Joint torque sensor zero calibration method, robot and control device thereof

Abstract

The invention discloses a zero calibration method for a joint torque sensor, a robot and a control device of the robot. The zero calibration method comprises the following steps that a joint is made to rotate to a reference point theta 1 in one direction; the reading tau 1 of a joint torque sensor is obtained, and the reading tau 1 comprises the offset tau offset of the torque sensor and the static friction force tau f borne by the torque sensor; rotating the joint part to the reference point theta1 in the other direction; the reading tau 2 of the joint point torque sensor is obtained, and the reading tau 2 comprises the offset tau offset of the torque sensor and the static friction force tau f borne by the torque sensor; and based on the two readings tau 1 and tau 2, eliminating the static friction force tau f in the two readings tau 1 and tau 2 to obtain the offset of the torque sensor, and taking the offset as the zero calibration data of the joint torque sensor to perform zero calibration on the joint.

Description

Technical Field

[0001] This invention relates to the field of robotics, and in particular to a method for zero-point calibration of a joint torque sensor, a robot, and a control device thereof. Background Technology

[0002] Adding torque sensors to robot joints can significantly improve the robot's dynamic performance, playing a crucial role in collision detection, dragging, admittance control, and impedance control. Torque sensors provide real-time feedback of joint torque information, offering precise dynamic data to the robot control system, thereby optimizing the robot's motion trajectory and force control. However, during use, the zero point of the torque sensor may shift due to various factors such as material creep, temperature changes, and mechanical wear. This zero-point shift leads to inaccurate sensor readings, affecting the robot's dynamic control performance. For example, in collision detection, zero-point shift may cause the robot to misjudge the collision force, impacting safety; in dragging and admittance control, zero-point shift reduces control accuracy, affecting the user experience. To address the issue of torque sensor zero-point shift, periodic zero-point calibration is necessary. Therefore, a simple and accurate method is needed to facilitate the calibration of the torque sensor's zero point. Summary of the Invention

[0003] To address the shortcomings of the prior art, the technical problem to be solved by the present invention is to provide a method for zero-point calibration of joint torque sensors that can improve accuracy, as well as a robot and its control device.

[0004] To solve the above-mentioned technical problems, one technical solution adopted by the present invention is: providing a method for zero-point calibration of a torque sensor at a non-gravity axis joint of a robot, comprising:

[0005] Rotate the joint in one direction to the reference point θ1;

[0006] Acquire the reading τ1 of the torque sensor at the joint, wherein the reading τ1 includes the bias τ of the torque sensor. offset and the static friction force τ on the torque sensor f ;

[0007] Rotate the joint in the other direction to the reference point θ1;

[0008] Obtain the reading τ2 of the joint torque sensor, where the reading τ2 includes the bias τ of the torque sensor. offset and the static friction force τ on the torque sensor f ;

[0009] Based on two readings τ1 and τ2, eliminate the static friction τ in the two readings τ1 and τ2. fObtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

[0010] Furthermore, in acquiring the reading τ1 from the joint torque sensor:

[0011] τ1=τ offset +τ f (one)

[0012] The static friction τ when the joint rotates in another direction to the reference point θ1 f The static friction τ when the joint rotates in one direction to the reference point θ1 f Forming a reverse;

[0013] In acquiring the reading τ2 from the joint torque sensor:

[0014] τ2=τ offset -τ f ; (two)

[0015] Based on two readings τ1 and τ2, the static friction τ in the two readings τ1 and τ2 is eliminated. f Obtain the bias τ of the torque sensor offset In the steps, equations (I) and (II) are combined to obtain the bias t of the torque sensor. offset :

[0016] t offset =0.5*(τ1+τ2) (III).

[0017] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is: providing a method for zero-point calibration of a torque sensor at a robot's gravity axis joint, comprising:

[0018] Rotate the joint in one direction to the reference point θ1;

[0019] Acquire the reading τ1 of the torque sensor at the joint, wherein the reading τ1 includes the bias τ of the torque sensor. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g1 ;

[0020] Rotate the joint in the other direction to the reference point θ1;

[0021] Obtain the reading τ2 of the joint torque sensor, where the reading τ2 includes the bias τ of the torque sensor. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g1;

[0022] Based on two readings τ1 and τ2, the static friction force τ is eliminated. f and gravity τ g1 Obtain the bias τ of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

[0023] Furthermore, based on two readings τ1 and τ2, the static friction τ in the two readings τ1 and τ2 is eliminated. f and gravity τ g1 Obtain the bias t of the torque sensor offset The step of using this as zero-point calibration data for the joint torque sensor to perform zero-point calibration specifically includes:

[0024] Based on two readings τ1 and τ2, the average value T1 of the two torque sensor readings τ1 and τ2 is obtained:

[0025] T1 = 0.5 * (τ1 + τ2) = τ offset +τ g1 (Four)

[0026] τ g1 =Mg*L*sin(θ1) (V)

[0027] Where M is the mass and L is the distance from the center of mass to the axis of rotation;

[0028] Rotate the joint forward and backward to the reference point θ1-90°, and obtain the torque sensor reading τ3 during forward rotation and the torque sensor reading τ4 during reverse rotation. Both readings τ3 and τ4 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g2 ;

[0029] Based on two readings τ3 and τ4, the average value T2 of the two torque sensor readings τ3 and τ4 is obtained:

[0030] T2=0.5*(τ3+τ4)=τ offset +τ g2 (six)

[0031] τ g2 =Mg*L*sin(θ1-90)=Mg*L*cos(θ1) (7)

[0032] Combining equations (iv), (v), (vi), and (vii), we get:

[0033] (T1-τoffset ) 2 +(T2-τ offset ) 2 =(Mg*) 2 (eight)

[0034] Rotate the joints forward and backward to the reference point θ2, where θ2 ≠ θ1 ≠ θ1 - 90°.

[0035] Acquire the torque sensor reading τ5 during forward rotation and the torque sensor reading τ6 during reverse rotation. Both readings τ5 and τ6 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g3 ;

[0036] Based on two readings τ5 and τ6, the average value T3 of the two torque sensor readings τ5 and τ6 is obtained:

[0037] T3 = 0.5 * (τ5 + τ6) = τ offset +τ g3 (Nine)

[0038] T g3 =Mg*L*sin(θ2) (x)

[0039] Rotate the joints forward and backward to the reference point θ2-90° respectively;

[0040] Acquire the torque sensor reading τ7 during forward rotation and the torque sensor reading τ8 during reverse rotation. Both readings τ7 and τ8 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g4 ;

[0041] Based on the two readings τ7 and τ8, the average value T4 of the two torque sensor readings τ7 and τ8 is obtained:

[0042] T4 = 0.5 * (τ7 + τ8) = τ offset +τ g4 (eleven)

[0043] τ g4 =Mg·L*sin(θ2-90)=Mg*L*cos(θ2) (XII)

[0044] Combining equations (IX), (X), (XI), and (XII), we obtain:

[0045] (T3-τ offset ) 2 +(T4-τoffset ) 2 =(Mg*L) 2 (Thirteen)

[0046] Combining equation (8) and equation (13), we get:

[0047] (T1-τ offset ) 2 +(T2-τ offset ) 2 =(T3-τ) offset ) 2 +(T4-τ offset ) 2 (fourteen)

[0048] Based on equation (14), the bias τ of the torque sensor is obtained. offset .

[0049] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is: to provide a method for zero-point calibration of a robot joint torque sensor, comprising:

[0050] Obtain the joint to be calibrated;

[0051] Determine whether the joint to be calibrated is a non-gravity axis joint or a gravity axis joint;

[0052] If the joint to be calibrated is a non-gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described above.

[0053] If the joint to be calibrated is a gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described above.

[0054] Furthermore, in the step of obtaining the joint to be calibrated: obtain the axis number of the joint to be calibrated;

[0055] In the step of determining whether the joint to be calibrated is a non-gravity axis joint or a gravity axis joint: the identification and judgment are based on the label pre-assigned to the axis number as either a non-gravity axis joint label or a gravity axis joint label.

[0056] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is: to provide a robot zero-point calibration method for zero-point calibration of torque sensors at joints of the first to nth axes of a multi-axis robot, the method comprising the following steps.

[0057] Obtain the axis number of the joint;

[0058] Determine if the axle number of the joint is greater than the total number of axles;

[0059] If the axle number of the joint is less than or equal to the total number of axles, then a judgment is made based on whether the axle number is a non-gravity axle joint or a gravity axle joint.

[0060] If the joint number of the joint corresponds to a non-gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described above.

[0061] If the joint to be calibrated is a gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described above.

[0062] After incrementing the current axis number by one, proceed to the axis number determination step for the joint.

[0063] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is: to provide a robot control device that generates zero-point calibration data for joint torque sensors, the control device comprising:

[0064] The motion control unit is used to control the joint to rotate forward and backward to the reference point θ1 respectively;

[0065] A torque sensor is used to detect the torque of the joint when it rotates forward and backward to a reference point θ1, so as to obtain the reading τ1 for forward rotation and the reading τ2 for reverse rotation, respectively. When the joint is a non-gravity axis joint, both readings τ1 and τ2 include the bias τ of the torque sensor. offset and the static friction force τ on the torque sensor f ;as well as

[0066] The data generation unit is used to eliminate the static friction τ in the two readings τ1 and τ2 based on the readings τ1 and τ2. f Obtain the bias t of the two readings τ1 and τ2 from the torque sensor. offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

[0067] Furthermore, when the joint is a gravity axis joint, both the readings τ1 and τ2 also include the gravity τ acting on the torque sensor. g1 ;

[0068] The data generation unit is also used to eliminate static friction τ based on two readings τ1 and τ2. f and gravity τ g1 Obtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

[0069] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is to provide a robot, including the control device described above.

[0070] The joint torque sensor zero-point calibration method, robot, and control device described in this invention involve rotating the non-gravity axis joint twice, forward and backward, to a reference point θ1. After reading the values ​​from the torque sensor twice, the static friction τ after each rotation is eliminated by rotating in both directions. f This allows us to obtain the zero-point calibration data of the torque sensor at the non-gravity axis joint, solving the problem of inaccurate values ​​caused by the creep and other characteristics of the torque sensor, and improving the accuracy of zero-point calibration. The zero-point calibration of the gravity axis joint torque sensor involves two forward and reverse rotations to reference points θ1 and θ2, followed by two more forward and reverse rotations to reference points θ1-90° and θ2-90°. The measured readings are then processed using sine and cosine functions to eliminate the gravity effect at each step, thus obtaining the τ of the gravity axis. offset This can solve the impact of sensor creep and other characteristic variations on robot control. Attached Figure Description

[0071] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0072] Figure 1 This is a state diagram of a six-axis robot in the origin posture according to one embodiment of the present invention.

[0073] Figure 2 for Figure 1 A schematic diagram of the posture of a six-axis robot after rotating its two-axis and four-axis axes by 90 degrees.

[0074] Figure 3 This is a flowchart of one embodiment of the zero-point calibration method for the torque sensor of the non-gravity axis joint of the robot according to the present invention.

[0075] Figure 4 This is a flowchart of one embodiment of the zero-point calibration method for the torque sensor of the robot gravity axis joint of the present invention.

[0076] Figure 4-1 This is a diagram showing the state of the robot's joints rotating forward and backward to the reference point θ1.

[0077] Figure 4-2 This is a diagram showing the state of the robot's joints rotating forward and backward to the reference point θ1-90°.

[0078] Figure 4-3 This is a diagram showing the state of the robot's joints rotating forward and backward to the reference point θ2.

[0079] Figure 4-4 The diagram shows the state of the robot's joints rotating in both directions to reference points θ2-90°.

[0080] Figure 5 This is a flowchart of one embodiment of the robot joint torque sensor zero-point calibration method of the present invention.

[0081] Figure 6 This is a flowchart of one embodiment of the robot zero-point calibration method of the present invention.

[0082] Figure 7 This is a block diagram of one embodiment of the control device for the robot of the present invention.

[0083] The diagrams in the instruction manual are labeled as follows:

[0084] Robot 100; One-axis J1; Two-axis J2; Three-axis J3; Four-axis J4; Five-axis J5; Six-axis J6; Motion control unit 101; Torque sensor 102; Data generation unit 103. Detailed Implementation

[0085] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0086] It should be noted that when a component is said to be "fixed to" another component, it can be directly attached to the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component.

[0087] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0088] Please see Figure 1 and Figure 2 , Figure 1The diagram illustrates the state of a six-axis robot in its origin posture. The origin posture means that axes J1, J2, J3, J4, J5, and the sixth axis are all at zero (0, 0, 0, 0, 0, 0). In this posture, axes J1, J4, and J6 are configured as non-gravity axes, J2 and J3 are configured as gravity axes, and J5 is a gravity axis. A gravity axis is one whose torque value is affected by gravity regardless of the robot's posture. A non-gravity axis is one where the robot's posture allows the torque value of that joint to remain constant during movement, unaffected by gravity. Figure 2 The diagram illustrates the orientation (0, 90, 0, 90, 0, 0) of a six-axis robot when axes J1, J3, J5, and J6 are all at the origin, and axes J2 and J4 are at 90 degrees. In this orientation, axes J1, J4, J5, and J6 are non-gravity axes, while axes J2 and J3 are gravity axes. The advantage of this approach is that by classifying as many axes as possible as non-gravity axes, the zero-point calibration logic becomes simpler, improving calibration efficiency.

[0089] It is known that each axis of a six-axis robot has a joint, and each joint is equipped with a drive motor, an encoder that detects the rotation angle of the drive shaft of the drive motor, and a torque sensor that detects the torque of each axis about the drive shaft. Due to the creep characteristic of the sensor, the strain of the torque sensor changes over time when the external torque of the joint remains constant, resulting in inaccurate sensor readings. Therefore, the zero-point calibration method for joint torque sensors of this invention calibrates the zero point of the joint torque sensor of a specific axis based on the value read by the torque sensor of that specific axis, thereby improving the accuracy of the zero-point calibration method. The aforementioned specific axis joint can refer to each axis joint of a multi-axis robot, or it can be a specified one or more axes joints. The zero-point calibration method for joint torque sensors of this invention is not limited to the calibration of six-axis robots as described above, but can be applied to the calibration of joint torque sensors of robots with other numbers of axes.

[0090] Please see Figure 3 , Figure 3 This is a flowchart of one embodiment of the zero-point calibration method for the torque sensor of the non-gravity axis joint of the robot according to the present invention. Figure 3 In the illustrated embodiment, the zero-point calibration method for the torque sensor of the robot's non-gravity axis joint includes the following steps:

[0091] S101, rotate the joint (non-gravity axis joint) in one direction to the reference point θ1.

[0092] S102. Obtain the reading τ1 of the torque sensor at the joint, wherein the reading τ1 includes the bias τ of the torque sensor. offset and the static friction force τ on the torque sensor f .

[0093] In this step, the torque sensor at the joint detects the value (i.e., reading τ1) when the joint rotates in one direction to the reference point θ1. The reading τ1 is expressed by the following formula:

[0094] τ1=τ offset +τ f (one)

[0095] S103. Rotate the joint in another direction to the reference point θ1. This other direction is the opposite of the first direction mentioned above. If the first direction is clockwise, the other direction is counterclockwise, and vice versa.

[0096] S104. Obtain the reading τ2 of the joint torque sensor, wherein the reading τ2 includes the bias τ of the torque sensor. offset and the static friction force τ on the torque sensor f .

[0097] In this step, the torque sensor at the joint detects the value (i.e., reading τ1) when the joint rotates in the other direction to the reference point θ1. The static friction force τ when the joint rotates in the other direction to the reference point θ1 is also measured. f The static friction τ when the joint rotates in one direction to the reference point θ1 f This results in a reversal, therefore the reading τ2 is expressed by the following formula:

[0098] τ2=τ offset -τ f ; (two)

[0099] S105. Based on two readings τ1 and τ2, eliminate the static friction τ in the two readings τ1 and τ2. f Obtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration. Specifically, based on the combination of equations (I) and (II), the bias t of the torque sensor is obtained. offset :

[0100] t offset =0.5*(τ1+τ2) (III)

[0101] In this embodiment, the non-gravity axis joint is rotated twice, forward and backward, to the reference point θ1. After reading the values ​​twice using a torque sensor, the static friction τ after each rotation is eliminated by rotating in both directions. f This allows us to obtain the zero-point calibration data of the torque sensor at the non-gravity axis joint, solving the problem of inaccurate values ​​caused by the creep characteristics of the torque sensor and improving the accuracy of zero-point calibration.

[0102] Please see Figure 4 , Figure 4 This is a flowchart of one embodiment of the zero-point calibration method for the torque sensor of the robot gravity axis joint of the present invention. Figure 4 In the illustrated embodiment, the zero-point calibration method for the robot gravity axis joint torque sensor includes the following steps:

[0103] S201, please refer to Figure 4-1 This causes the joint (gravity axis joint) to rotate in one direction to the reference point θ1.

[0104] S202. Obtain the reading τ1 of the torque sensor at the joint, wherein the reading τ1 includes the bias τ of the torque sensor. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g1 .

[0105] In this step, the torque sensor at the joint detects the value (i.e., reading τ1) when the joint rotates in one direction to the reference point θ1. The reading τ1 is expressed by the following formula:

[0106] τ1=τ offset +τ f +τ g1 (one)

[0107] S203. Rotate the joint in another direction to the reference point θ1. This other direction is the opposite of the first direction mentioned above. If the first direction is clockwise, the other direction is counterclockwise, and vice versa.

[0108] S204. Obtain the reading τ2 of the joint torque sensor, wherein the reading τ2 includes the bias τ of the torque sensor. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g1 .

[0109] In this step, the torque sensor at the joint detects the value (i.e., reading τ2) when the joint rotates in the other direction to the reference point θ1. The static friction force τ when the joint rotates in the other direction to the reference point θ1 is also measured. f The static friction τ when the joint rotates in one direction to the reference point θ1 f This results in a reversal, therefore the reading τ2 is expressed by the following formula:

[0110] τ2=τ offset -τ f +τ g1 ; (two)

[0111] S205. Based on two readings τ1 and τ2, eliminate static friction τ. f and gravity τ g1 Obtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration. This step specifically includes the following sub-steps:

[0112] S2051. Based on two readings τ1 and τ2, obtain the average value T1 of the two readings τ1 and τ2 from the torque sensor:

[0113] T1 = 0.5 * (τ1 + τ2) = τ offset +τ g1 (Four)

[0114] τ g1 =Mg*L*sin(θ1) (V)

[0115] Where M is the mass, and L is the distance from the center of mass to the axis of rotation. Equation (V) is used to eliminate the static friction τ generated by rotation. f .

[0116] S2052, please refer to Figure 4-2 Rotate the joint forward and backward to the reference point θ1-90°, and obtain the torque sensor reading τ3 during forward rotation and the torque sensor reading τ4 during reverse rotation. Both readings τ3 and τ4 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g2 In this step:

[0117] τ3=τ offset +τ f +τ g2 (VI-I)

[0118] τ4=τ offset -τ f +τ g2(VI-II)

[0119] Based on two readings τ3 and τ4, the average value T2 of the two torque sensor readings τ3 and τ4 is obtained:

[0120] T2=0.5*(τ3+τ4)=τ offset +τ g2 (six)

[0121] τ g2 =Mg·L*sin(θ1-90)=Mg*L*cos(θ1) (7)

[0122] Combining equations (iv), (v), (vi), and (vii), we get:

[0123] (T1-τ offset ) 2 +(T2-τ offset ) 2 =(Mg*L) 2 (eight)

[0124] S2053. Rotate the joint parts clockwise and counterclockwise to reference point θ2, where θ2 ≠ θ1 ≠ θ1 - 90°; please refer to [link / reference]. Figure 4-3 This is a diagram showing the state of the joint as it rotates forward and backward to the reference point θ2.

[0125] S2054. Obtain the torque sensor reading τ5 during forward rotation and the torque sensor reading τ6 during reverse rotation. Both readings τ5 and τ6 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g3 In this step:

[0126] τ5=τ offset +τ f +τ g3 (9-I)

[0127] τ6=τ offset -τ f +τ g3 (IX-II)

[0128] Based on two readings τ5 and τ6, the average value T3 of the two torque sensor readings τ5 and τ6 is obtained:

[0129] T3 = 0.5 * (τ5 + τ6) = τ offset +τ g3 (Nine)

[0130] τ g3 =Mg*L*sin(θ2) (x)

[0131] S2055. Rotate the joint in both directions to the reference point θ2-90° respectively. Please refer to 4-4 for a diagram showing the state of the joint rotated in both directions to the reference point θ2-90° respectively.

[0132] S2056. Obtain the torque sensor reading τ7 during forward rotation and the torque sensor reading τ8 during reverse rotation. Both readings τ7 and τ8 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g4 In this step:

[0133] τ7=τ offset +τ f +τ g4 (9-I)

[0134] τ8=τ offset -τ f +τ g4 (IX-II)

[0135] Based on the two readings τ7 and τ8, the average value T4 of the two torque sensor readings τ7 and τ8 is obtained:

[0136] T4 = 0.5 * (τ7 + τ8) = τ offset +τ g4 (eleven)

[0137] τ g4 =Mg*L*sin(θ2-90)=Mg*L*cos(θ2) (XII)

[0138] Combining equations (IX), (X), (XI), and (XII), we obtain:

[0139] (T3-τ offset ) 2 +(T4-τ offset ) 2 =(Mg*L) 2 (Thirteen)

[0140] Combining equation (8) and equation (13), we get:

[0141] (T1-τ offset ) 2 +(T2-τ offset ) 2 =(T3-τ) offset ) 2 +(T4-τ offset ) 2 (fourteen)

[0142] S2057. Based on equation (XIV), obtain the bias τ of the torque sensor. offset .

[0143] In this embodiment of the invention, the zero-point calibration of the torque sensor at the gravity axis joint is achieved by rotating it twice in both directions to reference points θ1 and θ2, and then rotating it twice in both directions to reference points θ1-90° and θ2-90°. The measured readings are then processed using sine and cosine functions to eliminate the gravity force at each step, thereby obtaining the τ of the gravity axis. offset This can solve the problem of sensor creep affecting robot control.

[0144] Please see Figure 5 , Figure 5 This is a flowchart of one embodiment of the robot joint torque sensor zero-point calibration method of the present invention. Figure 5 In the illustrated embodiment, the zero-point calibration method for the robot joint torque sensor includes the following steps:

[0145] S301. Obtain the joint to be calibrated, wherein the joint to be calibrated refers to the torque sensor of the joint to be calibrated.

[0146] In this step, it is preferable to obtain the axis number of the joint to be calibrated.

[0147] S302. Determine whether the joint to be calibrated is a non-gravity axis joint or a gravity axis joint.

[0148] In this step, it is preferable to identify the axis number based on whether the pre-assigned label is a non-gravity axis joint label or a gravity axis joint label. For the distinction between non-gravity axis joints and gravity axis joints, see [link to relevant documentation]. Figure 1 , Figure 2 And related descriptions in the text.

[0149] S303. If the joint to be calibrated is a non-gravity axis joint, then refer to... Figure 3 The zero-point calibration method for the non-gravity axis joint torque sensor described above is used to perform zero-point calibration on the joint torque sensor to be calibrated.

[0150] S304. If the joint to be calibrated is a gravity axis joint, then refer to... Figure 4 The zero-point calibration method for the joint torque sensor of the gravity axis described above is used to calibrate the joint torque sensor to be calibrated.

[0151] Please see Figure 6 , Figure 6 This is a flowchart of one embodiment of the robot zero-point calibration method of the present invention. Figure 6In the illustrated embodiment, the robot zero-point calibration method is used to perform zero-point calibration of the torque sensors at the joints of the first to nth axes of a multi-axis robot (n axes). The method includes the following steps:

[0152] S401, Obtain the axis number of the joint.

[0153] In this embodiment, combined with Figure 1 and Figure 2 The six-axis robot shown is used as an example for illustration. The six-axis robot has a total of 6 axes. The axis number of the first axis J1 can be assigned the value 1, the axis number of the second axis J2 can be assigned the value 2, the axis number of the third axis J3 can be assigned the value 3, the axis number of the fourth axis J4 can be assigned the value 4, the axis number of the fifth axis J5 can be assigned the value 5, and the axis number of the sixth axis J6 can be assigned the value 6.

[0154] S402. Determine whether the axis number of the joint is greater than the total number of axes. If it is less than or equal to the total number of axes, proceed to step S403. If it is greater than the total number of axes, proceed to step S408.

[0155] In this embodiment, taking the zero-point calibration from axis J1 to axis J6 in sequence as an example, firstly, obtain axis number 1 of axis J1 (axis number 1 is less than the total number of axes 6). After the zero-point calibration of axis J1 is completed, continue the zero-point calibration of axis J2 until all six axes of the six-axis robot are calibrated.

[0156] S403. If the axle number of the joint is less than or equal to the total number of axles, then a judgment is made based on whether the axle number is a non-gravity axle joint label or a gravity axle joint label. The joint corresponding to the axle number of the joint is determined to be either a non-gravity axle joint or a gravity axle joint.

[0157] S404. If the joint number of the joint is a non-gravity axis joint, then the torque sensor of the joint to be calibrated shall be zero-point calibrated in accordance with the zero-point calibration method of the non-gravity axis joint torque sensor described above.

[0158] Continue with Figure 1 and Figure 2 Taking the illustrated implementation as an example, the first axis J1, the fourth axis J4, the fifth axis J5, and the sixth axis J6 of the six-axis robot are all determined to be non-gravity axes. When the axis numbers are 1, 4, 5, and 6, the zero-point calibration is performed according to the non-gravity axis joint torque sensor zero-point calibration method described above.

[0159] S406. If the joint to be calibrated is a gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described in claim 3 or 4.

[0160] Continue with Figure 1 and Figure 2Taking the illustrated implementation as an example, the two axes J2 and the three axes J3 of the six-axis robot are both determined to be gravity axes. When the axis numbers are 2 and 3, the zero-point calibration is performed according to the zero-point calibration method of the gravity axis joint torque sensor described above.

[0161] S407. Increment the current axis number by one and then proceed to the axis number determination step for the joint.

[0162] S408. If the axis number of the joint is greater than the total number of axes, it is considered that the zero-point calibration of the torque sensors of all joints of the robot has been completed.

[0163] Please see Figure 7 , Figure 7 This is a block diagram of one embodiment of the control device for the robot of the present invention. Figure 7 In the illustrated embodiment, the robot's control device generates zero-point calibration data for the joint torque sensors by controlling the joints to rotate in the corresponding directions described in the above embodiment. The robot's control device includes an arithmetic processing unit (CPU) that serves as a processor, and a storage medium connected to the CPU via a bus. The control device inputs a pre-made motion program for performing the movements of each joint of the robot. The control device includes a motion control unit 101 that controls the robot's movements. The motion control unit 101 sends motion commands to the robot's joints based on the motion program (e.g., driving the robot's joints to move according to the methods described above). The robot joints perform corresponding movements based on the motion commands. The control device includes torque sensors 102 disposed at each joint; in this embodiment, six drive motors and six torque sensors 102 are configured for the six joints. The control device also includes a data generation unit 103.

[0164] Specifically, the motion control unit 101 is used to control the joint to rotate forward and backward to a reference point θ1, respectively. The torque sensor 102 is used to detect the torque of the joint when it rotates forward and backward to the reference point θ1, so as to obtain the reading τ1 for forward rotation and the reading τ2 for reverse rotation, respectively. Wherein: for non-gravity axis joints, both readings τ1 and τ2 include the bias τ of the torque sensor 102. offset and the static friction force τ acting on the torque sensor 102 f The data generation unit 103 is used to eliminate the static friction force τ in the two readings τ1 and τ2 based on the readings τ1 and τ2. f The bias t of the two readings τ1 and τ2 from the torque sensor 102 is obtained. offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration. For the gravity axis joint, both readings τ1 and τ2 also include the gravity τ acting on the torque sensor.g1 The data generation unit 103 is also used to eliminate static friction τ based on two readings τ1 and τ2. f and gravity τ g1 Obtain the bias t of torque sensor 102 offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

[0165] The present invention also discloses a robot, including the control device described above.

[0166] The above embodiments merely illustrate preferred implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention should be determined by the appended claims.

Claims

1. A method for zero-point calibration of a torque sensor at a non-gravity axis joint of a robot, characterized in that, include: Rotate the joint in one direction to the reference point θ1; Acquire the reading τ1 of the torque sensor at the joint, wherein the reading τ1 includes the bias τ of the torque sensor. offset And the static friction force τ on the torque sensor f ; Rotate the joint in the other direction to the reference point θ1; Obtain the reading τ2 of the joint torque sensor, where the reading τ2 includes the bias τ of the torque sensor. offset And the static friction force τ on the torque sensor f ; Based on two readings τ1 and τ2, eliminate the static friction τ in the two readings τ1 and τ2. f Obtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

2. The zero-point calibration method for a robot non-gravity axis joint torque sensor as described in claim 1, characterized in that, In acquiring the reading τ1 from the torque sensor at the joint: τ1=τ offset +t f (one) The static friction τ when the joint rotates in another direction to the reference point θ1 f The static friction τ when the joint rotates in one direction to the reference point θ1 f Forming a reverse; In acquiring the reading τ2 from the joint torque sensor: τ2 = τ offset -τ f ; (2) Based on two readings τ1 and τ2, the static friction τ in the two readings τ1 and τ2 is eliminated. f Obtain the bias t of the torque sensor offset In the steps, equations (I) and (II) are combined to obtain the bias t of the torque sensor. offset : t offset =0.5*(τ1+τ2)(three).

3. A method for zero-point calibration of a torque sensor at a robot's gravity axis joint, characterized in that, include: Rotate the joint in one direction to the reference point θ1; Acquire the reading τ1 of the torque sensor at the joint, wherein the reading τ1 includes the bias τ of the torque sensor. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g1 ; Rotate the joint in the other direction to the reference point θ1; Obtain the reading τ2 of the joint torque sensor, where the reading τ2 includes the bias τ of the torque sensor. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g1 ; Based on two readings τ1 and τ2, the static friction force τ is eliminated. f and gravity τ g1 Obtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

4. The zero-point calibration method for the torque sensor of the robot gravity axis joint as described in claim 3, characterized in that, Based on two readings τ1 and τ2, the static friction τ in the two readings τ1 and τ2 is eliminated. f and gravity τ g1 Obtain the bias t of the torque sensor offset The step of using this as zero-point calibration data for the joint torque sensor to perform zero-point calibration specifically includes: Based on two readings τ1 and τ2, the average value T1 of the two torque sensor readings τ1 and τ2 is obtained: T1 = 0.5 * (τ1 + τ2) = τ offset + τ g1 (IV) τ g1 = Mg * L * sin(θ1) (V) Where M is the mass and L is the distance from the center of mass to the axis of rotation; Rotate the joint forward and backward to the reference point θ1-90°, and obtain the torque sensor reading τ3 during forward rotation and the torque sensor reading τ4 during reverse rotation. Both readings τ3 and τ4 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g2 ; Based on two readings τ3 and τ4, the average value T2 of the two torque sensor readings τ3 and τ4 is obtained: T2 = 0.5 * (τ3 + τ4) = τ offset + τ g2 (VI) τ g2 = Mg * L * sin(θ1 - 90) = Mg * L * cos(θ1) (VII) Combining equations (iv), (v), (vi), and (vii), we get: (T1 - τ offset ) 2 +(T2 - τ offset ) 2 =(Mg * L) 2 (Eight) Rotate the joints forward and backward to the reference point θ2, where θ2 ≠ θ1 ≠ θ1 - 90°. Acquire the torque sensor reading τ5 during forward rotation and the torque sensor reading τ6 during reverse rotation. Both readings τ5 and τ6 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g3 ; Based on two readings τ5 and τ6, the average value T3 of the two torque sensor readings τ5 and τ6 is obtained: T3 = 0.5 * (τ5 + τ6) = τ offset + τ g3 (Nine) τ g3 = Mg * L * sin(θ2) (Equation (10)) Rotate the joints forward and backward to the reference point θ2-90° respectively; Acquire the torque sensor reading τ7 during forward rotation and the torque sensor reading τ8 during reverse rotation. Both readings τ7 and τ8 include the torque sensor bias τ. offset The static friction force τ experienced by the torque sensor f and the gravitational force τ acting on the torque sensor g4 ; Based on the two readings τ7 and τ8, the average value T4 of the two torque sensor readings τ7 and τ8 is obtained: T4 = 0.5 * (τ7 + τ8) = τ offset + τ g4 (XI) τ g4 = Mg * L * sin(θ2 - 90) = Mg * L * cos(θ2) (Equation 12) Combining equations (IX), (X), (XI), and (XII), we obtain: (T3-τ offset ) 2 +(T4-τ offset ) 2 =(Mg*L) 2 (Thirteen) Combining equation (8) and equation (13), we get: (T1 - τ offset ) 2 +(T2 - τ offset ) 2 =(T3 - τ offset ) 2 +(T4 - τ offset ) 2 (Fourteen) Based on equation (14), the bias τ of the torque sensor is obtained. offset .

5. A method for zero-point calibration of a robot joint torque sensor, characterized in that, include: Obtain the joint to be calibrated; Determine whether the joint to be calibrated is a non-gravity axis joint or a gravity axis joint; If the joint to be calibrated is a non-gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method of claim 1 or 2. If the joint to be calibrated is a gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described in claim 3 or 4.

6. The zero-point calibration method for a robot joint torque sensor as described in claim 5, characterized in that: In the step of obtaining the joint to be calibrated: obtain the axis number of the joint to be calibrated; In the step of determining whether the joint to be calibrated is a non-gravity axis joint or a gravity axis joint: the identification and judgment are based on the label pre-assigned to the axis number as either a non-gravity axis joint label or a gravity axis joint label.

7. A robot zero-point calibration method for zero-point calibration of torque sensors at joints of the first to nth axes of a multi-axis robot, the method comprising the following steps: Obtain the axis number of the joint; Determine if the axle number of the joint is greater than the total number of axles; If the axle number of the joint is less than or equal to the total number of axles, then a judgment is made based on whether the axle number is a non-gravity axle joint or a gravity axle joint. If the joint number of the joint is a non-gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method of claim 1 or 2. If the joint to be calibrated is a gravity axis joint, then the torque sensor of the joint to be calibrated is zero-point calibrated according to the method described in claim 3 or 4. After incrementing the current axis number by one, proceed to the axis number determination step for the joint.

8. A robot control device that generates zero-point calibration data for joint torque sensors, characterized in that, include: The motion control unit is used to control the joint to rotate forward and backward to the reference point θ1 respectively; A torque sensor is used to detect the torque of the joint when it rotates forward and backward to a reference point θ1, so as to obtain the reading τ1 for forward rotation and the reading τ2 for reverse rotation, respectively. When the joint is a non-gravity axis joint, both readings τ1 and τ2 include the bias τ of the torque sensor. offset And the static friction force τ on the torque sensor f ;as well as The data generation unit is used to eliminate the static friction τ in the two readings τ1 and τ2 based on the readings τ1 and τ2. f Obtain the bias t of the two readings τ1 and τ2 from the torque sensor. offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

9. The robot control device as described in claim 8, characterized in that, When the joint is a gravity axis joint, both readings τ1 and τ2 also include the gravity τ acting on the torque sensor. g1 ; The data generation unit is also used to eliminate static friction τ based on two readings τ1 and τ2. f and gravity τ g1 Obtain the bias t of the torque sensor offset This data is used as the zero-point calibration data for the joint torque sensor to perform zero-point calibration.

10. A robot, characterized in that, Includes the control device as described in claim 8.

Images