Intelligent robot automatic mold grabbing device for building components

Abstract

The invention discloses a building component intelligent robot automatic mold grabbing device which comprises a building component robot, the building component robot comprises hardware equipment and a software system, and the hardware equipment comprises an adjusting assembly, a lifting assembly, a grabbing assembly and a man-machine interaction assembly; the software system comprises a processing template, an execution template, a voice control template and a gesture recognition template. According to the invention, operation instructions are issued through the operation buttons and the instruction input area on the control panel, instructions can also be issued to the building component robot through the voice control template and the gesture recognition template, and efficient and convenient man-machine interaction is realized.

Description

Technical Field

[0001] The invention relates to a grasping device, in particular to an automatic grasping mold device of an intelligent robot for building components, belonging to the technical field of building construction robots. Background Art

[0002] At present, the product size and length of precast concrete piles are generally more than 15 meters, and their own mass is relatively large, generally reaching more than 6 tons. The equipment currently used to lift products or steel molds generally uses bridge cranes.

[0003] After searching, Chinese patent number CN209940296U discloses an intelligent robot automatic mold grabbing device for building components, including: walking module, lifting module, grabbing module and control module, which can automatically grab the mold on the production line, realize intelligent automatic control of the grabbing action on the building component production line, reduce the inertia during the grabbing process, reduce the mechanical wear between equipment, and improve production efficiency. However, although the intelligent robot can automatically complete most of the grabbing work, in some complex construction scenes, the participation and assistance of human workers are still required. Since the robot has its own set of operating procedures, it needs to cooperate with the machine manually during the human participation process. If the machine cannot cooperate with the human in time, it is easy to collide with and injure humans. Summary of the invention

[0004] The purpose of the present invention is to provide a building component intelligent robot automatic formwork grabbing device in order to solve the above problems.

[0005] The present invention achieves the above-mentioned object through the following technical scheme: an automatic mold grabbing device of a building component intelligent robot, comprising a building component robot, the building component robot comprising hardware equipment and a software system, the hardware equipment comprising an adjustment component, a lifting component, a grabbing component and a human-computer interaction component; The software system includes a processing template, an execution template, a voice control template and a gesture recognition template; The adjustment assembly includes a driving motor, a lead screw and a threaded block, and is used to adjust the position of the building component robot to improve the flexibility of the building component robot. Side panels are provided on both sides of the bottom end of the building component robot, a driving motor is provided at one end of the side panel, the lead screw is connected to the execution end of the driving motor, and a threaded block is provided on the lead screw; The lifting assembly includes a telescopic cylinder and a lifting rod, which are used to adjust the height of the building component robot. The lifting rod is arranged at the execution end of the telescopic cylinder; The grabbing assembly comprises a first mechanical arm and a second mechanical arm, a joint shaft is provided at the connection between the first mechanical arm and the second mechanical arm, a bottom plate is provided at the bottom end of the second mechanical arm, and a plurality of grabbing clamps are equidistantly provided on both sides of the bottom end of the bottom plate; The human-computer interaction component includes a control panel, a human image capturing camera and a microphone, which are used to trigger a voice control template and a gesture recognition template through gestures and voice to achieve convenient human-computer interaction. The building component robot is equipped with a control panel, human image capturing cameras are provided on both sides of the telescopic cylinder, and a microphone is provided on the control panel.

[0006] A conveyor belt is provided at the bottom end of the building component robot, and the side panels are arranged on both sides of the conveyor belt.

[0007] Preferably, a conveyor belt is provided at the bottom end of the building component robot, and the side panels are arranged on both sides of the conveyor belt.

[0008] Preferably, an adjustment groove is provided at the top end of the side plate, the lead screw is arranged inside the adjustment groove, and the threaded block is connected to the lead screw nut.

[0009] Preferably, the non-executing end of the telescopic cylinder is arranged at the top end of the side plate, and a connecting rod is connected between the non-executing end of the telescopic cylinder and the threaded block.

[0010] Preferably, a cross bar is provided between the top ends of the two lifting rods, the grabbing components are provided in two, the executing motor is provided at both ends of the cross bar, the executing end of the executing motor is located inside the cross bar and is connected to a rotating drum, and the rotating drum is movably connected to the cross bar.

[0011] Preferably, a movable shaft is connected between the rotating drum and the first mechanical arm, and a rotating shaft is connected between an end of the second mechanical arm away from the joint shaft and the base plate.

[0012] Preferably, a mounting block is provided at the top of the gripping clamp, a mounting hole is provided on the mounting block, a plurality of docking holes are equidistantly provided on both sides of the base plate, mounting screws are provided in the docking holes, and an electrode block is provided at the top of the mounting block.

[0013] Preferably, the docking hole passes through the outer wall of the base plate, the size of the docking hole is consistent with the size of the mounting hole, and the length of the mounting screw is consistent with the sum of the hole distances between the docking hole and the mounting hole.

[0014] Preferably, visual sensors are provided at both the front and rear ends of the bottom plate and in the middle.

[0015] The present invention has the following beneficial effects: 1. Add a human-machine interaction component to the building component robot, and issue operation instructions through the operation buttons and command input area on the control panel. Different instructions can be issued to the building component robot according to different batches of building components, realizing efficient and convenient human-machine interaction; 2. The voice control template uses hidden Markov model (HMM), deep neural network-hidden Markov model (DNN-HMM) hybrid model, and end-to-end speech recognition model (such as Transformer-based model) to recognize speech, and uses natural language processing (NLP) algorithm for semantic understanding to improve the position and accuracy of robot grasping components; 3. The gesture recognition template is based on the template matching algorithm, neural network algorithm, and hidden Markov model algorithm to implement gesture actions to issue instructions to the building component robot; 4. The position of the building component robot is adjusted forward and backward through the adjustment component, and the height of the top crossbar of the building component robot is realized according to the lifting component, so as to facilitate the grasping and adjustment of building components of different sizes with the grasping component. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a three-dimensional diagram of the overall structure of an intelligent robot automatic formwork grabbing device for building components proposed by the present invention; Figure 2 This is a structural stereogram of an intelligent robot automatic formwork grabbing device for building components proposed by the present invention; Figure 3 A three-dimensional diagram of the structure of a grabbing assembly of an automatic mold grabbing device of an intelligent robot for building components proposed by the present invention; Figure 4 This is a schematic diagram of the structure of a grabbing component of an automatic mold grabbing device of an intelligent robot for building components proposed by the present invention; Figure 5 This is a structural stereogram of an intelligent robot automatic formwork grabbing device for building components proposed by the present invention; Figure 6 This is a structural stereogram of an intelligent robot automatic formwork grabbing device for building components proposed by the present invention; Figure 7 This is a logical system diagram of an intelligent robot automatic formwork grabbing device for building components proposed by the present invention.

[0017] In the figure: 1. building component robot; 101. visual sensor; 2. conveyor belt; 3. adjustment component; 301. side panel; 302. adjustment slot; 303. drive motor; 304. lead screw; 305. threaded block; 306. connecting rod; 4. lifting component; 401. telescopic cylinder; 402. lifting rod; 403. cross bar; 5. grasping component; 501. execution motor; 502. rotating drum; 503. movable axis; 504. first mechanical arm; 505. joint axis; 506. second mechanical arm; 507. rotating axis; 508. bottom plate; 509. grasping clamp; 6. human-computer interaction component; 601. control panel; 602. human image capture camera; 603. microphone; 7. electrode block; 701. docking hole; 702. mounting hole; 703. mounting screw; 704. mounting block. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0019] Embodiment 1: Reference Figure 1 as well as Figure 7 , a building component intelligent robot automatic mold grabbing device, including a building component robot 1, the building component robot 1 includes hardware equipment and a software system, the hardware equipment includes an adjustment component 3, a lifting component 4, a grabbing component 5 and a human-computer interaction component 6; The software system includes a processing template, an execution template, a voice control template, and a gesture recognition template; The human-computer interaction component 6 includes a control panel 601, a human image capturing camera 602 and a microphone 603, which are used to trigger a voice control template and a gesture recognition template through gestures and voice to achieve convenient human-computer interaction. The building component robot 1 is equipped with a control panel 601, and human image capturing cameras 602 are provided on both sides of the telescopic cylinder 401. A microphone 603 is provided on the control panel 601.

[0020] In this embodiment, it should be noted that when the building component robot 1 is used to grab the building component, operation instructions can be issued through the operation buttons and instruction input area on the control panel 601.

[0021] The control panel 601 can take into account the operator's usage habits and operation procedures to achieve efficient and convenient human-computer interaction. The interface layout should be concise and clear, the functional modules should be reasonably divided, and the operation buttons and command input areas should be easy to operate. At the same time, it is necessary to use appropriate fonts, colors, icons and other elements to improve the readability and aesthetics of the interface.

[0022] When using the building component robot 1, instructions can also be given to the building component robot 1 through a voice control template, and a hidden Markov model (HMM), a deep neural network-hidden Markov model (DNN-HMM) hybrid model, or an end-to-end speech recognition model (such as a Transformer-based model) can be used.

[0023] Hidden Markov Model (HMM) is a statistical model that is widely used in speech recognition. It is based on the temporal characteristics of speech signals and regards speech signals as observation sequences generated by a series of hidden states (such as phonemes). During the training phase, a large amount of speech data is used to estimate the parameters of the model, including state transition probabilities, observation probabilities, etc. For example, for the voice command "grab building components", HMM will analyze the acoustic features of the speech signal and identify the corresponding text sequence through model matching.

[0024] Deep Neural Network - Hidden Markov Model (DNN - HMM) hybrid model, which combines the powerful feature learning ability of deep neural network (DNN) and the time series modeling ability of HMM. DNN is used to extract high-level acoustic features of speech signals and convert the original speech spectrum features into more discriminative feature representations. Then, HMM uses these features for time series modeling and speech recognition. For example, DNN - HMM can better handle the speech differences between different speakers and improve the accuracy of speech recognition.

[0025] End-to-end speech recognition models (such as Transformer-based models) directly convert speech signals into text sequences without the need for manual feature extraction and alignment steps in the middle. Transformer-based models have advantages in processing long sequence data (such as speech). They can capture the relationship between different time steps in speech signals through the self-attention mechanism. For example, in the voice control of building component intelligent robots, when users issue complex command sequences, end-to-end models can better understand the semantics of the entire command instead of recognizing it fragment by fragment.

[0026] Then, the natural language processing (NLP) algorithm is used for semantic understanding. For example, for the sentence "grab the rectangular building component", after word segmentation, we get "grab / that / rectangular / of / building components", and the part-of-speech tagging may be "verb / pronoun / noun / particle / noun phrase", to determine the semantic role of each component in the sentence, such as agent, patient, tool, etc. In "grab the rectangular building component", "robot (not explicitly mentioned, but inferred from the context)" is the agent, and "building component" is the patient. SRL can help robots understand the semantic relationship between actions and objects, so as to better execute instructions.

[0027] Once the intent of the voice command is determined, it needs to be converted into an action that the robot can perform. For example, if the intent is "grab that rectangular building component", the robot needs to plan the grasping action based on the location of the component (possibly obtained through a visual sensor) and its shape (rectangular). This may involve adjusting parameters such as the angle, position, and strength of the gripper.

[0028] Calculation of action parameters: Calculate specific action parameters based on the detailed information in the voice command and the current state of the robot. For example, if the command says "grab slowly", the robot needs to adjust the grasping speed parameters; if the command says "grab from the left", the robot needs to adjust the horizontal position parameters of the gripper. Mathematical models and the robot's kinematic and dynamic models can be used to calculate these parameters.

[0029] The gesture recognition template is based on the template matching algorithm, the neural network algorithm, and the hidden Markov model algorithm to implement gesture actions to issue instructions to the building component robot.

[0030] Algorithms based on template matching: First, a gesture template library is established, which contains images or feature data of various pre-defined gestures.

[0031] When a new gesture image is obtained, it is compared with the templates in the template library one by one to calculate the similarity. The similarity can be measured by methods such as image pixel difference and feature vector distance.

[0032] The gesture category corresponding to the template with the highest similarity is selected as the recognition result. For example, if the new gesture image has the highest similarity with the template of the "grab" gesture in the template library, it is recognized as a "grab" gesture.

[0033] Neural network based algorithms: Training phase: Collect a large amount of gesture image data and annotate them to indicate the gesture category corresponding to each image. Then, use a neural network model (such as a convolutional neural network CNN) to train this data. The neural network will automatically learn the relationship between the features of the gesture image and the gesture category, and adjust the model's weights and parameters to improve recognition accuracy.

[0034] Recognition stage: Input the new gesture image into the trained neural network model, and the model will output the prediction result of the category of the gesture. The neural network has a strong feature learning ability, can process complex gesture images, and has good adaptability to changes in different lighting, angles, etc.

[0035] Algorithms based on Hidden Markov Models: A gesture is considered as a stochastic process consisting of hidden states and observed states. The hidden states represent the intrinsic state of the gesture or the stage of the action, while the observed states are the gesture features extracted from the image.

[0036] By training a large amount of gesture data, the model parameters are estimated, including state transition probability and observation probability. The state transition probability indicates the possibility of a gesture transitioning from one state to another, and the observation probability indicates the probability of observing a specific gesture feature in a certain state.

[0037] During recognition, the most likely gesture state sequence is calculated based on the input gesture feature sequence using the model parameters to determine the category of the gesture. This algorithm is suitable for dynamic gesture recognition and can capture the time series information of gestures.

[0038] Embodiment 2: Different from the first embodiment, Figure 1-7 , this embodiment also has the following further contents: the adjustment component 3 includes a driving motor 303, a lead screw 304 and a threaded block 305, which is used to adjust the position of the building component robot 1 to improve the flexibility of the building component robot 1. Side plates 301 are provided on both sides of the bottom end of the building component robot 1. A driving motor 303 is provided at one end of the side plate 301. The lead screw 304 is connected to the execution end of the driving motor 303, and a threaded block 305 is provided on the lead screw 304; The lifting assembly 4 includes a telescopic cylinder 401 and a lifting rod 402 , and is used to adjust the height of the building component robot 1 . The lifting rod 402 is disposed at the execution end of the telescopic cylinder 401 .

[0039] A conveyor belt 2 is provided at the bottom end of the building component robot 1, side panels 301 are arranged on both sides of the conveyor belt 2, an adjustment groove 302 is provided at the top end of the side panel 301, a lead screw 304 is arranged inside the adjustment groove 302, a threaded block 305 is connected to the lead screw 304 nut, a non-executing end of the telescopic cylinder 401 is arranged at the top end of the side panel 301, a connecting rod 306 is connected between the non-executing end of the telescopic cylinder 401 and the threaded block 305, and visual sensors 101 are provided at the front and rear ends of the bottom plate 508 and in the middle position.

[0040] In this embodiment, it should be noted that: after the visual sensor 101 of the building component robot 1 collects the image of the building component, the image data of the building component collected by the visual sensor needs to be processed by a template, and image processing and pattern recognition technology are used to extract the characteristics of the component, such as shape, size, color, texture, etc.

[0041] These features can be used to identify different types of building components and determine the position and orientation of the components. For example, the edge contour of the component can be extracted through the edge detection algorithm, and the position and orientation of the component can be determined through the feature point matching algorithm.

[0042] After processing the information of the building component, the processing template sends its electrical signal to the execution template, and the execution template sends it to the adjustment component 3, the lifting component 4 and the grasping component 5. According to the shape and size of the building component, the building component robot 1 is driven by the driving motor 303 and the screw 304 to adjust the position, and the telescopic cylinder 401 drives the height of the building component robot 1 to adjust the height, and cooperates with the grasping clamp 509 to grasp the building component.

[0043] Implementation three: Reference Figure 1-7 Compared with the first and second embodiments, in this embodiment: the grasping assembly 5 includes a first robotic arm 504 and a second robotic arm 506, a joint shaft 505 is provided at the connection between the first robotic arm 504 and the second robotic arm 506, a bottom plate 508 is provided at the bottom end of the second robotic arm 506, and a plurality of grasping clamps 509 are equidistantly provided on both sides of the bottom end of the bottom plate 508.

[0044] A cross bar 403 is provided between the top ends of the two lifting rods 402, and the grabbing assembly 5 is arranged in two. The execution motor 501 is arranged at both ends of the cross bar 403. The execution end of the execution motor 501 is located inside the cross bar 403 and is connected to a rotating drum 502. The rotating drum 502 is movably connected to the cross bar 403. A movable shaft 503 is connected between the rotating drum 502 and the first robotic arm 504. A rotating shaft 507 is connected between the end of the second robotic arm 506 away from the joint shaft 505 and the bottom plate 508.

[0045] A mounting block 704 is provided at the top of the gripper 509, a mounting hole 702 is provided on the mounting block 704, a plurality of docking holes 701 are equidistantly provided on both sides of the base plate 508, the docking holes 701 are equipped with mounting screws 703, an electrode block 7 is provided at the top of the mounting block 704, the docking hole 701 passes through the outer wall of the base plate 508, the size of the docking hole 701 is consistent with the size of the mounting hole 702, and the length of the mounting screw 703 is consistent with the sum of the hole distances between the docking hole 701 and the mounting hole 702.

[0046] In this embodiment, it should be noted that the grabbing clip 509 at the bottom end of the base plate 508 can be disassembled and installed by connecting the mounting screw 703 with the mounting hole 702. During use, the number of grabbing clips 509 at the bottom end of the base plate 508 can be adjusted according to actual needs. At the same time, after a period of use, the aged or faulty grabbing clip 509 can be replaced.

Claims

1. An intelligent robot automatic formwork grabbing device for building components, comprising a building component robot (1), characterized in that: The building component robot (1) comprises hardware equipment and a software system, wherein the hardware equipment comprises an adjustment component (3), a lifting component (4), a grasping component (5) and a human-machine interaction component (6); The software system includes a processing template, an execution template, a voice control template and a gesture recognition template; The adjustment component (3) comprises a drive motor (303), a lead screw (304) and a threaded block (305), and is used to adjust the position of the building component robot (1) to improve the flexibility of the building component robot (1); side panels (301) are provided on both sides of the bottom end of the building component robot (1); a drive motor (303) is provided at one end of the side panel (301); the lead screw (304) is connected to the execution end of the drive motor (303); and a threaded block (305) is provided on the lead screw (304); The lifting assembly (4) comprises a telescopic cylinder (401) and a lifting rod (402), and is used to adjust the height of the building component robot (1); the lifting rod (402) is arranged at the execution end of the telescopic cylinder (401); The grasping assembly (5) comprises a first mechanical arm (504) and a second mechanical arm (506); a joint shaft (505) is provided at the connection between the first mechanical arm (504) and the second mechanical arm (506); a bottom plate (508) is provided at the bottom end of the second mechanical arm (506); and a plurality of grasping clamps (509) are equidistantly provided on both sides of the bottom end of the bottom plate (508); The human-machine interaction component (6) comprises a control panel (601), a human image capturing camera (602) and a microphone (603), and is used to trigger a voice control template and a gesture recognition template through gestures and voice, thereby realizing convenient human-machine interaction. The building component robot (1) is equipped with a control panel (601), human image capturing cameras (602) are provided on both sides of the telescopic cylinder (401), and a microphone (603) is provided on the control panel (601).

2. The intelligent robot automatic formwork grabbing device for building components according to claim 1, characterized in that: A conveyor belt (2) is provided at the bottom end of the building component robot (1), and the side panels (301) are arranged on both sides of the conveyor belt (2).

3. The intelligent robot automatic formwork grabbing device for building components according to claim 2, characterized in that: An adjustment groove (302) is provided at the top end of the side plate (301), the lead screw (304) is arranged inside the adjustment groove (302), and the threaded block (305) is connected to the lead screw (304) with a nut.

4. The intelligent robot automatic formwork grabbing device for building components according to claim 3, characterized in that: The non-executing end of the telescopic cylinder (401) is arranged at the top end of the side plate (301), and a connecting rod (306) is provided between the non-executing end of the telescopic cylinder (401) and the threaded block (305).

5. The intelligent robot automatic formwork grabbing device for building components according to claim 1, characterized in that: A cross bar (403) is provided between the top ends of the two lifting rods (402), the grabbing assembly (5) is provided in two, the execution motor (501) is provided at both ends of the cross bar (403), the execution end of the execution motor (501) is located inside the cross bar (403) and is connected to a rotating drum (502), and the rotating drum (502) is movably connected to the cross bar (403).

6. The intelligent robot automatic formwork grabbing device for building components according to claim 5, characterized in that: A movable shaft (503) is connected between the rotating drum (502) and the first mechanical arm (504), and a rotating shaft (507) is connected between an end of the second mechanical arm (506) away from the joint shaft (505) and the bottom plate (508).

7. The intelligent robot automatic formwork grabbing device for building components according to claim 6, characterized in that: A mounting block (704) is provided at the top of the gripper (509), a mounting hole (702) is provided on the mounting block (704), a plurality of docking holes (701) are equidistantly provided on both sides of the base plate (508), mounting screws (703) are provided in conjunction with the docking holes (701), and an electrode block (7) is provided at the top of the mounting block (704).

8. The intelligent robot automatic formwork grabbing device for building components according to claim 7, characterized in that: The docking hole (701) passes through the outer wall of the bottom plate (508), the size of the docking hole (701) is consistent with the size of the mounting hole (702), and the length of the mounting screw (703) is consistent with the sum of the hole distances between the docking hole (701) and the mounting hole (702).

9. The intelligent robot automatic formwork grabbing device for building components according to claim 8, characterized in that: The bottom plate (508) is provided with visual sensors (101) at both the front and rear ends and in the middle.

Images