A dynamic modeling method for multi-axis robots based on end-force control sensors

Abstract

This invention proposes a method for dynamic modeling of a multi-axis robot based on an end-point force control sensor. The method comprises the following steps: constructing a parameter identification function, calculating the estimated error between the actual trajectory of the robot end and the estimated trajectory, statistically obtaining multiple sets of control force and estimated error data pairs through multiple experiments, calculating the four parameter identification coefficients of the parameter identification function using the least squares method, and finally constructing the output expression of the robot's force controller. The method utilizes an external ATI mechanical sensor in conjunction with the robot's position control output. The method first establishes a parameter identification function through parameter identification, and then establishes a corresponding parameter model between input torque and output position by performing dynamic simulation on the robot. This modeling process requires fewer parameters, is simple to calculate, and has a low error.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a multi-axis robot dynamics modeling method based on an end force control sensor. Background Art

[0002] Industrial robots and collaborative robots typically operate using a robotic arm. When performing tasks involving force interactions with the environment, using position control alone can cause excessive forces due to position errors, potentially damaging the environment. Position control typically involves ensuring that the robotic arm achieves a specified joint angle in joint space, which corresponds to a specified spatial pose in Cartesian space. However, if the robotic arm encounters obstacles or is disturbed by external forces such as friction while moving to a specified position, the position control requires it to reach the specified position, and the robotic arm will increase its force output until it exceeds the limit. Therefore, force control is often used in conjunction with robotic arms when moving in confined environments.

[0003] During medical surgery, the robot senses the user's operating force control in different scenarios through human-machine collaboration, and generates a planned trajectory in real time to complete the operation of the robotic arm within the corresponding expected range; there are two main ways of force control in terms of interactive force detection. One is to read the current loop and convert the current loop value into the torque of each joint through mathematical formulas; the other method is to install a six-axis or three-axis force sensor at the end of the robotic arm.

[0004] However, during the control process, the mechanical control of the robotic arm is mainly troubled by the following two aspects: First, many robot manufacturers do not open the current loop interface or are not equipped with torque sensors, so developers cannot directly control and collect the corresponding torque information of the robot; second, even if some manufacturers open the relevant current loop and torque control interfaces, their robot dynamics models are analyzed from the perspective of joint modeling, with many parameters, complex calculations and large errors. Summary of the Invention

[0005] The present invention proposes a multi-axis robot dynamics modeling method based on an end force control sensor, which solves the problems of numerous parameters, complex calculations and large errors in the dynamics modeling process of the force control robot in the prior art.

[0006] The technical solution of the present invention is achieved as follows:

[0007] According to one aspect of the present invention, a method for dynamic modeling of a multi-axis robot based on an end force control sensor is provided, comprising the following steps:

[0008] Construct a parameter identification function:

[0009] y k+1 =a1yk +a2y k-1 +b1u k +b2u k-1

[0010] Where y represents the position information of the end of the robot arm; u represents the magnitude of the loading control force; y′ k+1 represents the estimated position of the end of the manipulator at the next moment; a1 represents the parameter identification coefficient related to the position of the end of the manipulator at the current moment; a2 represents the parameter identification coefficient related to the position of the end of the manipulator at the previous moment; b1 represents the parameter identification coefficient related to the magnitude of the force at the current moment; b2 represents the parameter identification coefficient related to the magnitude of the force at the previous moment;

[0011] The estimated error between the actual trajectory of the robot end and the estimated trajectory is:

[0012] Ess=y k+1 -y′ k+1

[0013] Among them, y k+1 Indicates the actual position of the end of the robotic arm at the next moment;

[0014] Assume the control error is:

[0015] e k+1 =y d，k -a1y k -a2y k-1 -b1u k -b2u k-1

[0016] Among them, y d，k Indicates the preset target position of the end of the robotic arm;

[0017] Let e k+1 =(1-p)e k

[0018] Then the output expression of the robot's force controller is:

[0019]

[0020] Where ρ represents the adjustment parameter.

[0021] The present invention uses an external ATI mechanical sensor in conjunction with the robot's position control output. First, a parameter identification function is established in a parameter identification manner. Then, by performing dynamic simulation of the robot, a corresponding parameter model between the input torque and the output position is established. The modeling process has few parameters, simple calculations, and small errors.

[0022] As a preferred solution of the present invention, the method for calculating the four parameter identification coefficients is: after multiple experiments, multiple sets of control force and estimation error data pairs are obtained, and the four parameter identification coefficients a1, a2, b1, and b2 are calculated using the least squares method.

[0023] As a preferred solution of the present invention, during the experiment, the constraints of the control force are as follows:

[0024]

[0025] That is, during the experiment, the loading of the control force is carried out in an increasing or decreasing manner.

[0026] As a preferred solution of the present invention, the period between the current moment and the adjacent moment is 0.015s.

[0027] As a preferred solution of the present invention, the dynamic modeling method also includes model verification, and an experiment is designed to verify the model based on the output expression of the robot's force controller to determine whether the control accuracy of the robotic arm meets the requirements. If not, the adjustment parameters need to be corrected.

[0028] According to another aspect of the present invention, a multi-axis robot based on an end force control sensor is also provided, including a controller and a robotic arm and a mechanical sensor electrically connected to the controller. The controller controls the displacement of the end of the robotic arm based on the above-mentioned dynamic model according to the torque data sensed by the mechanical sensor.

[0029] Beneficial effects

[0030] Compared with the prior art, the beneficial effect of the present invention lies in: the present invention uses an external ATI mechanical sensor, cooperates with the robot's position control output, first establishes a parameter identification function in a parameter identification manner, and then establishes a corresponding parameter model between the input torque and the output position by performing dynamic simulation of the robot. The modeling process has few parameters, simple calculations, and small errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0032] Figure 1 A schematic flow chart of a multi-axis robot dynamics modeling method based on an end force control sensor according to the present invention;

[0033] Figure 21 is a schematic diagram of an estimation error-timing simulation obtained according to the data in Table 1 in an embodiment of the present invention;

[0034] Figure 3 Schematic diagram of simulation obtained based on collected control force and time data in an embodiment of the present invention;

[0035] Figure 4 Schematic diagram of simulation obtained based on data of control error and time in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0037] Reference Figure 1 As shown, this embodiment provides a multi-axis robot dynamic modeling method based on the end force control sensor. This embodiment uses an external ATI mechanical sensor and cooperates with the robot's position control output. First, a parameter identification function is established in a parameter identification manner. Then, by performing dynamic simulation on the robot, a corresponding parameter model between the input torque and the output position is established. The specific steps are as follows:

[0038] Construct a parameter identification function:

[0039] y k+1 =a1y k +a2y k-1 +b1u k +b2u k-1

[0040] Where y represents the position information of the end of the manipulator (z-axis coordinate value); u represents the magnitude of the loading control force (the force on the end of the manipulator in the z-axis direction); y′ k+1 represents the estimated position of the end of the manipulator at the next moment; a1 represents the parameter identification coefficient related to the position of the end of the manipulator at the current moment; a2 represents the parameter identification coefficient related to the position of the end of the manipulator at the previous moment; b1 represents the parameter identification coefficient related to the magnitude of the force at the current moment; b2 represents the parameter identification coefficient related to the magnitude of the force at the previous moment; the period between the current moment and the adjacent moment is 0.015s;

[0041] Let [a1, a2, b1, b2] T =θ k ,[y k ,y k-1 ,u k ,uk-1 ] T =x k

[0042] Then, y k+1 =θ k ·x k

[0043]

[0044]

[0045] Then the estimated error between the actual trajectory of the robot end and the estimated trajectory is calculated as:

[0046] Ess=y k+1 -y k+1

[0047] Among them, y k+1 Indicates the actual position of the end of the robotic arm at the next moment;

[0048] A robot model was established to obtain data pairs of the magnitude of the z-axis loading torque at the end of the manipulator and the estimated error of the position of the manipulator end. After multiple experiments, multiple sets of control force (loading torque) and estimated error data pairs were obtained, as shown in Table 1 below:

[0049] Table 1 Statistics of control power and estimation error data

[0050]

[0051] Matlab is used to simulate the data in Table 1 above, and the simulation results are as follows Figure 2 As shown, Figure 2 In , the horizontal axis represents the time series, and the vertical axis represents the estimated error value;

[0052] The least squares method is used to calculate the four parameter identification coefficients a1, a2, b1, and b2, where:

[0053] a1=1.4024, a2=-0.4, b1=0.0143, b2=-0.0071

[0054] During the experiment, the constraints of the control force are as follows:

[0055]

[0056] That is, during the experiment, the loading of the control force is increased or decreased. When the control force increases and reaches the upper boundary of 40N, it begins to decrease. When the control force decreases and reaches the lower boundary of 0N, it begins to increase.

[0057] Assume the control error is:

[0058] e k+1 =y d，k -a1y k -a2y k-1 -b1u k -b2u k-1

[0059] Among them, y d，k Indicates the preset target position of the end of the robotic arm;

[0060] Let e k+1 =(1-ρ)e k

[0061] Then the output expression of the force controller for constructing the robot is:

[0062]

[0063] Among them, ρ = 0.5 represents the adjustment parameter;

[0064] Set the step distance to 1mm, collect the data of robot control force and time, and simulate it through matlab as follows Figure 3 As shown; at the same time, the control error and time data of the robot end position and the preset target position are collected, and the data is simulated by matlab as shown Figure 4 As shown;

[0065] from Figure 4 It can be seen that the end of the robotic arm finally reached the target point and maintained a very small jitter at the target point, achieving the expected effect; the maximum error between the moving path of the end of the robotic arm and the preset path was controlled within the range of [-0.01~0.025]mm, and the motion trajectory of the end of the robotic arm also achieved the expected effect.

[0066] This embodiment also provides a multi-axis robot based on an end force control sensor, including a controller and a robotic arm and a mechanical sensor electrically connected to the controller. The controller controls the displacement of the end of the robotic arm based on the above-mentioned dynamic model according to the torque data sensed by the mechanical sensor.

[0067] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-axis robot dynamics modeling method based on end force control sensor, characterized in that: The following steps are involved: Construct a parameter identification function: and' k+1 =a1y k +a2y k-1 +b1u k +b2u k-1 Where y represents the position information of the end of the robot arm; u represents the magnitude of the loading control force; y′ k+1 represents the estimated position of the end of the manipulator at the next moment; a1 represents the parameter identification coefficient related to the position of the end of the manipulator at the current moment; a2 represents the parameter identification coefficient related to the position of the end of the manipulator at the previous moment; b1 represents the parameter identification coefficient related to the magnitude of the force at the current moment; b2 represents the parameter identification coefficient related to the magnitude of the force at the previous moment; The estimated error between the actual trajectory of the robot end and the estimated trajectory is: Essy k+1 -and' k+1 Among them, y k+1 Indicates the actual position of the end of the robotic arm at the next moment; Assume the control error is: E k+1 =y d，k -a1y k -a2y k-1 -b1u k -b2u k-1 Among them, y d，k Indicates the preset target position of the end of the robotic arm; Let e k+1 =(1-ρ)e k Then the output expression of the robot's force controller is: Where ρ represents the adjustment parameter.

2. A multi-axis robot dynamic modeling method based on an end force control sensor according to claim 1, characterized in that: The method for calculating the four parameter identification coefficients is as follows: after multiple experiments, multiple sets of control force and estimation error data pairs are obtained, and the four parameter identification coefficients a1, a2, b1, and b2 are calculated using the least squares method.

3. A multi-axis robot dynamic modeling method based on an end force control sensor as claimed in claim 2, characterized in that: During the experiment, the constraints of the control force are as follows: That is, during the experiment, the loading of the control force is carried out in an increasing or decreasing manner.

4. The multi-axis robot dynamic modeling method based on the end force control sensor according to claim 1, characterized in that: The period between the current moment and the adjacent moment is 0.015s.

5. The multi-axis robot dynamic modeling method based on the end force control sensor according to claim 1, characterized in that: It also includes model verification. Experiments are designed based on the output expression of the robot's force controller to verify the model and determine whether the control accuracy of the robotic arm meets the requirements. If not, the adjustment parameters need to be corrected.

6. A multi-axis robot based on an end force control sensor, based on the dynamic modeling method according to any one of claims 1 to 5, characterized in that: The invention comprises a controller, a mechanical arm and a mechanical sensor electrically connected to the controller. The controller controls the end of the mechanical arm to move according to the torque data sensed by the mechanical sensor.

Images