A human-machine compliant interaction control method based on a multi-dimensional force sensor

Abstract

This invention belongs to the field of human-computer interaction technology and relates to a human-computer compliant interaction control method based on a multi-dimensional force sensor. This method involves setting characteristic parameters for switching human-computer interaction control modes; measuring the pre-grasping state using a multi-dimensional force sensor to obtain the gravitational disturbance of the gripper in the pre-grasping state; measuring the robot's grasping state using a multi-dimensional force sensor to obtain the gravitational disturbance of the object; calculating the automatic switching characteristic parameters for the human-computer interaction control mode; and achieving compliant human-computer interaction by comparing the automatic switching characteristic parameters with the switching characteristic parameters and by applying preset constraints. This invention achieves a quasi-balance between the human gripping force and the object's gravity at the moment of object transfer and release, thereby significantly reducing the impact during the human-computer interaction process and improving the compliance and safety of the interaction.

Description

Technical Field

[0001] This invention belongs to the field of human-computer interaction technology, specifically relating to a human-computer compliant interaction control method based on a multi-dimensional force sensor. Background Technology

[0002] Human-robot interaction refers to the process by which humans communicate, cooperate, and exchange ideas with robots to achieve information transfer, task execution, and knowledge sharing. Human-robot interaction can occur in various ways, including voice commands, gestures, touchscreens, and interaction through other external devices. These interaction methods are limited to information transfer; how humans and robots exchange objects has not yet been reported. Common applications of human-robot interaction include service robots, manufacturing, education and entertainment, and healthcare.

[0003] The existing technology still has the following problems that urgently need to be solved:

[0004] 1. Human-Computer Interaction Safety. The transfer of objects between humans and robots is a crucial step in human-computer interaction, but existing control strategies have significant shortcomings. When a robot transfers an object to a human, the object's gravitational load is rapidly applied to the human body within a very short time, causing a gravitational impact. This impact can not only cause injury to the operator but also lead to operational errors, seriously affecting the safety and reliability of the interaction. Therefore, current human-computer interaction control strategies still face significant challenges in applications with high safety requirements.

[0005] 2. Adaptability of Interaction. Most existing interaction methods are rather rigid and lack adaptability. For example, force-controlled interaction is easily interfered with under different robotic arm postures, leading to misjudgments. Specifically, when the robotic arm is in different spatial positions and postures, the environment and force direction of the sensors change, causing inconsistencies and fluctuations in sensor signals, thus affecting the accuracy and stability of force-controlled interaction. For instance, when the robotic arm moves from a horizontal to a vertical position, the force value detected by the sensor will change significantly due to the change in the direction of gravity. If the system cannot adjust and compensate for these changes in time, misjudgments are likely to occur, leading to interaction failure. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a human-machine compliant interaction control method based on a multi-dimensional force sensor. The aim is to achieve compliant transfer control of objects between humans and robots during human-machine interaction, improve the safety of human-machine interaction, and meet the application scenarios with strict safety requirements.

[0007] To achieve the objectives of this invention, the following technical solutions are adopted.

[0008] A human-computer compliant interaction control method based on a multi-dimensional force sensor, applied to a human-computer interaction system based on a multi-dimensional force sensor, includes the following steps:

[0009] S1. Set the human-computer interaction control mode switching characteristic parameters: F mode F threshold , where: F mode For the critical switching between incremental force control and gravity balance modes, F threshold The magnitude of the interaction force when the object is released;

[0010] S2. The human-machine interaction control system is initialized. The robot enters the grasping preparation state. The pre-grasping state is measured by a multi-dimensional force sensor to obtain the gravity disturbance F of the gripper in the pre-grasping state. o =FF g ;

[0011] S3. During the process of the robot grasping the object, the robot's grasping state is measured by a multi-dimensional force sensor to obtain the gravitational disturbance F of the object. o =FF g Calculate the gravitational disturbance F o The norm of ||F o || ∞ That is, the characteristic parameter ||F of the automatic switching of human-computer interaction control mode. o || ∞ ;

[0012] S4. During human-computer interaction, if ||F o || ∞ ≥F mode The robot then enters incremental interactive force control mode and prompts the user to apply an interactive force F by dragging the object. i To achieve human-computer interaction, the robot simultaneously calculates the interaction force F in real time. i In order to obtain the interaction force F i The norm of ||F i || ∞ , through ||F i || ∞ The control signal u is calculated, and the state of the robot hand is determined by the preset constraint condition (1) to achieve compliant human-machine interaction; wherein: the constraint condition (1) is as follows:

[0013]

[0014] If ||F o || ∞ <F mode The robot enters the gravity balance interactive control mode and prompts the user to apply an interactive force F to the object in the opposite direction of its gravity. iA multi-dimensional force sensor measures the external load response F in real time to obtain the norm of the external load response F, ||F||. ∞ , through ||F|| ∞ The control signal u is calculated, and the state of the robot hand is determined by the preset constraint condition (2) to achieve compliant human-machine interaction; wherein: the constraint condition (2) is as follows:

[0015]

[0016] Where: u g As a given constant, initial settings need to be completed before human-computer interaction control.

[0017] Furthermore, the interaction force F i The calculation formula is F i =FF g -F o .

[0018] Furthermore, in the incremental force control mode, when ||F i || ∞ <f threshold , u = u g When ||F|| i || ∞ ≥f threshold When u = 0, the robot hand releases the object.

[0019] Furthermore, in the incremental force control mode, when ||F i || ∞ <f threshold When, u=u g When ||F i || ∞ ≥f threshold At that time, u = 0.

[0020] Furthermore, in the gravity balance control mode, when the condition ‖‖F‖‖ is satisfied ∞ ≥f threshold , u = u g When ||F|| is satisfied, the robot hand maintains the object grasping state; when ||F|| is satisfied, the robot hand maintains the object grasping state. ∞ <f threshold When u = 0, the robot hand releases the object.

[0021] Furthermore, in the gravity balance control mode, when ||F|| ∞ ≥f threshold When, u=u g When ||F|| ∞ <f threshold At that time, u = 0.

[0022] Furthermore, the human-computer interaction control mode switching characteristic parameters are set based on experience.

[0023] Furthermore, the human-computer interaction system is composed of a robot, a multi-dimensional force sensor, and a robotic hand connected in series.

[0024] Beneficial effects

[0025] This invention discloses a human-computer compliant interaction control method based on a multi-dimensional force sensor, which includes an incremental interaction force control mode and a gravity balance interaction control mode. The incremental interaction force control mode is used for the transfer control of lightweight objects. For heavy objects, the gravity balance interaction control mode can guide the human body to resist the object's gravity. At the moment of object transfer and release, the human body's gripping force and the object's gravity are quasi-balanced, thereby greatly eliminating the impact during the human-computer interaction process and improving the compliance and safety of the interaction. Attached Figure Description

[0026] Figure 1 This is a structural diagram of the robot described in this invention;

[0027] Figure 2 This is a schematic diagram of the human-computer interaction process described in this invention;

[0028] Figure 3 This is a framework diagram of the human-computer interaction control described in this invention;

[0029] Figure 4 This is a diagram of the human-computer interaction process under the incremental force control mode described in this invention; wherein: Figure A shows the robot hand grasping and holding an object; Figure B shows the operator's finger applying a force to the object; Figure C shows the incremental force control mode being activated and the robot hand releasing the object;

[0030] Figure 5 Figure A shows the force response during human-computer interaction under the incremental interactive force control mode described in this invention; where: Figure A shows the response of the three-dimensional force sensor in human-computer interaction; Figure B shows a partial magnification of the three-dimensional force sensor response under the force control interaction mode.

[0031] Figure 6 This is a diagram illustrating the human-computer interaction process under the gravity balance force control mode described in this invention; wherein: Figure A shows the robot hand grasping and holding a heavy object; Figure B shows the operator's hand lifting the heavy object upwards to apply an operational force; Figure C shows the gravity balance force control mode being activated, and the robot hand releasing the object;

[0032] Figure 7 Figure A shows the force response during the human-computer interaction process under the gravity balance interactive force control mode described in this invention; where: Figure A shows the response of the three-dimensional force sensor in human-computer interaction; Figure B is a partial magnification of the three-dimensional force sensor response under the force control interaction mode.

[0033] Figure 8The diagrams show the human-computer interaction process and force response of the human-computer interaction control strategy described in this invention under tilted conditions; wherein: Figure A shows the robot hand in a vertically downward posture; Figure B shows the robot hand maintaining a 45-degree tilted posture; Figure C shows the robot hand maintaining a horizontal posture; Figure D shows the robot hand maintaining a 45-degree tilted posture again; Figure E shows the human-machine object transfer interaction process under a 45-degree tilted posture; Figure F shows the force response process from Figure A to Figure E.

[0034] Figure 9 Figure A shows the interaction sensitivity test embodiment of the human-computer interaction system described in this invention; wherein: Figure A shows the response of the three-dimensional force sensor in human-computer interaction; Figure B is a magnified view of the response of the three-dimensional force sensor in force control interaction mode.

[0035] Figure 10 The diagram shows the implementation process of the human-computer interaction system interaction sensitivity test according to the present invention, wherein: Figure A shows the robot hand grasping and holding the object; Figure B shows the operator's hand pinching one corner of the object and slowly dragging the object downward to apply the operating force; Figure C shows the incremental force control mode being activated and the robot hand releasing the object. Detailed Implementation

[0036] The present invention will be further described in conjunction with the embodiments and accompanying drawings.

[0037] Example 1, as Figure 1 As shown, the human-computer interaction system based on a multi-dimensional force sensor includes: robot 1, multi-dimensional force sensor 2, and robot hand 3. Robot 1, multi-dimensional force sensor 2, and robot hand 3 are connected in series with each other.

[0038] Example 2, as Figure 2 As shown, the gravitational disturbance of object 4 is F. o The robot hand's 3-gravity perturbation is F. g Human body 5 contacts object 4 and applies an interaction force F. i F o F g F i The resultant force acts on the multidimensional force sensor 2 to generate a response F.

[0039] Example 3, as Figure 3 As shown, a human-machine compliant interaction control method based on a multi-dimensional force sensor includes the following steps:

[0040] Step 1: Set the interaction mode switching feature parameter F mode F threshold .

[0041] Step 2: Human-machine interaction control initialization. The robot enters the grasping preparation state, and measures the pre-grasping state using multi-dimensional force sensors to obtain the gravitational disturbance F of the gripper in the pre-grasping state. g=F.

[0042] Step 3: Automatic switching of human-computer interaction control mode feature parameters ||F o || ∞ Calculation. The robot grasps and lifts the object, measuring the grasping state using multi-dimensional force sensors to obtain the gravitational disturbance F of the grasped object during the grasping state. o =FF g By calculating ||F o || ∞ Determining the human-computer interaction control mode is one of the key features of this invention.

[0043] Step 4: Human-computer interaction control, if ||F o || ∞ ≥F mode The robot enters an incremental interactive force control mode, which is one of the key features of this invention. A human hand grasps the object and drags it, applying an interactive force F. i via F i =FF g -F o Real-time calculation of interaction force F i Through ||F i || ∞ Calculate the control signal u to control the gripper to release the object and transfer it to the hand. The constraint condition (1) is as follows:

[0044]

[0045] Conversely, if ||F o || ∞ <F mode The robot enters the gravity balance interactive control mode, and the human hand grasps the object being grasped and applies an interactive force F to it in the opposite direction of gravity. i A multi-dimensional force sensor measures the external load response F in real time. This is one of the key features of this invention.

[0046] Through ||F o || ∞ Calculate the control signal u to control the gripper to release the object and transfer it to the human hand. The constraint condition (2) is as follows:

[0047]

[0048] Example 4 is an example of the human-computer interaction process under incremental interactive force control mode. For example... Figure 4 , Figure 5 As shown, where, Figure 4 Human-computer interaction process, Figure 5This describes the force response during human-computer interaction. The object being handled is a flexible packaging bag with a weight of approximately 0.2 N. After the clamping mechanism closes, the object's weight will act on the three-dimensional force sensor, producing a negative offset of approximately -0.2 N. Figure 5 As shown in Figure A. The corresponding grabbing states are as follows. Figure 4 As shown in Figure A. After approximately 10.5 seconds, the incremental force control mode is activated. At this point, the force measurement result is reset, and the amplitude immediately drops to zero. In the experiment, the interactive control threshold F was set... threshold The force was set to 1.5N. After approximately 15 seconds, an external interaction force was artificially applied to the object in the -z axis direction, such as... Figure 4 As shown in Figure B. During the interaction, F z The amplitude gradually increases and reaches a peak of approximately 1.64N during the interaction. According to constraint (1), when ||F i || ∞ ≥f threshold At that time, u will be by u g Switching to 0 will cause the robot gripper to open its gripping mechanism to release the object, such as... Figure 4 As shown in Figure C. After the object is released, F z The amplitude will immediately decrease, producing a positive deflection of approximately 0.2 N, roughly equal to the object's weight, such as... Figure 5 As shown in Figure B. Finally, at 22 seconds, a zero-point calibration command is sent to the microcontroller to reset the 3D force sensor measurements, preparing for the next grasping and interactive control.

[0049] Example 5 is an example of the human-computer interaction process under gravity balance force control mode, such as... Figure 6 , Figure 7 As shown. In gravity balance control, a dumbbell with a weight of 9.8N is used as the object to be grasped and transferred. F mode Set it to 5N to ensure that the gravity balance control mode can be activated under a 9.8N load. Also, set f... threshold Set to 0.2N. Figure 6 , Figure 7 This describes the interaction process and response of gripping and transferring a dumbbell under gravity balance control mode. During the experiment, the robot gripper grips the dumbbell and reaches a steady state, as shown... Figure 6 As shown in Figure A. Under the influence of the object's gravity, F z This produces an offset of approximately -9.8N, such as Figure 7 As shown in Figure A, at 11.5 seconds, the test subject gradually applied an interactive force along the z-axis to the dumbbell, as shown in Figure A. Figure 6 As shown in Figure B, F increases with the interaction force. z The amplitude gradually decreases, such as Figure 7 As shown in Figure B. At 13 seconds, F zThe amplitude decreased to approximately 0.02N, and the robot gripper released the dumbbell, as... Figure 6 As shown in Figure C. According to constraint (2), in ||F|| ∞ <f threshold Under constraints, the interaction force F i and the object's gravity F o They have equal magnitudes. The actual difference between them depends on f. threshold This means that during the process of transferring the object from the robot to the tester's hand, as long as f threshold It is small enough that the tester's hand can achieve a quasi-static balance with the object, enabling the object to be transferred smoothly without producing obvious impact.

[0050] Example 6 is an embodiment in which the human-computer interaction system is installed on a collaborative robot to further verify the effectiveness of the human-computer interaction control strategy in a tilted state. Figure 8 As shown, an underactuated robotic hand was mounted on the end joint of the robotic arm using a 3D force sensor, and human-robot interaction tests were conducted in different postures. First, in a vertical position, a circular object with a weight of 0.9N was grasped, as shown... Figure 8 As shown in Figure A. At this time, F z This produces a negative offset of approximately -0.9N, such as Figure 8 As shown in F. Then, in Figure 8 In diagrams B and C, the robot gripper gradually rotates from a vertical to a horizontal position as the robotic arm moves. During this process, F... x F z The measurement results are affected by gravitational disturbances. This is because, as the states of the gripper and the 3D force sensor change, the gripper's gravity generates attitude-dependent components along both the x and z axes. Figure 8 In diagrams A through C, reset operations were performed at 3.3 seconds, 6.7 seconds, and 9.4 seconds. Figure 8 In diagram D, the robot gripper returns to a 45° tilt position to perform the object transfer operation. At 11.5 seconds, the incremental force control mode is activated, and f is set. threshold The force is 3N. Then, at 13.7 seconds, the tester drags the object naturally, gradually increasing the interaction force. At 13.9 seconds, F satisfies ||F|| ∞ ≥f threshold The object is released by the gripper and successfully transferred to the tester's hand, such as Figure 8 As shown in Figure E.

[0051] Example 7 is an example of testing the interaction sensitivity of the human-computer interaction system, such as... Figure 9 , Figure 10 As shown. To test the limiting sensitivity of the interactive grasping system, at different f... thresholdHuman-computer interaction testing was conducted at various amplitude levels. In the sensitivity test, a flexible, easily deformable packaging bag was selected as the object for grasping and interaction. Figure 10 As shown in Figure A. During the human-computer interaction, the tester gently pinched one corner of the flexible packaging bag with their fingers and followed... Figure 10 The z-axis direction is slightly dragged as shown in Figure B. When F... z When it drops to -0.22 Newtons, as Figure 9 As shown in Figure A, the robot gripper is activated, and the object is released by the robot gripper, as... Figure 10 As shown in Figure C. (Through) Figure 9 As shown in Figure B, although the tester's finger primarily applied force in the -z axis direction, as F... z As amplitude increases, F x F y The measured peak values ​​reached approximately 0.18 N and 0.16 N, respectively. This indicates that, at a sensitivity setting of 0.2 N, F x F y The measurement results may erroneously trigger the release action. During the experiment, it was found that at this sensitivity level, the tester could hardly precisely control the magnitude and direction of the interactive force. The main reason for this phenomenon is that fingers and objects are soft and easily deformable; during interaction, fingers and objects deform, thus generating perturbations in multiple dimensions. Therefore, the limiting sensitivity of the interactive grasping system is approximately 0.2N. In practical applications, f threshold The setting should be greater than 0.2N to ensure the reliability of human-computer interaction.

[0052] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A human-computer compliant interaction control method based on a multi-dimensional force sensor, applied to a human-computer interaction system based on a multi-dimensional force sensor, characterized in that: Includes the following steps: S1. Set the human-computer interaction control mode switching characteristic parameters: F mode F threshold , where: F mode For the critical switching between incremental force control and gravity balance modes, F threshold The magnitude of the interaction force when the object is released; S2. Initialize the human-machine interaction system. The robot enters the grasping preparation state. The pre-grasping state is measured by a multi-dimensional force sensor to obtain the gravity disturbance F of the gripper in the pre-grasping state. o =FF g ; S3. During the process of the robot grasping the object, the robot's grasping state is measured by a multi-dimensional force sensor to obtain the gravitational disturbance F of the object. o =FF g Calculate the gravitational disturbance F o The norm of ||F o || ∞ That is, the characteristic parameter ||F of the automatic switching of human-computer interaction control mode. o || ∞ ; S4. During human-computer interaction, if ||F o || ∞ ≥F mode The robot then enters incremental interactive force control mode and prompts the user to apply an interactive force F by dragging the object. i To achieve human-computer interaction, the robot simultaneously calculates the interaction force F in real time. i In order to obtain the interaction force F i The norm of ||F i || ∞ , through ||F i || ∞ The control signal u is calculated, and the state of the robot hand is determined by the preset constraint condition (1) to achieve compliant human-machine interaction; wherein: the constraint condition (1) is as follows: If ||F o || ∞ <F mode The robot enters the gravity balance interactive control mode and prompts the user to apply an interactive force F to the object in the opposite direction of its gravity. i A multi-dimensional force sensor measures the external load response F in real time to obtain the norm of the external load response F, ||F||. ∞ , through ||F|| ∞ The control signal u is calculated, and the state of the robot hand is determined by the preset constraint condition (2) to achieve compliant human-machine interaction; wherein: the constraint condition (2) is as follows: Where: u g As a given constant, initial settings need to be completed before human-computer interaction control.

2. The human-machine compliant interaction control method based on a multi-dimensional force sensor according to claim 1, characterized in that: The interaction force F i The calculation formula is F i =FF g -F o .

3. The human-machine compliant interactive control method based on a multi-dimensional force sensor according to claim 1, characterized in that: In incremental force control mode, when ||F i || ∞ <f threshold , u = u g When ||F|| i || ∞ ≥f threshold When u = 0, the robot hand releases the object.

4. The human-machine compliant interactive control method based on a multi-dimensional force sensor according to claim 3, characterized in that: In incremental force control mode, when ||F i || ∞ <f threshold When, u=u g When ||F i || ∞ ≥f threshold At that time, u = 0.

5. The human-machine compliant interactive control method based on a multi-dimensional force sensor according to claim 1, characterized in that: In the gravity balance control mode, when ||F|| ∞ ≥f threshold , u = u g When ||F|| is satisfied, the robot hand maintains the object grasping state; when ||F|| is satisfied, the robot hand maintains the object grasping state. ∞ <f threshold When u = 0, the robot hand releases the object.

6. The human-machine compliant interactive control method based on a multi-dimensional force sensor according to claim 5, characterized in that: In the gravity balance control mode, when ||F|| ∞ ≥f threshold When, u=u g When ||F|| ∞ <f threshold At that time, u = 0.

7. The human-machine compliant interactive control method based on a multi-dimensional force sensor according to claim 1, characterized in that: The human-computer interaction control mode switching characteristic parameters are set based on experience.

8. The human-machine compliant interactive control method based on a multi-dimensional force sensor according to claim 1, characterized in that: The human-computer interaction system consists of a robot, a multi-dimensional force sensor, and a robotic hand connected in series.

Images