Concrete spreader based on depth visual control

Abstract

The invention discloses a concrete spreader based on depth visual control. The concrete spreader comprises a supporting frame and a supporting pipe. The main arm is rotatably connected to the upper part of the supporting pipe; the main arm rotating mechanism is fixed at the upper part of the supporting pipe; the material distribution pipe penetrates in from the bottom of the supporting pipe and penetrates out from the upper portion of the supporting pipe, the penetrating-out portion of the material distribution pipe is divided into a plurality of sections, and every two adjacent sections of the material distribution pipe are rotatably connected; the material distribution pipe rotating mechanism is fixed between two adjacent sections of material distribution pipes and is used for driving the two adjacent sections of material distribution pipes to rotate relatively; and the visual control module is electrically connected to the main arm driving mechanism and the plurality of material distribution pipe rotating mechanisms. Through the visual control module, a constructor makes corresponding gestures, the sensing module captures the gestures, the execution module executes corresponding instructions according to the instruction information, the concrete spreader can be controlled in the mode, operation is easy and efficient, the constructor does not need to make contact with a distributing pipe with hands, and the operation process is safe.

Description

Technical Field

[0001] The present invention belongs to the technical field of pile foundation engineering, and in particular relates to a concrete placing boom based on depth vision control. Background Art

[0002] During the concrete pouring phase, concrete placing booms are primarily used to deliver concrete to designated pouring locations. After the booms are set up, on-site operators must autonomously move the concrete placing hoses to the desired pouring locations. Currently, there are two ways to control the movement of the concrete placing hoses: manual movement of the discharging port by the operator, and remote control.

[0003] Traditional manual movement requires the operator to manually drag the distribution pipe to move it, which is time-consuming and labor-intensive, increasing the worker's workload. Furthermore, the operator needs to constantly hold the distribution pipe, increasing the workload. Furthermore, manual manual operation requires workers to maintain close contact with the distribution pipe for extended periods of time, posing safety risks in confined spaces or near edges. For example, during the pouring of mid- to high-rise concrete wall columns, the operator needs to stand on the edge formwork to manually hold the distribution pipe.

[0004] Although controlling the movement operation through remote control can reduce the workload of workers to a certain extent, the operation of professional remote control is complicated. It is difficult to master the movement operation of the end fabric mouth without long-term training. The operation threshold is high, and the operation of unskilled personnel can easily lead to a decrease in efficiency. Summary of the Invention

[0005] The present invention provides a concrete placing boom based on deep vision control. By providing a visual control module, a worker can make corresponding gestures with his hands to control the concrete placing boom in a simple and efficient manner.

[0006] The cam is connected to the top of the support tube to rotate with the top of the support tube, and the cam is connected to the top of the support tube to rotate with the top of the support tube.

[0007] A further improvement of the present invention is that the protruding part of the fabric tube is divided into three sections, namely, a first section which is arranged above the main arm, extends along the direction of the main arm to the end of one side of the main arm and bends vertically downward; a second section which bends from the end of the first section to the bottom of the main arm and extends below the main arm in the opposite direction to the first section; and a third section which bends vertically downward from the end of the second section and extends downward. The second section can rotate with the end of the first section bent vertically downward as the axis, and the third section can rotate with the second section as the axis.

[0008] A further improvement of the present invention is that the visual control module includes: a perception module for capturing gesture information; a data processing module for analyzing the gestures captured by the perception module to obtain instruction information; and an execution module for controlling the main arm drive structure and the multiple fabric tube rotation mechanisms through the instruction information.

[0009] A further improvement of the present invention is that the perception module includes: a camera for capturing gesture images.

[0010] A further improvement of the present invention is that the perception module also includes an infrared fill light for filling light for the camera and providing a night vision function.

[0011] A further improvement of the present invention is that the data processing module includes: a gesture recognition algorithm for recognizing captured gestures to obtain gesture information; and a custom gesture library for matching the obtained gesture information with custom gesture information to obtain instruction information.

[0012] A further improvement of the present invention is that the execution module includes a reverse calculation module, which is equipped with an inverse kinematics algorithm for calculating the specific angles of rotation of each joint that drives the cloth opening movement according to instruction information.

[0013] A further improvement of the present invention is that it also includes a start-stop switch that can control the start and stop of the concrete placing boom, and the visual control module includes a feedback correction module, which is electrically connected to the start-stop switch. The feedback correction module includes: a force sensor for collecting the reaction force of concrete flow; and a laser radar for scanning obstacles on the construction surface.

[0014] A further improvement of the present invention is that a counterweight is fixed to the other end of the main arm, opposite to the distribution pipe, for balancing the weight of the distribution pipe. The present invention provides a concrete placing boom based on deep vision control. By providing a visual control module, a worker makes corresponding gestures, which are captured by a perception module. A data processing module converts the gesture information into command information, and an execution module executes the corresponding command based on the command information. This method allows for simple and efficient control of the concrete placing boom. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a structural diagram of a concrete placing boom according to the present invention;

[0016] Figure 2 This is a hardware architecture diagram of the visual control module of the present invention;

[0017] Figure 3 is a system logic diagram of the visual control system of the present invention;

[0018] In the figure: 1. Support tube; 101. Feed port; 102. Inner section of tube; 103. First section; 104. Second section; 105. Third section; 106. Discharge port; 2. Main arm; 3. Counterweight; 4. Main arm rotation mechanism; 5. First distribution tube rotation mechanism; 6. Second distribution tube rotation mechanism; 7. Visual control module; 8. Support frame. DETAILED DESCRIPTION

[0019] like Figure 1 As shown, the present invention provides a concrete placing boom based on depth vision control, comprising: a support frame 8, a support tube 1 arranged in a vertical direction is fixed on the top of the support frame 8; a main arm 2, the main arm 2 is arranged in a horizontal direction and is rotatably connected to the upper part of the support tube 1 with the support tube 1 as an axis; a main arm rotating mechanism 4 fixed to the upper part of the support tube 1 for driving the main arm 2 to rotate; a distribution pipe, the distribution pipe passes through the bottom of the support tube 1, passes through the upper part of the support tube 1 and extends along one side of the main arm 2 to form a discharge port, the passing part of the distribution pipe is divided into multiple sections, and adjacent two sections of distribution pipes are rotatably connected; a distribution pipe rotating mechanism fixed between two adjacent sections of distribution pipes for driving the two adjacent sections of distribution pipes to rotate relative to each other; a visual control module 7 for controlling the concrete placing boom by recognizing hand commands, and the visual control module 7 is electrically connected to the main arm driving mechanism and multiple distribution pipe rotating mechanisms.

[0020] like Figure 1 As shown, in this embodiment, the portion of the fabric pipe located inside the support pipe 1 is called the inner section 102, and the protruding portion of the fabric pipe is divided into three sections, namely, a first section 103 provided above the main arm 2, extending along the direction of the main arm 2 to the end of one side of the main arm 2 and bending vertically downward; a second section 104 bending from the end of the first section 103 to the bottom of the main arm 2 and extending below the main arm 2 in the opposite direction to the first section 103; a third section 105 bending vertically downward from the end of the second section 104 and extending downward, the second section 104 can be rotated about the end of the first section 103 bent vertically downward as an axis, and the third section 105 can be rotated about the second section as an axis, wherein the first fabric pipe rotating mechanism 5 is located between the first section 103 and the second section 104, and the second fabric pipe rotating mechanism 6 is located between the second section 104 and the third section 105.

[0021] like Figures 2 and 3 As shown, the visual control module 7 includes: a perception module for capturing gesture information; a data processing module for analyzing the gestures captured by the perception module to obtain instruction information; and an execution module for controlling the main arm rotation mechanism 4 and multiple fabric pipe rotation mechanisms through the instruction information.

[0022] like Figure 2 As shown, the perception module includes: a camera for capturing gesture images.

[0023] like Figure 2 As shown, the perception module also includes an infrared fill light for the camera and providing night vision function.

[0024] Preferably, if Figures 2 and 3 As shown, in this embodiment, the camera uses a TOF camera to collect the three-dimensional depth information of the operator's hand in real time. At the same time, the TOF camera has a certain ability to resist dust interference, which is particularly suitable for the dust environment of the construction site. Compared with the traditional RGB camera, the recognition rate in the dust environment is improved by 40%.

[0025] Preferably, if Figures 2 and 3 As shown, in this embodiment, the infrared fill light uses an infrared LED array, which can dynamically adjust the lighting intensity to adapt to the low-light environment at night.

[0026] like Figures 2 and 3 As shown, in this embodiment, the data processing module is an embedded AI computing unit, including: a gesture recognition algorithm for recognizing captured gestures to obtain gesture information; and a custom gesture library for matching the obtained gesture information with custom gesture information to obtain instruction information.

[0027] In this example, the gesture recognition algorithm is a multi-target detection gesture recognition algorithm developed based on the YOLO model framework. Through a custom data set and using annotation tools such as Labelme, the training set gestures are annotated with boxes to train a gesture library that meets the requirements of field use. The algorithm engine combines hand skeleton points (MediaPipe) with motion trajectory analysis to reduce the misjudgment rate and add dynamic trajectory correction capabilities. Based on the recognized gestures, combined with the custom gesture library, the operator's intention is analyzed and different actions are executed based on the recognized intention. The customizable gesture commands in the gesture library can expand gesture recognition and its corresponding functions, making it scalable.

[0028] Preferably, this embodiment also includes a start / stop switch for controlling the start and stop of the concrete placing boom. The visual control module 7 also includes a feedback correction module electrically connected to the start / stop switch. The feedback correction module includes: a force sensor for collecting the reaction force of concrete flow; and a laser radar for scanning for obstacles on the construction surface. The force sensor can detect the reaction force of concrete flow. If the reaction force is abnormal, the concrete placing boom can be shut down via the start / stop switch to prevent overloading or splashing of the concrete. The laser radar can also be used to map obstacles on the construction surface (such as rebar and formwork), generating 3D point cloud data to assist in path avoidance. When an obstacle is encountered, the concrete placing boom can be shut down via the start / stop switch to prevent damage to the concrete placing boom due to collisions with obstacles.

[0029] Preferably, if Figure 1 As shown, a counterweight 3 is fixed to the other end of the main arm 2, opposite the distribution pipe, to balance the weight of the distribution pipe. In this embodiment, the distribution pipe is primarily located on the right side of the main arm 2, so a counterweight is installed on the left end of the main arm 2 to prevent the equipment from tipping over due to unbalanced gravity at both ends.

[0030] Preferably, in this embodiment, the main arm rotating mechanism 3 and the cloth rotating mechanism are both driven by servo motors, and the servo motors can receive signals sent by the data processing module and execute corresponding instructions.

[0031] Preferably, in this embodiment, if Figure 3 As shown, the execution module includes a reverse calculation module, which is equipped with an inverse kinematics algorithm and can calculate the specific rotation angles of each joint that drives the cloth mouth movement according to the instruction information.

[0032] Preferably, if Figure 3As shown, in this example, the data processing module also divides gesture information into two categories. The first category is start-stop instructions, which are used to control the start-stop switch to directly start and stop the concrete placing boom. The second category is movement instructions. When the data processing module recognizes that the worker's intention is a movement instruction, it will issue a motion instruction according to the gesture following algorithm. The gesture following algorithm is developed based on MediaPipe, captures the camera video stream through OpenCV, and identifies the center point of the operator's palm as the target joint point. The TOF camera captures the coordinates (X, Y, Z) of the hand in three-dimensional space in real time, where the Z axis is the front-to-back direction (perpendicular to the camera's field of view), completing three-dimensional depth perception and positioning. At the same time, if the on-site environment is low-light or dusty, a millimeter-wave radar can be installed according to the actual on-site environment to assist in detecting the hand movement speed and improve the accuracy of dynamic tracking. After analyzing the gestures using the gesture following algorithm to obtain the target placement opening position, a kinematic chain is constructed based on the DH parameters (Denavit-Hartenberg) of the concrete placing boom arm, and the target angles of each joint are obtained through inverse kinematics. The joint angle commands are then sent to the main arm rotation mechanism or the material placing rotation mechanism, causing the servo motor to move according to a trapezoidal speed curve. Simultaneously, a force sensor detects the concrete recoil force in real time and dynamically adjusts the servo motor's torque.

[0033] The present invention includes the following beneficial effects:

[0034] (1) Improved the continuity and stability of construction and ensured high standards of construction quality;

[0035] (2) Improved on-site construction efficiency, reduced customer costs, and improved overall economic benefits;

[0036] (3) Improved the construction environment for operators and reduced the workload of traditional construction;

[0037] (4) It has improved the level of intelligent construction in the construction industry and increased the degree of automation on site.

[0038] In summary, the present invention is an iterative solution for a concrete placing boom with significant innovation and practicality, which is of great significance for improving construction efficiency and improving the construction environment.

[0039] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent replacements for some of the technical features therein. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A concrete placing boom based on deep vision control, characterized in that: include: A support frame, wherein a support tube arranged in a vertical direction is fixed on the top of the support frame; a main arm, the main arm being arranged in a horizontal direction and being rotatably connected to the upper portion of the support tube with the support tube as an axis; a main arm rotating mechanism fixed to the upper portion of the support tube and used for driving the main arm to rotate; A distribution pipe, which passes through the bottom of the support pipe, passes through the upper part of the support pipe and extends along one side of the main arm to form a discharge port. The passing portion of the distribution pipe is divided into multiple sections, and two adjacent sections of the distribution pipe are rotatably connected; A distribution pipe rotation mechanism is fixed between two adjacent sections of the distribution pipes and is used to drive the two adjacent sections of the distribution pipes to rotate relative to each other; a visual control module controls the concrete placing boom by recognizing hand commands, and the visual control module is electrically connected to the main arm drive mechanism and the multiple distribution pipe rotation mechanisms.

2. The concrete placing boom based on depth vision control according to claim 1, characterized in that: The outlet portion of the fabric pipe is divided into three sections, which are respectively a first section arranged above the main arm, extending along the direction of the main arm to the end of one side of the main arm and bending vertically downward; a second section bent from an end of the first section to below the main arm and extending below the main arm in a direction opposite to the first section; The third section is bent vertically downward from the end of the second section and extends downward. The second section can rotate about the end of the first section bent vertically downward as an axis, and the third section can rotate about the second section as an axis.

3. The concrete placing boom based on depth vision control according to claim 1, characterized in that The visual control module includes: a perception module for capturing gesture information; a data processing module for analyzing the gestures captured by the perception module to obtain instruction information; and an execution module for controlling the main arm rotation mechanism and the multiple fabric tube rotation mechanisms through the instruction information.

4. The concrete placing boom based on depth vision control according to claim 3, characterized in that: The perception module includes: a camera for capturing gesture images.

5. The concrete placing boom based on depth vision control according to claim 4, characterized in that: The sensing module also includes an infrared fill light for filling light for the camera and providing a night vision function.

6. The concrete placing boom based on depth vision control according to claim 3, characterized in that: The data processing module includes: a gesture recognition algorithm for recognizing captured gestures to obtain gesture information; and a custom gesture library for matching the obtained gesture information with custom gesture information to obtain instruction information.

7. The concrete placing boom based on depth vision control according to claim 3, characterized in that: The execution module includes an inverse calculation module equipped with an inverse kinematics algorithm for calculating the specific rotation angles of each joint driving the cloth opening according to instruction information.

8. The concrete placing boom based on depth vision control according to claim 1, characterized in that: It also includes a start-stop switch that can control the start and stop of the concrete placing boom. The visual control module includes a feedback correction module, which is electrically connected to the start-stop switch. The feedback correction module includes: a force sensor for collecting the reaction force of concrete flow; and a laser radar for scanning obstacles on the construction surface.

9. The concrete placing boom based on depth vision control according to claim 1, characterized in that: A counterweight block for balancing the weight of the material distribution pipe is fixed on the other end of the main arm relative to the material distribution pipe.

Images