Self-adaptive manipulator and method

Abstract

The invention relates to the technical field of robots, in particular to a self-adaptive manipulator which comprises a sensing system and a control system. A multi-degree-of-freedom motion joint assembly is arranged at the lower end of the base, a driving device for providing power is installed at the lower end of the top of the base, the joint assembly comprises a lifting sleeve and a first connecting base, a lifting plate is fixedly connected to the lower end of the lifting sleeve, and second connecting bases annularly distributed at equal intervals are fixedly connected to the lower end of the lifting plate; a first rotating joint arm is rotatably connected in the first connecting seat, a third connecting seat is fixedly connected to the inner side of the first rotating joint arm, a second rotating joint arm is rotatably connected between the third connecting seat and the second connecting seat, and the control system is arranged in the base; the shape and hardness of an object can be sensed in real time, the grabbing force and posture are dynamically adjusted, it is ensured that different objects are stably grabbed, the grabbing strategy can be automatically learned and optimized, and high precision and reliability of grabbing are ensured through a high-precision sensor and a real-time feedback control mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of robots, and in particular to an adaptive manipulator and method. Background Art

[0002] Manipulators mainly complete various expected operations through programming. Manipulators can imitate the actions of human hands and arms to grasp, transport objects or operate tools according to fixed programs, and are widely used in multiple fields, which can effectively liberate people from dangerous working environments.

[0003] Manipulators mainly include an execution mechanism, a driving mechanism, a control system and a base. Manipulators are driven by pneumatic, hydraulic or electric means, receive instructions through the control system, and control the movement of components such as the arm, wrist and gripper. The gripper holds an object, and the wrist and arm work together to achieve precise operation and movement. Existing manipulators usually adopt a rigid structure, which is difficult to adapt to objects of different shapes and hardnesses, resulting in unstable grasping or damage to the object. In addition, the control methods of existing manipulators are complex and it is difficult to achieve fast and accurate adaptive grasping.

[0004] Therefore, aiming at the problems that the above-mentioned rigid structure is difficult to adapt to objects of different shapes and hardnesses, the control method is complex and it is difficult to achieve adaptive grasping, an adaptive manipulator and method can be designed. Summary of the Invention

[0005] In order to overcome the problems that a rigid structure is difficult to adapt to objects of different shapes and hardnesses, the control method is complex and it is difficult to achieve adaptive grasping.

[0006] The technical solution of the present invention is: an adaptive manipulator, including a base, a joint assembly, a driving device and flexible fingers; it also includes a sensing system and a control system; a joint assembly with multiple degrees of freedom is arranged at the lower end of the base, and a driving device providing power is installed at the lower end of the top of the base. The joint assembly includes a lifting sleeve and a first connecting seat. A lifting plate is fixedly connected to the lower end of the lifting sleeve, and a second connecting seat distributed annularly and equidistantly is fixedly connected to the lower end of the lifting plate. A first rotating joint arm is rotatably connected in the first connecting seat, a third connecting seat is fixedly connected to the inner side of the first rotating joint arm, a second rotating joint arm is rotatably connected between the third connecting seat and the second connecting seat, and a flexible finger is fixedly connected to the lower end of the first rotating joint arm. The sensing system is distributed at the fingers and the joint assembly, and the control system is arranged in the base;

[0007] The sensing system includes a force sensing unit, a tactile sensing unit and a vision sensing unit;

[0008] The control system includes a central processing unit, a drive module, and a communication module. The central processing unit is used to process the data of the sensing system and adjust the movement of the flexible finger according to a preset algorithm. The drive module drives the joint assembly to move according to the instructions of the central processing unit. The communication module is used to communicate with external devices;

[0009] An adaptive algorithm is also integrated in the control system. The adaptive algorithm includes a shape recognition algorithm, a force control algorithm, and an attitude adjustment algorithm. The shape recognition algorithm identifies the shape and size of an object through the data of the visual sensing unit and the tactile sensing unit. The force control algorithm adjusts the grasping force in real time according to the feedback of the force sensing unit. The attitude adjustment algorithm adjusts the attitude of the flexible finger according to the shape and position of the object.

[0010] Preferably, the surface of the flexible finger is covered with a flexible material, such as silicone or rubber.

[0011] Preferably, the driving device adopts any one of a screw lift mechanism, a pneumatic lift mechanism, an oil cylinder lift mechanism, and a motor plus a rack and pinion.

[0012] Preferably, for the shape recognition of the shape recognition algorithm, machine learning or deep learning algorithms are used to classify the fused features to identify the object shape. In the size analysis of the shape recognition algorithm, the visual data estimates the object size through geometric methods or deep learning, and the tactile data estimates the object size through the contact area and pressure distribution.

[0013] Preferably, the specific steps of the force control algorithm are as follows:

[0014] A. Collect force sensing data at a fixed frequency and measure the grasping force in real time;

[0015] B. Data preprocessing, including filtering and normalization;

[0016] C. Target force setting

[0017] Preset target force: Set the target grasping force according to the object characteristics;

[0018] Dynamic adjustment: Dynamically adjust the target force according to the task requirements;

[0019] D. Force error calculation, calculate the difference between the current grasping force and the target force: e = F t -F c ; where F t is the target force and F c is the current force;

[0020] E. Adjust the grasping force by using any one of the algorithms of PID control, fuzzy control, and adaptive control.

[0021] Preferably, the posture adjustment adopts any one of the inverse kinematics algorithm, force feedback-based control, optimization algorithm, and machine learning method.

[0022] Preferably, the force sensing unit is installed at the flexible finger for real-time monitoring of the grasping force, the tactile sensing unit is distributed on the surface of the flexible finger for perceiving the shape and material of the object, and the vision sensing unit is installed on the base for identifying the position and shape of the object.

[0023] The adaptive manipulator method includes the adaptive manipulator as described above, and the steps are as follows:

[0024] Step 1: Object recognition

[0025] The force sensing unit monitors the grasping force in real time, the tactile sensing unit perceives the shape and material of the object, and the vision sensing unit identifies the position and shape of the object;

[0026] Step 2: Grasping planning

[0027] The central processor plans the movement trajectory and grasping force of the flexible finger according to the data of the sensing unit;

[0028] Step 3: Grasping execution

[0029] The driving module drives the joint assembly to move according to the instructions of the central processor, and the flexible finger completes the grasping action;

[0030] Step 4: Real-time adjustment

[0031] During the grasping process, the force sensing unit and the tactile sensing unit monitor the grasping state in real time, and the central processor adjusts the grasping force and posture according to the feedback data to ensure stable grasping.

[0032] The beneficial effects of the present invention: By combining flexible joints and various sensors, it can perceive the shape and hardness of the object in real time, dynamically adjust the grasping force and posture, ensure stable grasping of different objects, adopt a multi-degree-of-freedom flexible joint design, can simulate the flexible movement of the human hand, is suitable for various operation tasks, by integrating machine learning algorithms, can automatically learn and optimize the grasping strategy, reduce manual intervention and complex parameter adjustment, and ensure the high precision and reliability of the grasping process through high-precision sensors and real-time feedback control mechanisms. Description of the drawings

[0033] Figure 1 Shown is the first three-dimensional structural schematic diagram of the adaptive manipulator of the present invention;

[0034] Figure 2 Shown is the three-dimensional structural schematic diagram of the joint assembly and the flexible finger in the adaptive manipulator of the present invention;

[0035] Figure 3 The figure shows a schematic diagram of the framework structure of the adaptive manipulator sensing system of the present invention;

[0036] Figure 4 The figure shows a schematic diagram of the framework structure of the adaptive manipulator control system of the present invention;

[0037] Figure 5 The figure shows a flowchart of the adaptive manipulator method of the present invention.

[0038] Description of the reference numerals: 1, base; 21, lifting sleeve; 22, first connecting seat; 23, lifting plate; 24, second connecting seat; 25, first rotating joint arm; 26, third connecting seat; 27, second rotating joint arm; 3, driving device; 4, flexible finger. Detailed implementation manners

[0039] The present invention will be further described below in conjunction with the drawings and embodiments.

[0040] Please refer to Figures 1 - 5 , the present invention provides an embodiment: an adaptive manipulator, including a base 1, a joint assembly, a driving device 3 and a flexible finger 4; further including a sensing system and a control system; a joint assembly with multi-degree-of-freedom movement is arranged at the lower end of the base 1, and a driving device 3 providing power is installed at the lower end of the top of the base 1. The joint assembly includes a lifting sleeve 21 and a first connecting seat 22. A lifting plate 23 is fixedly connected to the lower end of the lifting sleeve 21, and a second connecting seat 24 distributed annularly and equidistantly is fixedly connected to the lower end of the lifting plate 23. A first rotating joint arm 25 is rotatably connected inside the first connecting seat 22. A third connecting seat 26 is fixedly connected to the inner side of the first rotating joint arm 25. A second rotating joint arm 27 is rotatably connected between the third connecting seat 26 and the second connecting seat 24. A flexible finger 4 is fixedly connected to the lower end of the first rotating joint arm 25. The sensing system is distributed at the fingers and the joint assembly, and the control system is arranged inside the base 1;

[0041] The sensing system includes a force sensing unit, a tactile sensing unit and a vision sensing unit;

[0042] The control system includes a central processing unit, a driving module and a communication module. The central processing unit is used to process the data of the sensing system and adjust the movement of the flexible finger 4 according to a preset algorithm. The driving module drives the joint assembly to move according to the instruction of the central processing unit, and the communication module is used to communicate with external devices;

[0043] The control system is also integrated with an adaptive algorithm, which includes a shape recognition algorithm, a force control algorithm, and an attitude adjustment algorithm. The shape recognition algorithm identifies the shape and size of an object through the data of the visual sensing unit and the tactile sensing unit. The force control algorithm adjusts the grasping force in real time according to the feedback of the force sensing unit. The attitude adjustment algorithm adjusts the attitude of the flexible finger 4 according to the shape and position of the object.

[0044] Please refer to Figures 1 - 2 , in this embodiment, the surface of the flexible finger 4 is covered with a flexible material, such as silica gel or rubber, and the driving device 3 adopts any one of a screw rod lifting mechanism, a pneumatic lifting mechanism, an oil cylinder lifting mechanism, and a motor plus a rack and pinion.

[0045] Preferably, for the shape recognition of the shape recognition algorithm, machine learning or deep learning algorithms are used to classify the fused features to identify the object shape. In the size analysis of the shape recognition algorithm, the visual data estimates the object size through geometric methods or deep learning, and the tactile data estimates the object size through the contact area and pressure distribution.

[0046] Preferably, the specific steps of the force control algorithm are as follows:

[0047] A. Collect force sensing data at a fixed frequency and measure the grasping force in real time;

[0048] B. Data preprocessing, including filtering and normalization;

[0049] C. Target force setting

[0050] Preset target force: Set the target grasping force according to the object characteristics;

[0051] Dynamic adjustment: Dynamically adjust the target force according to the task requirements;

[0052] D. Calculate the force error, calculate the difference between the current grasping force and the target force: e = F t -F c ; where F t is the target force and F c is the current force;

[0053] E. Use any one of the PID control, fuzzy control, and adaptive control algorithms to adjust the grasping force.

[0054] Preferably, the attitude adjustment adopts any one of the inverse kinematics algorithm, the force feedback-based control, the optimization algorithm, and the machine learning method.

[0055] Preferably, the force sensing unit is installed at the flexible finger 4 for real-time monitoring of the grasping force, the tactile sensing unit is distributed on the surface of the flexible finger 4 for perceiving the shape and material of the object, and the visual sensing unit is installed on the base 1 for identifying the position and shape of the object.

[0056] An adaptive manipulator method, which includes the adaptive manipulator as described above, and the steps are as follows:

[0057] Step 1: Object recognition

[0058] The force sensing unit monitors the grasping force in real time, the tactile sensing unit perceives the shape and material of the object, the visual sensing unit identifies the position and shape of the object. In the shape recognition of the shape recognition algorithm, machine learning or deep learning algorithms are used to classify the fused features to identify the object shape. In the size analysis of the shape recognition algorithm, the visual data estimates the object size through geometric methods or deep learning, and the tactile data estimates the object size through the contact area and pressure distribution;

[0059] Step 2: Grasping planning

[0060] The central processor plans the movement trajectory and grasping force of the flexible finger 4 according to the data of the sensing unit. The specific steps of the force control are as follows:

[0061] A. Collect force sensing data at a fixed frequency to measure the grasping force in real time;

[0062] B. Data preprocessing, including filtering and normalization;

[0063] C. Target force setting

[0064] Preset target force: Set the target grasping force according to the object characteristics;

[0065] Dynamic adjustment: Dynamically adjust the target force according to the task requirements;

[0066] D. Calculate the force error, calculate the difference between the current grasping force and the target force: e = F t -F c ; where, F t is the target force, and F c is the current force;

[0067] E. Adjust the grasping force by using any one of the PID control, fuzzy control, and adaptive control algorithms;

[0068] Step 3: Grasping execution

[0069] The driving module drives the joint components to move according to the instructions of the central processor, and the flexible finger 4 completes the grasping action;

[0070] Step 4: Real-time adjustment

[0071] During the grasping process, the force sensing unit and the tactile sensing unit monitor the grasping state in real time, and the central processing unit adjusts the grasping force and posture according to the feedback data to ensure stable grasping.

[0072] When it is working, its steps are as follows:

[0073] Step 1: Object recognition

[0074] The force sensing unit monitors the grasping force in real time, the tactile sensing unit senses the shape and material of the object, and the vision sensing unit recognizes the position and shape of the object;

[0075] Step 2: Grasping planning

[0076] The central processing unit plans the movement trajectory and grasping force of the flexible finger 4 according to the data of the sensing unit;

[0077] Step 3: Grasping execution

[0078] The driving module drives the joint components to move according to the instructions of the central processing unit, and the flexible finger 4 completes the grasping action;

[0079] Step 4: Real-time adjustment

[0080] During the grasping process, the force sensing unit and the tactile sensing unit monitor the grasping state in real time, and the central processing unit adjusts the grasping force and posture according to the feedback data to ensure stable grasping.

[0081] Through the above steps, by combining flexible joints and various sensors, it is possible to sense the shape and hardness of the object in real time, dynamically adjust the grasping force and posture, ensure stable grasping of different objects, and by integrating machine learning algorithms, it is possible to automatically learn and optimize the grasping strategy, reducing manual intervention and complex parameter adjustment. Through high-precision sensors and real-time feedback control mechanisms, the high precision and reliability of the grasping process are ensured to solve the problems that rigid structures are difficult to adapt to objects with different shapes and hardnesses, the control method is complex, and it is difficult to achieve adaptive grasping.

[0082] The embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the purpose of the present invention.

Claims

1. An adaptive manipulator, comprising a base (1), a joint assembly, a driving device (3) and a flexible finger (4); characterized in that: It also includes a sensing system and a control system; a joint assembly with multi-degree-of-freedom movement is provided at the lower end of the base (1), a driving device (3) for providing power is installed at the lower end of the top of the base (1), the joint assembly includes a lifting sleeve (21) and a first connecting seat (22), a lifting plate (23) is fixedly connected to the lower end of the lifting sleeve (21), a second connecting seat (24) distributed annularly and equidistantly is fixedly connected to the lower end of the lifting plate (23), a first rotating joint arm (25) is rotatably connected inside the first connecting seat (22), a third connecting seat (26) is fixedly connected to the inner side of the first rotating joint arm (25), a second rotating joint arm (27) is rotatably connected between the third connecting seat (26) and the second connecting seat (24), a flexible finger (4) is fixedly connected to the lower end of the first rotating joint arm (25), the sensing system is distributed at the fingers and the joint assembly, and the control system is arranged inside the base (1); The sensing system includes a force sensing unit, a tactile sensing unit and a visual sensing unit; The control system includes a central processor, a driving module and a communication module. The central processor is used to process the data of the sensing system and adjust the movement of the flexible finger (4) according to a preset algorithm. The driving module drives the joint assembly to move according to the instruction of the central processor, and the communication module is used to communicate with external devices; An adaptive algorithm is also integrated in the control system. The adaptive algorithm includes a shape recognition algorithm, a force control algorithm and an attitude adjustment algorithm. The shape recognition algorithm identifies the shape and size of an object through the data of the visual sensing unit and the tactile sensing unit. The force control algorithm adjusts the grasping force in real time according to the feedback of the force sensing unit. The attitude adjustment algorithm adjusts the attitude of the flexible finger (4) according to the shape and position of the object.

2. The adaptive manipulator according to claim 1, wherein: The surface of the flexible finger (4) is covered with a flexible material, such as silica gel or rubber.

3. The adaptive manipulator according to claim 2, wherein: The driving device (3) adopts any one of a screw rod lifting mechanism, a pneumatic lifting mechanism, an oil cylinder lifting mechanism, and a motor plus a rack and pinion.

4. The adaptive manipulator according to claim 3, wherein: For the shape recognition of the shape recognition algorithm, machine learning or deep learning algorithms are used to classify the fused features to identify the object shape. In the size analysis of the shape recognition algorithm, the visual data estimates the object size through geometric methods or deep learning, and the tactile data estimates the object size through the contact area and pressure distribution.

5. The adaptive manipulator according to claim 4, wherein: The specific steps of the force control algorithm are as follows: A. Collect force sensing data at a fixed frequency to measure the grasping force in real time; B. Data preprocessing, including filtering and normalization; C. Target force setting Preset target force: Set the target grasping force according to the object characteristics; Dynamic adjustment: Dynamically adjust the target force according to the task requirements; D. Force error calculation, calculate the difference between the current grasping force and the target force: e = F t - F c ; where F t is the target force, and F c is the current force; E. Adjust the grasping force by using any one of the algorithms of PID control, fuzzy control, and adaptive control.

6. The adaptive manipulator according to claim 5, characterized in that: The attitude adjustment adopts any one of the algorithms of inverse kinematics algorithm, force feedback-based control, optimization algorithm and machine learning method.

7. The adaptive manipulator according to claim 6, wherein: The force sensing unit is installed at the flexible finger (4) for monitoring the grasping force in real time. The tactile sensing unit is distributed on the surface of the flexible finger (4) for sensing the shape and material of the object. The visual sensing unit is installed on the base (1) for identifying the position and shape of the object.

8. Adaptive manipulator method, characterized in that Including the adaptive manipulator according to claim 7, the steps are as follows: Step 1: Object recognition The force sensing unit monitors the grasping force in real time, the tactile sensing unit senses the shape and material of the object, and the vision sensing unit recognizes the position and shape of the object; Step 2: Grasping planning The central processor plans the movement trajectory and grasping force of the flexible finger (4) according to the data of the sensing unit; Step 3: Grasping execution The drive module drives the joint assembly to move according to the instructions of the central processor, and the flexible finger (4) completes the grasping action; Step 4: Real-time adjustment During the grasping process, the force sensing unit and the tactile sensing unit monitor the grasping state in real time, and the central processor adjusts the grasping force and posture according to the feedback data to ensure stable grasping.

Images