Mechanical arm man-machine cooperation self-adaptive admittance control method and device, terminal and medium

Abstract

The invention belongs to the field of robots, and particularly discloses a mechanical arm man-machine cooperation self-adaptive admittance control method and device, a terminal and a medium, a man-machine interaction model omitting virtual stiffness is constructed, and admittance parameters in the man-machine interaction model comprise virtual inertia and virtual damping; designing a stability observer of the physical man-machine interaction system, monitoring an external force applied by an operator by the stability observer, and performing spectral analysis on a force signal to obtain a system instability index; constructing an admittance parameter adaptive controller through a system instability index, wherein the admittance parameter adaptive controller comprises a virtual inertia adaptive controller and a virtual damping adaptive controller; and the virtual inertia and the virtual damping are adjusted based on the admittance parameter self-adaptive controller, so that the physical man-machine interaction system returns to a stable state. The method realizes self-adaptive adjustment of controller parameters, adapts to a complex environment, and keeps stability of physical man-machine interaction.

Description

Technical Field

[0001] This invention belongs to the field of robotics, specifically relating to a human-machine collaborative adaptive admittance control method, device, terminal, and medium for robotic arms. Background Technology

[0002] Physical human-machine interaction (HMI), as a crucial feature of human-machine collaboration, has seen widespread research and application in recent years in fields such as drag-and-drop teaching, compliant assembly, and assisted rehabilitation. From the perspective of equipment, it can be mainly divided into HMI based on dynamic models and HMI based on external sensors. In terms of control methods, it is mainly divided into impedance control and admittance control. Impedance control directly outputs torque commands to the servo drive, while admittance control outputs position commands, adding an inner control loop compared to impedance control. Generally, robot impedance control based on dynamic models allows for stable dynamic interactions between the robot and a rigid environment. However, due to uncertainties such as friction, gear backlash, and environmental factors, it is difficult to establish an accurate robot dynamic model. Furthermore, when the robot interacts with a soft environment or moves in free space, these uncertainties can also significantly interfere with the controller, leading to poor impedance control accuracy. In contrast, admittance control can still maintain high control accuracy in non-contact tasks or interactions with soft environments.

[0003] The main challenge of admittance control in human-robot interaction tasks stems from the dynamic and unknown interaction forces generated during the interaction process, which depend on the operator's habits and the given task. Related neurocontrol research shows that humans have the ability to adapt to external disturbances by adjusting arm stiffness, and research in control engineering also indicates that robot stability largely depends on the admittance parameters in the human-robot interaction task. If the parameters are not chosen properly, the system will become unstable, meaning the robot's behavior deviates from expectations, ultimately leading to oscillations and an unsafe physical human-robot interaction state.

[0004] This paper discloses a stability control method for a humanoid flexible arm system. Virtual damping and inertia parameters are selected as adjustment admittance parameters. The method improves the stability of the humanoid flexible arm system by adjusting these virtual damping and inertia parameters. The control equations are constructed based on the rated output torque of the motor, the instantaneous electromagnetic torque of the motor, the reduction ratio of the reducer, the speed of the connecting rod, and the inherent rotational inertia of the joint. While this method can improve system stability, it has the following drawbacks: 1. Based on a second-order differential model of the flexible joint, it relies on many inherent hardware design parameters and physical structure parameters (such as spring stiffness) to describe the system. This modeling method is relatively fixed, and when the physical structure or hardware parameters of the system change, the applicability and adjustment of the model become difficult. 2. The practical application of virtual parameters is constrained by the performance of the DC power supply, which limits the adjustable range and flexibility of the parameters. 3. Stability is evaluated by calculating the system's natural oscillation angular frequency and damping ratio using fixed formulas. This method is based on a static model and fixed parameter calculations, making it difficult to reflect the dynamic changes and potential instability factors of the system during operation in real time. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method, device, terminal, and medium for adaptive admittance control of a robotic arm in human-machine collaboration. The method involves performing spectral analysis on force signals to obtain system instability indices, constructing an adaptive admittance parameter controller based on these instability indices, and then adjusting virtual inertia and virtual damping based on the adaptive admittance parameter controller. This achieves adaptive adjustment of controller parameters, adapting to complex environments and maintaining stable physical human-machine interaction.

[0006] In a first aspect, the technical solution of the present invention provides a human-machine collaborative adaptive admittance control method for a robotic arm, comprising the following steps: Construct a human-computer interaction model that omits virtual stiffness. The admittance parameters in the human-computer interaction model include virtual inertia and virtual damping. Design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index. Admittance parameter adaptive controllers are constructed using system instability indices. These adaptive controllers include virtual inertia adaptive controllers and virtual damping adaptive controllers. The virtual inertia and virtual damping are adjusted by an adaptive controller based on admittance parameters, so that the physical human-computer interaction system returns to a stable state.

[0007] In an optional implementation, a human-computer interaction model omitting virtual stiffness is constructed, specifically including: Establish the admittance control equations for the end effector of the robotic arm, where contact occurs only. (1) in, and These represent the desired speed at the end effector of the robotic arm and the actual speed of the robotic arm, respectively. , , These represent virtual inertia, virtual damping, and virtual stiffness, respectively. Indicates the actual applied force; Omitting the virtual stiffness, the admittance governing equations are simplified to: (2) By transforming and discretizing equation (2), we obtain: (3) in, , , It is the velocity of the joint, obtained by differentiating the joint position. The acceleration of the joint is obtained by performing differentiation. , It is the Jacobian matrix of the robotic arm. Indicates the first One cycle, Indicates the duration of the sampling period; The human-computer interaction is modeled as a first-order system. The simplified human-computer interaction model represents the muscle characteristics of the human arm. The admittance control equation (2) is rewritten in the form of an admittance function. (4) in, Represent the complex variables in the Laplace transform; The closed-loop transfer function of a physical human-computer interaction system, with contact force as input and the measured pose of the robotic arm's end effector as output, is established as follows: (5) in, The stiffness of a human arm, This indicates the time delay of human-computer interaction response. Indicates measurement location and command location The transfer function between them.

[0008] In an optional implementation, constructing a human-computer interaction model that omits virtual stiffness further includes: A dynamic model of the robotic arm was obtained through experiments, and the measurement position was estimated using the dynamic model. and command location Transfer function between .

[0009] In one optional implementation, spectral analysis of the force signal is performed to obtain system instability indices, specifically including: right A window of the latest force signal measurements is used to perform a Fast Fourier Transform calculation to obtain the frequency components of the force signal in the frequency domain. amplitude ; By the magnitude of the frequency component Calculate the instability index of the calculation system.

[0010] In an optional implementation, the amplitude of the frequency component is used. The instability metrics of a computational system include: Determine the frequency range boundary using the Nyquist-Shannon sampling theorem. ; Determine the crossover frequency used to distinguish between stable and unstable motion. ; Determine the stability threshold ; The system instability index is calculated using the following formula. (6) in, It is the lowest frequency of the Fast Fourier Transform.

[0011] In an optional implementation, an adaptive controller for admittance parameters is constructed using system instability indices, specifically including: The system instability index calculated by formula (6) Normalization is performed to obtain new system instability indices. ; Construct an adaptive controller for the admittance parameters according to the following formula. (7) in, , To maintain a stable minimum virtual inertia and virtual damping for physical human-computer interaction systems, , To maintain stable inertia and damping reference values ​​for physical human-computer interaction systems.

[0012] In one optional implementation, , The physical human-computer interaction system is designed to maintain a critically stable value even when the operator's arm stiffness is at its maximum. , .

[0013] Secondly, the technical solution of the present invention provides a robotic arm human-machine collaborative adaptive admittance control device, comprising: The human-computer interaction model construction module is used to construct a human-computer interaction model that omits virtual stiffness. The human-computer interaction model includes admittance parameters, which include virtual inertia and virtual damping. The system instability index calculation module is used to design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index. The admittance parameter adaptive controller construction module is used to construct an admittance parameter adaptive controller based on system instability indices. The admittance parameter adaptive controller includes an adaptive controller with virtual inertia and virtual damping. The admittance parameter adjustment module is used to adjust the virtual inertia and virtual damping based on the admittance parameter adaptive controller, so as to keep the physical human-computer interaction system in a stable state.

[0014] Thirdly, the technical solution of the present invention provides a terminal, comprising: The memory is used to store the adaptive admittance control program for human-robot collaboration of the robotic arm. A processor is configured to implement the steps of the robotic arm human-machine collaborative adaptive admittance control method as described above when executing the robotic arm human-machine collaborative adaptive admittance control program.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing a robotic arm human-machine collaboration adaptive admittance control program, wherein when the robotic arm human-machine collaboration adaptive admittance control program is executed by a processor, it implements the steps of the robotic arm human-machine collaboration adaptive admittance control method as described in any of the above claims.

[0016] The present invention provides a human-machine collaborative adaptive admittance control method, device, terminal and medium for robotic arms, which has the following advantages compared with the prior art: First, a human-machine interaction model including virtual inertia and virtual damping is constructed. During the operation of the robotic arm, the external force applied by the operator is monitored by a stability observer. Then, the system instability index is obtained by spectrum analysis of the force signal. Based on the system instability index, an adaptive admittance parameter controller is constructed. The virtual inertia and virtual damping are adjusted by the constructed adaptive admittance parameter controller to keep the system stable. This invention: 1) Performs spectral analysis on force signals, converting time-domain force signals to the frequency domain. This allows for precise capture of different frequency components and their amplitude variations. Abnormal fluctuations in high-frequency components often indicate potential oscillation risks in the system. By monitoring subtle changes in the force signal, instability can be predicted in advance, providing a more timely and accurate basis for system stability assessment. Furthermore, the constructed system instability index integrates the characteristic information of the force signal at multiple frequencies, providing a comprehensive and quantitative value that can more meticulously describe the stable state of the system, making stability assessment more objective and accurate; 2) Constructs an admittance parameter adaptive controller based on the system instability index. This controller can dynamically adjust parameters such as virtual inertia and virtual damping according to the real-time calculated index. When the system detects signs of instability, it can quickly increase damping to suppress oscillations, or adjust the virtual inertia... 3) Compared to stability control based on fixed physical models and parameters, which is difficult to make timely and effective adjustments when facing complex and ever-changing working environments (such as differences in the operating habits of different operators), this invention can quickly perceive the impact of environmental changes on system stability by analyzing the force signal spectrum in real time, reflect the dynamic changes and potential unstable factors of the system during operation in real time, and adjust the admittance parameters in a timely manner through an adaptive controller, so that the robotic arm can maintain stable operation under various complex working conditions; 4) The adaptive control method mainly adjusts parameters based on the spectral characteristics of the force signal and the system instability index, and is less directly constrained by factors such as hardware power supply, and can flexibly adjust parameters in a wider range to adapt to the dynamic changes of the system. Attached Figure Description

[0017] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of a human-machine collaborative adaptive admittance control method for a robotic arm, provided as an embodiment of the present invention.

[0019] Figure 2This is a schematic block diagram of a robotic arm human-machine collaborative adaptive admittance control device provided in an embodiment of the present invention.

[0020] Figure 3 This is a schematic diagram of the structure of a terminal provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the specific embodiments. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. 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.

[0022] 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.

[0023] The key terms used in this invention will be explained below.

[0024] RHSO: Relative Hamiltonian Stability Observer.

[0025] pHRI: Physical Human-Robot Interaction.

[0026] FFT: Fast Fourier Transform.

[0027] Figure 1 This is a schematic flowchart of a human-machine collaborative adaptive admittance control method for a robotic arm, provided as an embodiment of the present invention. Figure 1 The executing entity can be a robotic arm-human-machine collaborative adaptive admittance control system. The robotic arm-human-machine collaborative adaptive admittance control method provided in this embodiment of the invention is executed by a computer device; correspondingly, the robotic arm-human-machine collaborative adaptive admittance control system runs within the computer device. Depending on different requirements, the order of the steps in this flowchart can be changed, and some steps can be omitted.

[0028] like Figure 1 As shown, the method includes the following steps.

[0029] S1. Construct a human-computer interaction model that omits virtual stiffness. The admittance parameters in the human-computer interaction model include virtual inertia and virtual damping.

[0030] The purpose of this step is to construct a human-computer interaction model, which includes admittance parameters. In this embodiment, considering that the human-computer interaction process in free space does not require restoring force, the virtual stiffness in the admittance parameters is omitted. The admittance parameters in the human-computer interaction model constructed in this embodiment include virtual inertia and virtual damping. Subsequently, the stability of the physical human-computer interaction system is controlled based on the human-computer interaction model constructed in this step, mainly by adjusting the virtual inertia and virtual damping to ensure stability.

[0031] S2. Design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index.

[0032] The purpose of this step is to perform spectral analysis on the force signal to obtain system instability indicators. First, a stability observer is used to monitor the external force applied to the robotic arm. Then, spectral analysis is performed on the force signal to obtain instability indicators for the physical human-machine interaction system. Force signals often fail to intuitively reflect internal instability factors in the time domain, but by converting them to the frequency domain through spectral analysis, the amplitudes of different frequency components can be clearly displayed. Variations in the force signal at different frequencies may correspond to different operating states of the system; for example, high-frequency components may be related to oscillations or unstable motion. By analyzing these frequency components and their amplitudes, the dynamic characteristics of the system can be accurately captured, providing a more detailed and accurate basis for judging whether the system is stable.

[0033] S3 constructs an adaptive controller for admittance parameters based on system instability indices. The adaptive controller for admittance parameters includes an adaptive controller for virtual inertia and an adaptive controller for virtual damping.

[0034] The purpose of this step is to construct an adaptive controller for admittance parameters. In this embodiment, an adaptive controller for admittance parameters is constructed based on the system instability index, including an adaptive controller for virtual inertia and an adaptive controller for virtual damping.

[0035] In human-machine collaboration, the external environment and operator behavior are constantly changing, making it difficult to guarantee stable system operation under various conditions with fixed admittance parameters. This embodiment utilizes an adaptive controller built based on system instability indices. This controller automatically adjusts parameters such as virtual inertia and virtual damping in the admittance controller according to real-time calculated instability indices. When the system detects signs of instability, the controller can promptly increase damping to suppress oscillations or adjust the virtual inertia to alter the system's dynamic response characteristics, thereby enabling the system to quickly return to a stable state.

[0036] S4, based on the admittance parameter adaptive controller, adjusts the virtual inertia and virtual damping to bring the physical human-computer interaction system back to a stable state.

[0037] Step S3 above constructs an adaptive controller for virtual inertia and an adaptive controller for virtual damping. When the physical human-computer interaction system is unstable, this step adjusts the virtual inertia and virtual damping through these two adaptive controllers, so that the physical human-computer interaction system can quickly return to a stable state.

[0038] In some implementations, step S1, which involves constructing a human-computer interaction model that omits virtual stiffness, specifically includes the following steps.

[0039] S1.1, Establish the admittance control equations for the end effector of the robotic arm where contact only occurs. (1) in, and These represent the desired speed at the end effector of the robotic arm and the actual speed of the robotic arm, respectively. , , These represent virtual inertia, virtual damping, and virtual stiffness, respectively. This indicates the actual force applied.

[0040] S1.2, omitting the virtual stiffness, simplifies the admittance governing equations to: (2) S1.3, transforming and discretizing equation (2), we obtain, (3) in, , , It is the velocity of the joint, obtained by differentiating the joint position. The acceleration of the joint is obtained by performing differentiation. , It is the Jacobian matrix of the robotic arm. Indicates the first One cycle, Indicates the duration of the sampling period.

[0041] It should be noted that the sampling period is relatively short. and The difference is small, so we use... Replaced ,and =0, this is the initial value for the configuration.

[0042] Considering that digital control systems typically operate with discrete time steps, this embodiment performs discretization processing. The discretized equations are suitable for processing in a digital control system. By converting continuous admittance control equations into discrete form, they can be combined with a computer control system, facilitating real-time control and adjustment using the computer's computing power, thereby improving the system's control accuracy and stability.

[0043] S1.4, the human-computer interaction is modeled as a first-order system. The simplified human-computer interaction model represents the muscle characteristics of the human arm. The admittance control equation (2) is rewritten in the form of an admittance function. (4) in, This represents a complex variable in the Laplace transform.

[0044] This embodiment models human-computer interaction as a first-order system, and the simplified model represents the muscle characteristics of a human arm. The modeling method in this embodiment closely resembles the actual human-computer interaction process because the muscles of the human arm exhibit certain inertia and damping characteristics during movement, which a first-order system can effectively simulate. By employing the modeling method of this embodiment, the dynamic behavior during human-computer interaction can be accurately described, improving the accuracy and reliability of the model.

[0045] S1.5, the closed-loop transfer function of the physical human-computer interaction system, with contact force as input and the measured pose of the robotic arm's end effector as output, is: (5) in, The stiffness of a human arm, This indicates the time delay of human-computer interaction response. Indicates measurement location and command location The transfer function between them. This embodiment obtains the dynamic model of the robotic arm through experiments, and estimates the measurement position using the dynamic model. and command location Transfer function between .

[0046] An optional implementation involves first preparing for the experiment, including ensuring the robotic arm is in normal working order, checking the mechanical structure for robustness, joint movement for smoothness, and the functionality of all sensors (such as position sensors and force sensors) to collect data. Data acquisition equipment, such as a data acquisition card, should be prepared to record relevant motion data from the robotic arm. A suitable experimental site should be selected, ensuring environmental stability and avoiding external interference (such as vibration or electromagnetic interference) that could affect the experimental results. The robotic arm should be properly secured and calibrated to ensure its initial position and posture are accurate.

[0047] Then, dynamic model experiments are performed to obtain the data, including designing suitable excitation signals, data acquisition experiments, data preprocessing, and dynamic model identification. During the data acquisition experiment, the designed excitation signals are input to the robotic arm, and data acquisition equipment is used to record the input signals (such as the control voltage of the drive motor) and output signals of each joint of the robotic arm. The acquired data is preprocessed, including noise removal and filtering. During dynamic model identification, a suitable dynamic model structure is selected, including a Lagrange model or a Newton-Euler model.

[0048] The transfer function estimation is then performed, including defining input and output variables, frequency domain analysis, and transfer function fitting. When defining input and output variables, the specific meanings of the measured position and command position are clarified, and the corresponding measurement methods and sensors are determined. For example, the measured position can be obtained through an encoder mounted on the joint, and the command position can be the target position signal sent by the controller to the drive motor. During frequency domain analysis, Fourier transforms are performed on the acquired command position signal and measured position signal to convert them to the frequency domain. The amplitude ratio and phase difference of the command position signal and measured position signal at different frequencies are calculated. These amplitude ratios and phase differences constitute the amplitude-frequency characteristic and phase-frequency characteristic of the transfer function. Based on the amplitude-frequency characteristic and phase-frequency characteristic obtained from the frequency domain analysis, a suitable transfer function model structure is selected, and a curve fitting method (such as the least squares method) is used to fit the parameters of the transfer function model with the experimental data to obtain the transfer function between the measured position and the command position.

[0049] This embodiment obtains the dynamic model of the robotic arm through experiments and uses it to estimate the measurement position. and command location Transfer function between Compared with theoretical modeling, experimental methods can take into account various factors in the actual operation of the robotic arm, such as friction and gear backlash, making the transfer function more consistent with reality. This improves the accuracy of the closed-loop transfer function of the established physical human-machine interaction system, and can more realistically reflect the input-output relationship of the system, providing a more reliable basis for subsequent system analysis and control.

[0050] These implementation methods, along with the constructed human-computer interaction model omitting virtual stiffness and the obtained closed-loop transfer function, facilitate system stability analysis. The establishment of this model lays the foundation for subsequently constructing an adaptive controller based on admittance parameters using system instability indices. An accurate human-computer interaction model can more accurately reflect the system state, making the calculation of system instability indices more reliable. The adaptive controller constructed based on these indices can improve the effectiveness of adjusting admittance parameters to adapt to different working environments and task requirements, thereby improving the system's adaptability and control performance.

[0051] In some implementations, step S2 involves performing spectral analysis on the force signal to obtain system instability indicators, specifically including the following steps.

[0052] It should be noted that the force signal passes through a stability observer. The stability observer of the physical human-machine interaction system is designed according to requirements, and it monitors the external force applied by the operator. To monitor the external force applied by the operator, appropriate sensors need to be selected, such as force sensors and torque sensors. Force sensors directly measure the magnitude and direction of the force and are used to measure linear forces; torque sensors are used to measure rotational torque and are suitable for force measurement at joints. For example, installing a six-dimensional force sensor on the end effector of a robotic arm can simultaneously measure forces and torques in three directions.

[0053] Based on the system model and sensor measurement information, a suitable stability observer structure is selected. Observer structures include, but are not limited to, Kalman filters, sliding mode observers, and high-gain observers. Among these, Kalman filters are suitable for linear systems and can provide optimal estimation of the system state even in the presence of noise; sliding mode observers have strong robustness and are suitable for systems with uncertainties. Then, based on the system's dynamic characteristics and performance requirements, the observer parameters are designed, and the designed observer algorithm is implemented in a computer. The observer program can be written using programming languages ​​(such as Python, MATLAB, etc.) or processed in real time using an embedded system.

[0054] The stability observer is used to monitor the external force applied by the operator in real time. Based on the signals measured by the sensors and the system model, the observer can estimate the magnitude and direction of the external force applied by the operator.

[0055] S2.1, for A window of the latest force signal measurements is used to perform a Fast Fourier Transform calculation to obtain the frequency components of the force signal in the frequency domain. amplitude .

[0056] Using a force sensor mounted on the end effector of a robotic arm, at a fixed sampling period... Real-time measurement of force signals during human-computer interaction. For example, if the sampling period... Then the force signal value is recorded every 0.01s.

[0057] Select the latest force signal data sequence obtained from continuous measurements Each data point is treated as a window of data. For example, when When the time is right, take the 100 most recent force signal measurements. (Window data length) It will affect frequency resolution. The larger the value, the higher the frequency resolution, but the computational load will also increase accordingly.

[0058] Use the FFT algorithm to select The data in each window is processed, and the FFT sampling window period is [value missing]. Then the frequency resolution is The Fast Fourier Transform (FFT) is an efficient algorithm for calculating the Discrete Fourier Transform (DFT), which can convert a force signal in the time domain into a frequency domain representation. Through FFT calculation, the force signal at different frequency components can be obtained. Complex representation of .

[0059] Complex numbers obtained from FFT By taking the modulus, the frequency components of the force signal in the frequency domain can be obtained. amplitude .

[0060] S2.2, based on the amplitude of the frequency component Calculate the instability index of the calculation system.

[0061] S2.2.1, determine the frequency range boundary using the Nyquist-Shannon sampling theorem. .

[0062] According to the Nyquist-Shannon sampling theorem, in order to recover the original signal from the sampled signal without distortion, the sampling frequency... It should be at least twice the highest frequency of the original signal. Therefore, the frequency range boundary... For example, if the sampling period Then the sampling frequency Frequency range boundary .

[0063] S2.2.2, Determine the crossover frequency used to distinguish between stable and unstable motion. .

[0064] By analyzing experimental data, we can observe the distribution characteristics of the force signal frequency components under stable and unstable system states. For example, we can statistically analyze the frequency ranges corresponding to the occurrence of instability in the system under different operating conditions and select a suitable frequency value as the crossover frequency. Alternatively, a preliminary estimate can be made based on the system's dynamic characteristics and empirical knowledge, followed by adjustments and optimizations in practical applications.

[0065] S2.2.3, Determine the stability threshold .

[0066] Similarly, based on experimental data, the range of values ​​for system instability indices under stable and unstable states can be analyzed. This can be achieved by setting an appropriate threshold. The system's operating state can be divided into two categories: stable and unstable. For example, when the system instability index is less than... When the index is greater than a certain value, the system is considered to be in a stable state; when the index is greater than a certain value, the system is considered to be in a stable state. At that time, it was believed that the system might be unstable.

[0067] S2.2.4, the system instability index is calculated using the following formula. (6) in, It is the lowest frequency of the Fast Fourier Transform.

[0068] Formula (6) represents the sum of the amplitudes of the high-frequency components. The ratio to the sum of the amplitudes of all frequencies.

[0069] In these implementations, by calculating the amplitude of the frequency components of the force signal using FFT, the characteristics of the force signal can be analyzed in depth from a frequency domain perspective. Variations in different frequency components may correspond to different operating states of the system, and abnormal increases in high-frequency components are often related to unstable motion of the system. Therefore, the spectral analysis methods in these implementations can provide the accuracy to capture potential instability factors in the system, detect abnormal situations in advance, and provide timely and accurate information for subsequent control and adjustment.

[0070] In these implementations, the system instability index is a quantified value that integrates the amplitude information of the force signal at different frequencies. This index provides a clear numerical description of the system's stability, improving the objectivity and accuracy of system stability assessments and providing a clear reference for subsequent adaptive control.

[0071] In these implementations, the calculated system instability index is a key input for constructing an admittance parameter adaptive controller. Based on this index, the adaptive controller can adjust the admittance parameters (virtual inertia and virtual damping) in real time, enabling the system to quickly adapt to different working environments and task requirements, maintaining stable operation. These implementations, based on adaptive control methods using spectrum analysis and instability indices, improve the robustness and adaptability of the system, effectively avoiding instability phenomena such as system oscillations.

[0072] In some implementations, step S3 constructs an adaptive controller for admittance parameters using system instability indices, specifically including the following steps.

[0073] S3.1, the system instability index calculated by formula (6) Normalization is performed to obtain new system instability indices. .

[0074] First, determine the range of values ​​for the new system instability index after normalization. Normalization can be done to the interval [0, 1], which facilitates subsequent processing and comparison.

[0075] By using experimental data or long-term monitoring of the system's operating status, the system instability index calculated by formula (6) is statistically derived. maximum value and minimum value These extreme values ​​represent the boundary values ​​that the system's instability indicators may exhibit under different operating conditions.

[0076] The new system instability index is calculated using the following normalization formula. :

[0077] For each system instability index calculated by formula (6) Substituting into the above formula (6) for calculation, we obtain the normalized new index. .

[0078] Normalization unifies system instability indicators to a fixed range, eliminating differences in the dimensions and numerical ranges of indicators under different operating conditions. This ensures the comparability of instability indicators calculated under different scenarios and conditions, facilitating subsequent analysis and processing. Normalized indicators more intuitively reflect the degree of system instability, allowing adaptive controllers to adjust parameters based on these standardized indicators, improving the accuracy of adjustments. Simultaneously, it avoids the problem of controller sluggishness or over-adjustment caused by excessively large or small indicator numerical ranges, thus improving the controller's adaptability to different system states.

[0079] S3.2, Construct an adaptive controller for the admittance parameters according to the following formula. (7) in, , To maintain a stable minimum virtual inertia and virtual damping for physical human-computer interaction systems, , To maintain stable inertia and damping reference values ​​for physical human-computer interaction systems.

[0080] Determine the minimum virtual inertia for a physical human-computer interaction system to maintain stability through experiments or theoretical analysis. Minimum virtual damping and inertia reference value and damping reference value Specifically, the determination of these parameters takes into account the system's dynamic characteristics, task requirements, and stability performance during actual operation. For example, when the system is in a stable operating state, the corresponding virtual inertia and virtual damping values ​​can be recorded as a reference. , The physical human-computer interaction system is designed to maintain a critically stable value even when the operator's arm stiffness is at its maximum. It should be noted that... , .

[0081] When the robot is in an unstable state, fix Increasing the damping coefficient reduces the oscillation amplitude. However, simply increasing the damping coefficient does not decrease the oscillation frequency. (Fixed damping coefficient) ,Increase The oscillation frequency decreases, but the oscillation amplitude remains unchanged. Therefore, in these implementations, the adaptive law of the above formula (7) is designed to simultaneously adjust the admittance controller parameters so that the entire human-computer interaction system can maintain a balanced amplitude-frequency characteristic.

[0082] In these implementations, the constructed adaptive controller can adjust the virtual inertia and virtual damping in real time according to the system instability index. When the system instability increases (whether it is stable can be detected by Equation 6), the virtual inertia and virtual damping will increase accordingly to enhance the system stability; when the system tends to be stable, the parameters will be appropriately reduced to maintain good dynamic performance. The adaptive adjustment mechanism enables the physical human-computer interaction system to maintain stable operation under different working conditions and external disturbances.

[0083] These implementations introduce minimum virtual inertia and minimum virtual damping to ensure that the system has a certain stability foundation even when the system instability index is low. This avoids the risk of system instability degradation due to excessive parameter adjustments and provides a reliable guarantee for the stable operation of the system.

[0084] In these implementations, the inertia reference value and the damping reference value provide a reasonable upper limit for parameter adjustment. When the system is highly unstable, the parameters will be adjusted towards the reference value, but will not increase indefinitely, ensuring that the dynamic performance of the system is not affected by excessively large parameters, while also avoiding excessive oscillations during the parameter adjustment process.

[0085] In some implementations, an adaptive rate of formula (7) is designed to activate the physical human-computer interaction system in each sampling period. Internally, the force signal is continuously measured and its spectrum analyzed to calculate new system instability indices, which are then normalized to obtain... Based on the new normalized instability index, the updated virtual inertia is calculated again using formula (7). and virtual damping To determine if the updated system is stable, if... If the system is stable, then parameter adjustment is stopped; otherwise, let... Continue sampling and conduct the next round of indicator updates and parameter adjustments.

[0086] The foregoing has described in detail an embodiment of a robotic arm human-machine collaborative adaptive admittance control method. Based on the robotic arm human-machine collaborative adaptive admittance control method described in the above embodiment, this invention also provides a robotic arm human-machine collaborative adaptive admittance control device corresponding to the method.

[0087] Figure 2 This is a schematic block diagram of a robotic arm human-machine collaborative adaptive admittance control device provided in an embodiment of the present invention. In this embodiment, the robotic arm human-machine collaborative adaptive admittance control device 200 can be divided into multiple functional modules according to the functions it performs. A module, as referred to in this invention, is a series of computer program segments that can be executed by at least one processor and perform a fixed function, and is stored in memory.

[0088] The human-computer interaction model construction module 210 is used to construct a human-computer interaction model that omits virtual stiffness. The human-computer interaction model includes admittance parameters, which include virtual inertia and virtual damping.

[0089] The system instability index calculation module 220 is used to design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index.

[0090] The admittance parameter adaptive controller construction module 230 is used to construct an admittance parameter adaptive controller based on system instability indices. The admittance parameter adaptive controller includes an adaptive controller with virtual inertia and virtual damping.

[0091] The admittance parameter adjustment module 240 is used to adjust the virtual inertia and virtual damping based on the admittance parameter adaptive controller, so as to keep the physical human-computer interaction system in a stable state.

[0092] The robotic arm human-machine collaboration adaptive admittance control device of this embodiment is used to implement the aforementioned robotic arm human-machine collaboration adaptive admittance control method. Therefore, the specific implementation of this device can be found in the embodiment section of the robotic arm human-machine collaboration adaptive admittance control method above. Thus, its specific implementation can be referred to the description of the corresponding embodiments, and will not be elaborated here.

[0093] Furthermore, since the robotic arm human-machine collaboration adaptive admittance control device in this embodiment is used to implement the aforementioned robotic arm human-machine collaboration adaptive admittance control method, its function corresponds to the function of the above method, and will not be described again here.

[0094] Figure 3 A schematic diagram of a terminal 300 provided in an embodiment of the present invention includes: a processor 310, a memory 320, and a communication unit 330. The processor 310 is used to implement the following steps when executing the adaptive admittance control program for human-machine collaboration of the robotic arm stored in the memory 320: Construct a human-computer interaction model that omits virtual stiffness. The admittance parameters in the human-computer interaction model include virtual inertia and virtual damping. Design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index. Admittance parameter adaptive controllers are constructed using system instability indices. These adaptive controllers include virtual inertia adaptive controllers and virtual damping adaptive controllers. The virtual inertia and virtual damping are adjusted by an adaptive controller based on admittance parameters, so that the physical human-computer interaction system returns to a stable state.

[0095] The present invention also provides a computer storage medium, which may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0096] The computer storage medium stores a robotic arm human-machine collaborative adaptive admittance control program, which, when executed by a processor, performs the following steps: Construct a human-computer interaction model that omits virtual stiffness. The admittance parameters in the human-computer interaction model include virtual inertia and virtual damping. Design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index. Admittance parameter adaptive controllers are constructed using system instability indices. These adaptive controllers include virtual inertia adaptive controllers and virtual damping adaptive controllers. The virtual inertia and virtual damping are adjusted by an adaptive controller based on admittance parameters, so that the physical human-computer interaction system returns to a stable state.

[0097] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A human-machine collaborative adaptive admittance control method for a robotic arm, characterized in that, Includes the following steps: Construct a human-computer interaction model that omits virtual stiffness. The admittance parameters in the human-computer interaction model include virtual inertia and virtual damping. Design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index. Admittance parameter adaptive controllers are constructed using system instability indices. These adaptive controllers include virtual inertia adaptive controllers and virtual damping adaptive controllers. The virtual inertia and virtual damping are adjusted by an adaptive controller based on admittance parameters, so that the physical human-computer interaction system returns to a stable state.

2. The adaptive admittance control method for human-machine collaboration of a robotic arm according to claim 1, characterized in that, Constructing a human-computer interaction model that omits virtual stiffness specifically includes: Establish the admittance control equations for the end effector of the robotic arm, where contact occurs only. （1） in, and These represent the desired speed at the end effector of the robotic arm and the actual speed of the robotic arm, respectively. , , These represent virtual inertia, virtual damping, and virtual stiffness, respectively. Indicates the actual applied force; Omitting the virtual stiffness, the admittance governing equations are simplified to: （2） By transforming and discretizing equation (2), we obtain: （3） in, , , It is the velocity of the joint, obtained by differentiating the joint position. The acceleration of the joint is obtained by performing differentiation. , It is the Jacobian matrix of the robotic arm. Indicates the first One cycle, Indicates the duration of the sampling period; The human-computer interaction is modeled as a first-order system. The simplified human-computer interaction model represents the muscle characteristics of the human arm. The admittance control equation (2) is rewritten in the form of an admittance function. （4） in, Represent the complex variables in the Laplace transform; The closed-loop transfer function of a physical human-computer interaction system, with contact force as input and the measured pose of the robotic arm's end effector as output, is established as follows: （5） in, The stiffness of a human arm, This indicates the time delay of human-computer interaction response. Indicates measurement location and command location The transfer function between them.

3. The adaptive admittance control method for human-machine collaboration of a robotic arm according to claim 2, characterized in that, Constructing a human-computer interaction model that omits virtual stiffness also includes: A dynamic model of the robotic arm was obtained through experiments, and the measurement position was estimated using the dynamic model. and command location Transfer function between .

4. The adaptive admittance control method for human-machine collaboration of a robotic arm according to claim 2, characterized in that, Spectral analysis of force signals is performed to obtain system instability indices, specifically including: right A window of the latest force signal measurements is used to perform a Fast Fourier Transform calculation to obtain the frequency components of the force signal in the frequency domain. amplitude ; By the magnitude of the frequency component Calculate the instability index of the calculation system.

5. The adaptive admittance control method for human-machine collaboration of a robotic arm according to claim 4, characterized in that, By the magnitude of the frequency component The instability metrics of a computational system include: Determine the frequency range boundary using the Nyquist-Shannon sampling theorem. ; Determine the crossover frequency used to distinguish between stable and unstable motion. ; Determine the stability threshold ; The system instability index is calculated using the following formula. （6） in, It is the lowest frequency of the Fast Fourier Transform.

6. The adaptive admittance control method for human-machine collaboration of a robotic arm according to claim 5, characterized in that, Constructing an adaptive controller based on admittance parameters using system instability indices specifically includes: The system instability index calculated by formula (6) Normalization is performed to obtain new system instability indices. ; Construct an adaptive controller for the admittance parameters according to the following formula. （7） in, , To maintain a stable minimum virtual inertia and virtual damping for physical human-computer interaction systems, , To maintain stable inertia and damping reference values ​​for physical human-computer interaction systems.

7. The adaptive admittance control method for human-machine collaboration of a robotic arm according to claim 6, characterized in that, , The physical human-computer interaction system is designed to maintain a critically stable value even when the operator's arm stiffness is at its maximum. , .

8. A robotic arm human-machine collaborative adaptive admittance control device, characterized in that, include: The human-computer interaction model construction module is used to construct a human-computer interaction model that omits virtual stiffness. The human-computer interaction model includes admittance parameters, which include virtual inertia and virtual damping. The system instability index calculation module is used to design a stability observer for a physical human-computer interaction system. The stability observer monitors the external force applied by the operator and performs spectral analysis on the force signal to obtain the system instability index. The admittance parameter adaptive controller construction module is used to construct an admittance parameter adaptive controller based on system instability indices. The admittance parameter adaptive controller includes an adaptive controller with virtual inertia and virtual damping. The admittance parameter adjustment module is used to adjust the virtual inertia and virtual damping based on the admittance parameter adaptive controller, so as to keep the physical human-computer interaction system in a stable state.

9. A terminal, characterized in that, include: The memory is used to store the adaptive admittance control program for human-robot collaboration of the robotic arm. A processor is configured to implement the steps of the adaptive admittance control method for human-machine collaboration of the robotic arm as described in any one of claims 1-7 when executing the robotic arm human-machine collaboration adaptive admittance control program.

10. A computer-readable storage medium, characterized in that, The readable storage medium stores a robotic arm human-machine collaborative adaptive admittance control program, which, when executed by a processor, implements the steps of the robotic arm human-machine collaborative adaptive admittance control method as described in any one of claims 1-7.

Images