.) Figure 25 The center point of the color block is located to the upper right of the frame center point. The robot is then controlled to spin to the right in place until the line connecting the color block center point and the frame center point is parallel to the frame's Y-axis. The rotation stops then, and the robot moves forward until the color block center point coincides with the frame center point. The specific implementation is as follows.

[0181] The image captured by the camera is divided into seven parts, as shown in the attached figure. Figure 26 As shown, these represent the top left, bottom left, top right, bottom right, Y-axis area, X-axis area, and center area. The region where a color block is located is determined by the coordinates of its center point, and different operations are performed based on the different regions.

[0182] When controlling the robot to follow, only two movements need to be controlled: left and right rotation and forward and backward movement. In the following text, R, L, F, and B will represent right rotation, left rotation, forward movement, and backward movement, respectively. When the color block is in the upper left area, both R and F movements are performed. When the center point of the color block is in the Y-axis area, the R movement stops, and only the F movement is performed. All movements stop only when the center point of the color block is in the center area. A detailed process is attached. Figure 27 As shown. Where R, L, F and B represent right rotation, left rotation, forward and backward rotation, R(1) is right rotation, R(0) is stopping right rotation, and so on.

[0183] Visual recognition and grasping

[0184] Visual recognition and grasping is also based on color block recognition. Its basic process is the same as the visual recognition follower in the previous section. It also determines the position by the coordinates of the center point of the color block, then adjusts the position of the robot so that the center point of the color block is located in the center area of ​​the image, and finally controls the vertical lift, telescopic arm and gripper of the body to grasp.

[0185] Since the robot's movement is controlled to keep the center of the color block in a fixed position after each color block recognition, the position of the color block and the robot is the same every time a grasping action is performed. Therefore, the movement of the body parts (vertical elevator, telescopic arm, gripper) is fixed. This makes the control of the body parts simpler, as the grasping parameters only need to be set once.

[0186] The distance between the center point of the image (O1) and the center point of the robot chassis is S1. When the body is fully reset, the telescopic arm and gripper are in a horizontal state. The distance between the end effector (Q) and the center point of the robot chassis (O) is S2, and the height above the ground is H. See attached figure for details. Figure 28 As shown:

[0187] Depend on

[0188]

[0189] The tilt angle of the telescopic boom is:

[0190]

[0191] Depend on

[0192]

[0193] The telescopic arm's extension length is:

[0194] S = S3 - S2

[0195] because

[0196] θ1=θ

[0197] Therefore, the angle of the gripper's pitch is θ.

[0198] System testing

[0199] Module testing

[0200] Mobile chassis test

[0201] The mobile chassis test primarily assesses the chassis's basic functions, such as straight-line driving, turning, rotating, and stopping, ensuring accurate command response and smooth operation. Specific test details are shown in Table 1-7.

[0202] Table 1-7 Test Results of Mobile Chassis

[0203]

[0204]

[0205] Vertical lift test

[0206] The vertical lift is a core component of the telescopic robot system. This test primarily focuses on evaluating its key performance indicators, including the control accuracy of lifting speed, positional accuracy during lifting, stability when carrying heavy loads, precision of pitch movements, and accuracy of rotational motion. Specific tests aim to verify whether the vertical lift can accurately and smoothly complete various actions when executing corresponding commands, ensuring it meets design standards and practical application requirements. Detailed test results are shown in Tables 1-8.

[0207] Table 1-8 Test Results of Vertical Lifts

[0208]

[0209] Telescopic arm test

[0210] The telescopic arm is the most important component of a telescopic robot system, and its control precision directly affects the robot's accuracy when grasping objects. The telescopic arm employs a unique synchronous rope-driven control method, which presents some challenges in ensuring accuracy; however, some errors can still be compensated for through software. The telescopic arm test mainly examines its telescopic accuracy, the pushing and pulling force during telescopic extension, and its load capacity when fully extended. Detailed test results are shown in Table 1-9.

[0211] Table 1-9 Test Results of Telescopic Boom

[0212]

[0213] gripper test

[0214] The gripper test mainly examines whether each joint of the gripper can rotate accurately and the gripping force of the gripper. Detailed test results are shown in Table 1-10:

[0215] Table 1-10 Gripper Test Results

[0216]

[0217] Functional testing

[0218] Keyboard control test

[0219] Keyboard control is the most basic control method. Detailed test results are shown in Table 1-11:

[0220] Table 1-11 Keyboard Control Test Results

[0221]

[0222] Mobile phone control test

[0223] Mobile phone control is a remote wireless control method via the network. Detailed test results are shown in Table 1-12:

[0224] Table 1-12 Mobile Phone Control Test Results

[0225]

[0226] Visual recognition control test

[0227] Visual recognition control includes color block recognition and following, and color block recognition and grasping. Detailed test results are shown in Table 7-7:

[0228] Table 7-7 Test Results of Visual Recognition Control

[0229]

[0230] This invention overcomes the cost and technological barriers of traditional complex robotic arm structures by using a telescopic arm and lifting rod to replace the complex multi-joint robotic arm, reducing development difficulty and manufacturing costs. In the system, the mobile chassis achieves flexible movement through two-wheel differential drive, the vertical lifting platform controls lifting and pitching attitudes using a screw lifting mechanism and stepper motor, the telescopic arm adopts a four-segment sleeve structure, and uses synchronous rope drive technology to ensure the accuracy of extension and retraction, while the gripper achieves precise grasping action through servo motors.

[0231] In terms of hardware design, the system selected motors and drive solutions with moderate performance and low cost. For example, a 520 encoder DC geared motor is used for the mobile chassis, a 42 stepper motor with a closed-loop servo drive module is used for the lifting platform and telescopic arm, and a high-performance serial bus servo motor is used for the gripper. The main controller uses an STM32 chip and integrates various sensors and motor control interfaces through the OpenCTR B20 control board. It uses the ROS2 operating system for programming and communication management of the upper-level control logic.

[0232] In terms of motion control, the system achieves efficient communication between the host computer and the underlying STM32 controller through node design under the ROS2 framework, enabling the robot's motion to be controlled by keyboard and mobile phone, while also realizing visual recognition control.

[0233] In the description of the embodiments of the present invention, it should be noted that if terms such as "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, terms such as "first," "second," and "third" are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0234] Furthermore, the use of terms such as "horizontal," "vertical," and "sag" does not imply that the component must be absolutely horizontal or suspended, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0235] In the description of the embodiments of the present invention, "multiple" means at least two.

[0236] In the description of the embodiments of the present invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention according to the specific circumstances.

[0237] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A ROS-based telescopic structure robot system, characterized in that: Includes a mobile chassis (1), a vertical lift (2) mounted on the mobile chassis (1), a telescopic arm (3), a gripper (4), a main controller, and multiple sensors; The mobile chassis (1) includes a wheel mounting base (101), a plurality of support columns (102) fixed above the wheel mounting base (101), an upper platform (103) fixed to the upper end of the support columns (102), two drive wheels (104), and two passive wheels (105). The two drive wheels (104) are symmetrically arranged at both ends of the wheel mounting base (101) and are driven independently by their respective connected motors. The two passive wheels (105) are symmetrically arranged at the other two ends of the wheel mounting base (101). The vertical lift (2) includes a 360-degree rotating base (201), two sets of screw lifting mechanisms (202), and two sets of connecting modules. The 360-degree rotating base (201) is supported by a hollow rotary bearing and equipped with a stepper motor and gear linkage to achieve free rotation in all directions. The screw lifting mechanism (202) is assembled from 2020 aluminum profile, 8mm screw, support seat and slider. The telescopic arm (3) is connected to two sliders through two sets of connecting modules. The function of the connecting module is to connect the slider and the telescopic arm (3). When the two sliders move synchronously, the overall lifting state of the telescopic arm (3) is controlled. When the two sliders move with a height difference, the overall pitch state of the telescopic arm (3) is controlled. The telescopic robot system is designed based on ROS2 and includes a vision recognition and control module. The telescopic arm (3) is constructed from four square aluminum alloy sleeves, each 180mm long. The outer diameters of the four sleeves decrease sequentially, being 45mm, 40mm, 35mm, and 30mm respectively, and the thickness is uniformly 1.5mm. A small gap of 1mm is maintained between adjacent sleeves. An outer support seat and an inner support seat are installed on both sides of the sleeve. Each outer support seat and inner support seat is equipped with eight miniature pulleys with a diameter of 5mm.

2. The ROS-based telescopic robot system according to claim 1, characterized in that: The gripper (4) includes four servo motors, a connecting base and a mechanical gripper. The mechanical gripper is a conventional product on the market, with a maximum opening of 92mm and a maximum gripping weight of 500g. It achieves four degrees of freedom through four servo motors, namely horizontal rotation, vertical pitch, spin and gripper tension.

3. The ROS-based telescopic robot system according to claim 1, characterized in that: The main controller is an STM32, and the main controller uses the OpenCTR B20 control board from Tucker Innovations.

Citation Information

Patent Citations

  • High-working efficiency industrial robot

    CN106112959A

  • AUPM570694A0