A mechanical arm compliance control method based on sensing monitoring

Abstract

The present application relates to the technical field of mechanical arm control, and particularly relates to a kind of mechanical arm compliance control method based on sensing monitoring.The method comprises: obtaining the relative load degree characteristic value of a monitoring position at a moment;Obtain the relative rigidity degree estimation coefficient of each moment of each monitoring position, and then obtain the rigidity estimation credibility of each moment of each monitoring position by combining motor speed;Obtain the final decomposition weight of each moment of a monitoring position, and then decompose the relative rigidity degree estimation coefficient of each moment of the monitoring position based on STL algorithm to obtain trend term and periodic term;Then predict the relative rigidity degree estimation coefficient prediction value of next moment;Based on the relative rigidity degree estimation coefficient prediction value of next moment of a monitoring position, the relative rigidity degree estimation coefficient of each historical moment and the relative rigidity degree estimation coefficient prediction value, the predicted rigidity matrix adjustment coefficient is obtained, and the mechanical arm is controlled.The present application can effectively control the mechanical arm.

Description

Technical Field

[0001] This invention relates to the field of robotic arm control technology, and more specifically to a compliant control method for robotic arms based on sensor monitoring. Background Technology

[0002] With the continuous improvement of the intelligence and automation level of power systems, the demand for robotic arms in power operation and maintenance is increasing. Especially in high-risk, unstructured work scenarios such as substations, robotic arms can replace manual labor in tasks such as inspection, maintenance, and switch operations, effectively reducing labor costs and safety risks. However, under conditions of diverse equipment types, confined spaces, and sudden environmental changes, traditional robotic arms often lack compliance and adaptability, easily leading to insufficient operational accuracy or even collision risks. In recent years, with the introduction of sensor technology and intelligent algorithms, the perception and control performance of robotic arms has improved, but the efficiency of compliant control strategies during operation remains limited, making it difficult to simultaneously ensure safety and high operational precision and flexibility. Therefore, achieving compliant control of robotic arms in power equipment operation and maintenance, and improving their stability and accuracy under complex tasks, plays a crucial role in the high-precision operation of power equipment and the efficient completion of complex tasks.

[0003] In compliant control of robotic arms, existing variable impedance control (VIC) methods often rely on static or preset parameters when adjusting the stiffness matrix, resulting in insufficient response to changes in environmental stiffness. For example, when a robotic arm contacts switches or cables of different stiffness during substation operation, a fixed stiffness setting may cause excessive impact on hard switches, while the action on flexible cables may be insufficient. The influence of sensor accuracy and the robotic arm's own dynamic interference can lead to inaccurate or delayed stiffness adjustment, resulting in poor adaptability of the robotic arm to nonlinear or dynamic environmental characteristics, making it difficult to ensure safe and efficient compliant operation of the robotic arm under complex tasks. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide a compliant control method for a robotic arm based on sensor monitoring. The specific technical solution adopted is as follows:

[0005] One embodiment of the present invention provides a compliant control method for a robotic arm based on sensor monitoring, the method comprising:

[0006] The system collects the input current, loop current, input voltage, loop voltage, and motor speed of the joint motors at each monitoring position of the robotic arm at each moment; and obtains the relative load level characteristic value at a given moment based on the input current, loop current, input voltage, and loop voltage of a monitoring position at a given moment.

[0007] The relative stiffness estimation coefficient at a given moment is obtained based on the relative load characteristic value of a monitoring location at a given moment and the previous moment, and the motor speed; the stiffness estimation reliability at a given moment is obtained based on the relative stiffness estimation coefficients of each neighboring moment of a monitoring location at a given moment and the motor speed.

[0008] The final decomposition weights at each time point are obtained based on the reliability of the rigidity estimate at a monitoring location and the initial decomposition weights of the STL algorithm. The trend term and periodic term are obtained by combining the final decomposition weights at each time point of a monitoring location with the relative rigidity estimation coefficients at each time point of the monitoring location based on the STL algorithm.

[0009] The relative stiffness estimation coefficient for the next moment is predicted based on the trend and periodic terms corresponding to a monitoring position; the stiffness matrix adjustment coefficient for the next moment is obtained based on the relative stiffness estimation coefficient for the next moment at a monitoring position, the relative stiffness estimation coefficients for each historical moment, and the predicted relative stiffness estimation coefficient; the robotic arm is controlled by combining the stiffness matrix adjustment coefficients for the next moment at each monitoring position with a variable impedance control strategy.

[0010] Preferably, the relative load level characteristic value at a given moment is obtained based on the input current, loop current, input voltage, and loop voltage at a monitoring location, including:

[0011] The ratio of the loop current to the input current at a monitoring location at a given moment is recorded as the first ratio; the ratio of the input voltage to the loop voltage at the same monitoring location at that moment is recorded as the second ratio; the first ratio and the second ratio are multiplied together to obtain the relative load characteristic value of the monitoring location at that moment.

[0012] Preferably, the relative stiffness estimation coefficient at a given moment is obtained based on the relative load characteristic value at a monitoring location at a given moment and the moment preceding that moment, and the motor speed, including:

[0013] The difference between the relative load characteristic values ​​at a monitoring location at a given time and the time preceding that time is negatively correlated using an exponential function with a base of the natural constant to obtain a first mapped value at that time. The difference between the motor speed at the monitoring location at that time and the time preceding that time is negatively correlated using an exponential function with a base of the natural constant to obtain a second mapped value. The first mapped value and the second mapped value are compared to obtain the relative stiffness estimation coefficient at the monitoring location at that time.

[0014] Preferably, the reliability of the rigidity estimate at a given moment is obtained based on the relative rigidity estimation coefficients of each neighboring moment at a given monitoring location and the motor speed, including:

[0015] Starting from a given moment, a predetermined number of moments are taken before and after that moment, and these moments are recorded as the neighborhood moments of that moment. A negative correlation mapping is performed on the time interval between a neighborhood moment and the given moment using an exponential function with a base of the natural constant to obtain the time distance weight of that neighborhood moment. For a monitoring location, the absolute value of the difference between the relative rigidity estimation coefficients of that moment and its neighboring moments is weighted and averaged using the time distance weights of the neighborhood moments of that monitoring location to obtain the first average difference. The absolute value of the difference between the motor speeds of that moment and its neighboring moments is weighted and averaged using the time distance weights of the neighborhood moments of that monitoring location to obtain the second average difference. A negative correlation mapping is performed on the ratio of the first average difference to the product of the second average difference and the motor speed at that moment using an exponential function with a base of the natural constant to obtain the rigidity estimation reliability of that monitoring location at that moment.

[0016] Preferably, the final decomposition weights for each time step are obtained based on the rigidity estimation reliability of a monitoring location at each time step and the initial decomposition weights of the STL algorithm, including:

[0017] The reliability of the rigidity estimate at a monitoring location at a given time is compared with the sum of the reliability of the rigidity estimates at all times within the local range when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm. This sum is then multiplied by the initial decomposition weights when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm to obtain the weight factor at that time. The final decomposition weight at that monitoring location at that time is then compared with the sum of the weight factors at all times within the local range when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm to obtain the final decomposition weight at that monitoring location at that time.

[0018] Preferably, the predicted value of the relative stiffness estimation coefficient for the next moment is predicted based on the trend term and periodic term corresponding to a monitoring location, including:

[0019] The GRU algorithm is used to predict the trend and periodic terms corresponding to a monitoring location to obtain the predicted trend and periodic values ​​for the next moment. The predicted trend and periodic values ​​for the next moment are added together to obtain the predicted relative stiffness estimation coefficient for the next moment.

[0020] Preferably, the stiffness matrix adjustment coefficient for the next moment is obtained based on the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment, the relative stiffness estimation coefficients at each historical moment, and the predicted value of the relative stiffness estimation coefficient, including:

[0021] If the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment is greater than or equal to the estimated value of the relative stiffness at the current moment; the time distance weight of the historical moment is obtained by negatively mapping the time interval between the historical moment and the current moment using an exponential function with the natural constant as the base; the third average difference is obtained by weighting and averaging the absolute values ​​of the differences between the predicted value of the relative stiffness estimation coefficient at each historical moment and the relative stiffness estimation coefficient based on the time distance weight of each historical moment, and the third average difference is obtained by negatively mapping the third average difference using an exponential function with the natural constant as the base; the relative change degree is obtained by subtracting the ratio of the predicted value of the relative stiffness estimation coefficient at the next moment to the relative stiffness estimation coefficient at the current moment from the first preset value and taking the absolute value; the adjustment range is obtained by multiplying the stiffness matrix adjustment coefficient at the current moment, the relative change degree, and the third mapping value; the stiffness matrix adjustment coefficient at the current moment is added to the adjustment range to obtain the stiffness matrix adjustment coefficient at the next moment.

[0022] If the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment is less than the relative stiffness estimation coefficient at the current moment, the stiffness matrix adjustment coefficient at the current moment is subtracted from the adjustment amplitude to obtain the stiffness matrix adjustment coefficient at the next moment.

[0023] The embodiments of the present invention have at least the following beneficial effects: This application collects the input current, loop current, input voltage, loop voltage, and motor speed of the joint motor at each monitoring location at various times using sensors. Then, it analyzes the input current, loop current, input voltage, and loop voltage at a monitoring location at a given time to obtain the relative load characteristic value at that time. Next, it combines the relative load characteristic value from the previous time and the motor speed to obtain the relative stiffness estimation coefficient for that time. Then, based on the relative stiffness estimation coefficients of the neighboring times at a monitoring location at a given time and the motor speed, it obtains the stiffness estimation reliability for that time and obtains the final decomposition weights for each time at that monitoring location. The final decomposition weights at each monitoring location are based on the STL algorithm to decompose the relative stiffness estimation coefficients at each monitoring location into trend and periodic terms, thereby predicting the relative stiffness estimation coefficients for the next monitoring location. Finally, based on the predicted relative stiffness estimation coefficients for the next monitoring location, the relative stiffness estimation coefficients for each historical time, and the predicted relative stiffness estimation coefficients, the stiffness matrix adjustment coefficients for the next monitoring location are obtained. The robotic arm is then controlled using a variable impedance control strategy based on the stiffness matrix adjustment coefficients for each monitoring location at the next monitoring location. This effectively controls the robotic arm, facilitating the adjustment of stiffness and flexibility at each monitoring location, making its application more stable and reliable. Attached Figure Description

[0024] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0025] Figure 1 This is a flowchart illustrating a compliant control method for a robotic arm based on sensor monitoring, provided as an embodiment of the present invention. Detailed Implementation

[0026] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a sensor-based robotic arm compliant control method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0027] 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.

[0028] The following description, in conjunction with the accompanying drawings, details a specific scheme for a sensor-based compliant control method for a robotic arm provided by the present invention.

[0029] Example:

[0030] The main application scenario of this invention is: using sensors to collect data on the position of a robotic arm, then analyzing the data, obtaining the stiffness matrix adjustment coefficient for future moments based on the analysis results, and then performing compliant control on the robotic arm.

[0031] Please see Figure 1 The diagram illustrates a flowchart of a compliant control method for a robotic arm based on sensor monitoring, provided by an embodiment of the present invention. The method includes the following steps:

[0032] Step S1: Collect the input current, loop current, input voltage, loop voltage, and motor speed of the joint motors at each monitoring position of the robotic arm at each moment; obtain the relative load characteristic value at that moment based on the input current, loop current, input voltage, and loop voltage of a monitoring position at a given moment.

[0033] This invention primarily focuses on effectively controlling the compliance of robotic arms in application scenarios to improve operational stability. To achieve this goal, it is necessary to acquire the joints and key positions of the robotic arm as monitoring locations to reflect changes in its movement.

[0034] Furthermore, sensors are deployed at each monitoring location to collect the input current, loop current, input voltage, loop voltage, and motor speed of the joint motor at each monitoring location at any given time. At the same time, these data are standardized and dimensionless to facilitate subsequent analysis.

[0035] In power equipment maintenance scenarios, robotic arms are mainly used for operations such as switch control. The output of their joint motors typically employs variable impedance control (VIC) to achieve stiffness and flexibility adjustment. However, fixed or preset stiffness coefficient matrices often fail to meet requirements under varying operating environments, maintenance objects, and the robotic arm's own real-time operating conditions. Without an adaptive adjustment mechanism, the robotic arm may experience excessive impact when operating rigid equipment or perform poorly when operating flexible equipment. Therefore, it is necessary to adaptively adjust the stiffness coefficient matrix based on the dynamic characteristics of the working environment and operating status to achieve efficient, safe, and compliant control of the robotic arm in complex tasks.

[0036] Therefore, the first step is to use the collected data to obtain the relative load characteristic value at any monitoring location at any time. Specifically, the ratio of the loop current to the input current at a monitoring location at a given time is recorded as the first ratio; the ratio of the input voltage to the loop voltage at the same monitoring location at that time is recorded as the second ratio; and the first and second ratios are multiplied to obtain the relative load characteristic value at that monitoring location at that time.

[0037] The specific calculation model for the relative load level characteristic value is as follows:

[0038] ,

[0039] Where H represents the relative load level characteristic value of a monitoring location at a given time; This indicates the measured value of the current in the motor circuit at that monitoring location and at that moment, which is also the circuit current; This indicates the input value of the motor current at that monitoring location at that moment, which is also known as the input current. This indicates the motor voltage input value at this monitoring location at this moment; This indicates the measured voltage value in the motor circuit at this monitoring location and at this moment; The first ratio represents the ratio of the measured current in the motor circuit at the monitoring location at the current moment to the current input value of the motor at that moment. The larger this value is, the greater the load at that location may be. However, it needs to be analyzed in conjunction with the input voltage at that moment. The second ratio represents the ratio of the measured voltage in the motor circuit at the monitoring location at the current moment to the motor voltage input at that moment, thus obtaining the voltage drop ratio at this moment. If the voltage drop is greater at this moment, it indicates that the motor is under greater load.

[0040] Based on the above, the relative load level characteristic value at any monitoring position of the robotic arm at any time can be obtained. Therefore, during its operation, a monitoring position can obtain a sequence of relative load level characteristic values ​​in time sequence.

[0041] Step S2: Obtain the relative stiffness estimation coefficient at a given moment based on the relative load characteristic value of a monitoring location at a given moment and the previous moment, and the motor speed; obtain the stiffness estimation reliability at a given moment based on the relative stiffness estimation coefficients of each neighboring moment of a monitoring location at a given moment and the motor speed.

[0042] After obtaining the relative load characteristic value sequence at any given time using the above method, if the load characteristic value at a certain monitoring position increases significantly in adjacent time intervals, while the motor speed at that position changes only slightly, it indicates that the robotic arm is under a large load at that position, but the speed has not yet responded significantly, reflecting high rigidity of the robotic arm. Based on this, estimating the relative rigidity coefficient at each monitoring position at each time interval allows for dynamic evaluation of the rigidity characteristics of the robotic arm under different operating conditions, providing a reliable reference for adjusting compliant control strategies and monitoring operational safety.

[0043] Therefore, the relative stiffness estimation coefficient at a given moment is obtained based on the relative load characteristic value and motor speed at a monitoring location at a given moment and the moment before that moment.

[0044] Specifically, a first mapping value is obtained by negatively mapping the difference between the relative load characteristic values ​​at a monitoring location at a given time and the time preceding that time using an exponential function with a base of the natural constant; a second mapping value is obtained by negatively mapping the difference between the motor speed at the monitoring location at that time and the time preceding that time using an exponential function with a base of the natural constant; and the relative stiffness estimation coefficient at the monitoring location at that time is obtained by comparing the first mapping value and the second mapping value.

[0045] The specific calculation model for the relative stiffness estimation coefficient is as follows:

[0046] ,

[0047] in, This represents the relative stiffness estimation coefficient at a monitoring location at time t; e represents the natural constant, and the purpose of using e for mapping is to prevent each part from being 0; and These represent the relative load characteristic values ​​at time t and time t-1 at the monitoring location, respectively. , These represent the motor speeds at time t and t-1 at the monitoring location, respectively. The first mapping value represents the amplification factor of the relative load characteristic value at time t and time t-1 at the monitoring location. The larger the value, the greater the increase in load. This represents the increase factor of the motor speed at time t and time t-1 at the monitoring location. The larger the value, the greater the increase in motor speed. This represents the ratio of the increase factor of the relative load characteristic value at time t and time t-1 at the monitoring location to the increase factor of the motor speed. The larger this value is, the greater the increase in load characteristics and the smaller the increase in speed, which indicates that the relative rigidity is greater at this time.

[0048] The model described above calculates the relative stiffness estimation coefficients for different monitoring positions and times of the robotic arm. However, due to inertial effects during robotic arm movement and dynamic coupling between joints, the load change of a single joint may be affected by the movement of other joints, leading to fluctuations or deviations in the relative stiffness estimation coefficients. To improve control stability, different decomposition weights can be assigned to the relative stiffness estimation coefficients at any monitoring position and at any time. The STL (Seasonal-Trend decomposition using Loess) method is then used to decompose the sequence of relative stiffness estimation coefficients at each time corresponding to a monitoring position, extracting more accurate trend and periodic terms. Subsequently, a prediction algorithm can be applied to the decomposed data to predict the stiffness in real time, and the stiffness coefficient matrix in the variable impedance control (VIC) can be dynamically adjusted based on the prediction results, thereby ensuring the compliance and control stability of the robotic arm in complex operating environments.

[0049] Within a local range, if the relative stiffness estimation coefficient G at a certain moment shows a significant abrupt change compared to the neighborhood, while the speed difference with the neighborhood is small and the speed itself is low, it indicates that the robot arm's response to load changes is lagging and may be affected by external disturbances or joint coupling. In this case, the stiffness estimation value may not accurately reflect the actual stiffness change of the robot arm, and should be given a lower weight in the decomposition and weight allocation to avoid misjudging the stiffness change.

[0050] Therefore, the reliability of the rigidity estimate at a given moment is obtained by estimating the relative rigidity of each neighboring moment at a given monitoring location and the motor speed, which reflects the possible degree of change in the actual rigidity.

[0051] Specifically, starting from a given moment, a predetermined number of moments are taken before and after that moment, and these moments are recorded as the neighborhood moments of that moment. A negative correlation mapping is performed on the time interval between a neighborhood moment and the given moment using an exponential function with a base of the natural constant to obtain the time distance weight of that neighborhood moment. For a monitoring location, the absolute value of the difference between the relative rigidity estimation coefficients of that moment and its neighboring moments is weighted and averaged using the time distance weights of the neighborhood moments of that monitoring location to obtain the first average difference. The absolute value of the difference between the motor speeds of that moment and its neighboring moments is weighted and averaged using the time distance weights of the neighborhood moments of that monitoring location to obtain the second average difference. Finally, a negative correlation mapping is performed on the ratio of the first average difference to the product of the second average difference and the motor speed at that moment using an exponential function with a base of the natural constant to obtain the rigidity estimation reliability of that monitoring location at that moment.

[0052] The specific calculation model for the reliability of rigid estimation is as follows:

[0053] ,

[0054] in, The represents the reliability of the rigidity estimate at a monitoring location at time t, indicating the possibility that the relative rigidity estimate coefficient at any monitoring location at any time may be the true rigidity change; exp represents the exponential function with the natural constant e as the base; n represents the number of neighboring times at time t. Preferably, the reference value in this application is 8, that is, a preset number of times are taken before time t and a preset number of times are taken after time t, for a total of 8. The preset number is 4. When obtaining neighboring times, if the number of times before or after a time is insufficient, all times can be obtained. This represents the coefficient representing the relative stiffness estimate at monitoring location t at time t. This represents the relative stiffness estimation coefficient of the s-th neighboring time in the neighborhood of time t. This represents the time interval between the s-th neighboring time at time t and time t itself. The smaller this value, the better. The larger, This represents the temporal distance weight of the s-th neighboring time step; , These represent the motor speed at time t at the monitoring location and the motor speed at the s-th neighboring time, respectively.

[0055] The first average difference represents the weighted average of the absolute values ​​of the differences between the relative stiffness estimation coefficient at time t at the monitoring location and the relative stiffness estimation coefficient at each neighboring time. In other words, it represents the difference characteristics between the relative stiffness estimation coefficient at that time and the relative stiffness estimation coefficients at each neighboring time within its neighborhood.

[0056] The second average difference represents the weighted average of the absolute values ​​of the differences between the motor speed at time t at the monitoring location and the motor speed at each neighboring time within that neighborhood. In other words, it represents the difference characteristic between the motor speed at time t and the motor speed at each neighboring time. The smaller this value, and the smaller the motor speed at that time, the better. The smaller the value, the greater the difference between the rigidity and the neighborhood, i.e. The larger the value, the greater the probability that it is caused by random disturbances, and the smaller the actual probability. Therefore, its weight will be smaller in the subsequent decomposition process, and vice versa.

[0057] Therefore, the reliability of the rigid estimate for each monitoring location at each time point can be obtained.

[0058] Step S3: Obtain the final decomposition weights for each moment based on the reliability of the rigidity estimate at each moment of a monitoring location and the initial decomposition weights of the STL algorithm; combine the final decomposition weights for each moment of a monitoring location with the relative rigidity estimation coefficients at each moment of the monitoring location based on the STL algorithm to obtain the trend term and the periodic term.

[0059] The above steps can yield the rigid estimate reliability for each monitoring location at each time point. Then, by combining the local weighted decomposition in STL, the weights assigned to the local data are corrected, resulting in a more effective final decomposition weight.

[0060] Therefore, the final decomposition weights at each time point are obtained based on the rigidity estimation reliability of a monitoring location at each time point and the initial decomposition weights of the STL algorithm.

[0061] Specifically, the reliability of the rigidity estimate at a monitoring location at a given time is compared with the sum of the reliability of the rigidity estimates at all times within the local range when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm. This sum is then multiplied by the initial decomposition weights when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm to obtain the weight factor at that time. Finally, the weight factor at a monitoring location at that time is compared with the sum of the weight factors at all times within the local range when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm to obtain the final decomposition weight at that monitoring location at that time.

[0062] The specific calculation model for the final decomposition weights at a monitoring location at a given time is as follows:

[0063] ,

[0064] In the formula The relative stiffness estimation coefficient at time t at a monitoring location represents the final decomposition weight when the STL algorithm is used for local weighted decomposition. This represents the initial weights (i.e., the weights assigned by the STL algorithm) when performing local weighted decomposition of the stiffness estimation coefficients at time t at a monitoring location. This represents the reliability of a rigid estimate at time t at a given monitoring location. This represents the number of time intervals contained within the local range when the STL algorithm estimates the relative stiffness at time t using coefficient decomposition. This represents the reliability of the rigid estimate at time a within the local range corresponding to time t. This represents the initial decomposition weight at time a within the local range corresponding to time t; The ratio of the reliability of the rigidity estimate at monitoring location t at time t to the sum of the reliability of the rigidity estimates at all times within the local range corresponding to that time represents the probability weight of the relative rigidity estimation coefficient at monitoring location t as the degree of authenticity. The larger this value is, the higher its final decomposition weight. This is the weighting factor at time t for the monitoring location.

[0065] The above method yields the final decomposition weights when performing local weighted decomposition on the relative stiffness estimation coefficients at each time point at a monitoring location. Then, by combining the final decomposition weights at each time point, the STL algorithm is used to decompose the relative stiffness estimation coefficients at all times at that monitoring location, resulting in trend and periodic data.

[0066] Step S4: Based on the trend term and period term corresponding to a monitoring position, predict the relative stiffness estimation coefficient for the next moment; based on the predicted relative stiffness estimation coefficient for the next moment at a monitoring position, the relative stiffness estimation coefficients for each historical moment, and the predicted relative stiffness estimation coefficient, obtain the stiffness matrix adjustment coefficient for the next moment; based on the stiffness matrix adjustment coefficients for each monitoring position for the next moment, combine with the variable impedance control strategy to control the robotic arm.

[0067] The above describes obtaining the trend and periodic terms corresponding to a monitoring location. Further analysis of these two terms yields the predicted value of the relative stiffness estimation coefficient for the next moment. Specifically, the GRU algorithm is used to predict the trend and periodic terms corresponding to a monitoring location to obtain the predicted trend and periodic values ​​for the next moment. The predicted trend and periodic values ​​for the next moment are then added together to obtain the predicted value of the relative stiffness estimation coefficient for the next moment. It should be noted that the predicted relative stiffness estimation coefficient is the same physical quantity as the relative stiffness estimation coefficient calculated above.

[0068] Furthermore, by comparing the predicted value of the relative stiffness estimation coefficient at the next moment with the relative stiffness estimation coefficient at the current moment, the adjustment coefficient of the stiffness matrix is ​​adaptively adjusted, thereby achieving adaptive control of the rigidity and flexibility of the robotic arm.

[0069] In current applications of robotic arms, if the relative stiffness estimation coefficient for the next moment, predicted through feature sequences, increases compared to the current moment, it typically indicates that the environment or the manipulated object exhibits higher resistance to deformation, i.e., it becomes "harder." To ensure stability and compliance during the robot's interaction with the environment, the stiffness matrix adjustment coefficient should be increased accordingly, allowing the robot to output greater resistance to match the high stiffness of the environment, thereby avoiding positional errors or control deviations caused by excessive compliance. Conversely, if the predicted relative stiffness estimation coefficient decreases, it indicates that the environment becomes "softer," requiring a reduction in the stiffness matrix adjustment coefficient to prevent the system from becoming too stiff and generating oscillations or unnecessary forces. For example, in a robot grinding a workpiece, when it is predicted that the workpiece surface material will gradually harden, the controller will increase the stiffness adjustment coefficient to maintain a stable contact force for the grinding head; while when the workpiece edge area is softer, the stiffness adjustment coefficient will be reduced so that the end effector can adapt to surface deformation without damaging the workpiece.

[0070] Therefore, the stiffness matrix adjustment coefficient for the next moment is obtained based on the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment, the relative stiffness estimation coefficient at each historical moment, and the predicted value of the relative stiffness estimation coefficient.

[0071] Specifically, if the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment is greater than or equal to the estimated value of the relative stiffness at the current moment; the time distance weight of the historical moment is obtained by negatively mapping the time interval between the historical moment and the current moment using an exponential function with the natural constant as the base; the absolute value of the difference between the predicted value of the relative stiffness estimation coefficient at each historical moment and the relative stiffness estimation coefficient at each historical moment is weighted and averaged to obtain the third average difference, and the third average difference is obtained by negatively mapping the third average difference using an exponential function with the natural constant as the base; the relative change degree is obtained by subtracting the ratio of the predicted value of the relative stiffness estimation coefficient at the next moment to the relative stiffness estimation coefficient at the current moment from the first preset value and taking the absolute value; the adjustment range is obtained by multiplying the stiffness matrix adjustment coefficient at the current moment, the relative change degree, and the third mapping value; the stiffness matrix adjustment coefficient at the current moment is added to the adjustment range to obtain the stiffness matrix adjustment coefficient at the next moment.

[0072] If the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment is less than the relative stiffness estimation coefficient at the current moment, the stiffness matrix adjustment coefficient at the current moment is subtracted from the adjustment amplitude to obtain the stiffness matrix adjustment coefficient at the next moment.

[0073] The specific calculation model for the stiffness matrix adjustment coefficient at the next moment is as follows:

[0074] ,

[0075] ,

[0076] in, This represents the stiffness matrix adjustment coefficient at the next moment for a monitoring location. This represents the stiffness matrix adjustment coefficient at the monitoring location at the next moment before the previous moment; exp() represents an exponential function with the natural constant as the base. This indicates the number of historical moments preceding the next moment, including the current moment. , These represent the predicted value of the relative stiffness estimation coefficient at the next moment from the current moment at the monitoring location and the estimated value of the relative stiffness estimation coefficient at the current moment (excluding residuals, i.e., the sum of the periodic term and the trend term). The relative degree of change is represented by the absolute value of the difference between the predicted value of the relative stiffness estimation coefficient at the current time and the estimated value of the relative stiffness estimation coefficient at the current time at the monitoring location and 1. This value indicates the relative degree of change between the two at these two times. The larger this value is, the greater the degree of adjustment of the stiffness matrix adjustment coefficient at the next time.

[0077] , Let and represent the predicted value and the estimated coefficient of relative stiffness at the ns-th historical moment at the monitoring location, respectively; e represents the natural constant. This represents the time interval between the nth historical moment (the current moment) and the nsth historical moment. Because there is a certain error between the predicted value and the actual value, by calculating the magnitude of the error between the historical data and the predicted value over time, we can measure the error of the predicted value at the current moment and use it as a proportional coefficient for adjusting the stiffness. This avoids the impact of prediction error on the stability of the robotic arm by making hasty or large-scale adjustments to the stiffness matrix coefficients. The third average difference represents the weighted average of the current prediction error as it changes over time. A smaller value indicates a potentially smaller prediction error, allowing for a higher adjustment ratio in the stiffness matrix; conversely, a smaller adjustment ratio indicates a lower error. This is the third mapping value;

[0078] It should be noted that the relative stiffness estimation coefficients for each historical moment, i.e., the actual values ​​of the relative stiffness estimation coefficients for each historical moment, need to be obtained by performing STL decomposition on the relative stiffness estimation coefficients obtained from the above calculation. Then, the trend term and period term corresponding to each historical moment are added together to obtain the relative stiffness estimation coefficients for each historical moment in the formula. This is because the predicted value of the relative stiffness estimation coefficient for the next moment is obtained by adding the predicted value of the trend and the predicted value of the period for the next moment. This ensures that the comparison dimension is consistent and reflects the difference between the prediction and the actual value. The residual term is discarded during the prediction in order to avoid the fluctuation of the adjustment and improve the smoothness and stability of the robotic arm's operation.

[0079] After obtaining the stiffness matrix adjustment coefficients for the next moment at each monitoring position, these coefficients can be applied in real time to the variable impedance control (VIC) strategy of each joint of the robotic arm to dynamically adjust the joint stiffness. This allows the robotic arm to adaptively change its stiffness and flexibility when contacting different work objects or being subjected to external disturbances. Simultaneously, by continuously monitoring signals such as current, voltage, and joint speed through sensors, the stiffness estimation coefficients are updated in real time. Combined with a predictive model, the stiffness matrix for the next moment is continuously adjusted, thereby achieving closed-loop control of the robotic arm's stiffness and flexibility. This ensures that it possesses sufficient flexibility to protect the work object while maintaining operational accuracy and motion stability in complex working environments.

[0080] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0081] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0082] 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 within the protection scope of the present invention.

Claims

1. A compliant control method for a robotic arm based on sensor monitoring, characterized in that, The method includes: The system collects the input current, loop current, input voltage, loop voltage, and motor speed of the joint motors at each monitoring position of the robotic arm at each moment; and obtains the relative load level characteristic value at a given moment based on the input current, loop current, input voltage, and loop voltage of a monitoring position at a given moment. The relative stiffness estimation coefficient at a given moment is obtained based on the relative load characteristic value of a monitoring location at a given moment and the previous moment, and the motor speed; the stiffness estimation reliability at a given moment is obtained based on the relative stiffness estimation coefficients of each neighboring moment of a monitoring location at a given moment and the motor speed. The final decomposition weights at each time point are obtained based on the reliability of the rigidity estimate at a monitoring location and the initial decomposition weights of the STL algorithm. The trend term and periodic term are obtained by combining the final decomposition weights at each time point of a monitoring location with the relative rigidity estimation coefficients at each time point of the monitoring location based on the STL algorithm. The relative stiffness estimation coefficient for the next moment is predicted based on the trend and periodic terms corresponding to a monitoring position; the stiffness matrix adjustment coefficient for the next moment is obtained based on the relative stiffness estimation coefficient for the next moment at a monitoring position, the relative stiffness estimation coefficients for each historical moment, and the predicted relative stiffness estimation coefficient; the robotic arm is controlled by combining the stiffness matrix adjustment coefficients for the next moment at each monitoring position with a variable impedance control strategy.

2. The compliant control method for a robotic arm based on sensor monitoring according to claim 1, characterized in that, The process of obtaining the relative load level characteristic value at a given moment based on the input current, loop current, input voltage, and loop voltage at a monitoring location includes: The ratio of the loop current to the input current at a monitoring location at a given moment is recorded as the first ratio; the ratio of the input voltage to the loop voltage at the same monitoring location at that moment is recorded as the second ratio; the first ratio and the second ratio are multiplied together to obtain the relative load characteristic value of the monitoring location at that moment.

3. The compliant control method for a robotic arm based on sensor monitoring according to claim 1, characterized in that, The process of obtaining the relative stiffness estimation coefficient at a given moment based on the relative load characteristic value at a monitoring location and the previous moment, and the motor speed, includes: The difference between the relative load characteristic values ​​at a monitoring location at a given time and the time preceding that time is negatively correlated using an exponential function with a base of the natural constant to obtain a first mapped value at that time. The difference between the motor speed at the monitoring location at that time and the time preceding that time is negatively correlated using an exponential function with a base of the natural constant to obtain a second mapped value. The first mapped value and the second mapped value are compared to obtain the relative stiffness estimation coefficient at the monitoring location at that time.

4. The compliant control method for a robotic arm based on sensor monitoring according to claim 1, characterized in that, The process of obtaining the reliability of the rigidity estimate at a given moment based on the relative rigidity estimation coefficients of each neighboring moment at a given monitoring location and the motor speed includes: Starting from a given moment, a predetermined number of moments are taken before and after that moment, and these moments are recorded as the neighborhood moments of that moment. A negative correlation mapping is performed on the time interval between a neighborhood moment and the given moment using an exponential function with a base of the natural constant to obtain the time distance weight of that neighborhood moment. For a monitoring location, the absolute value of the difference between the relative rigidity estimation coefficients of that moment and its neighboring moments is weighted and averaged using the time distance weights of the neighborhood moments of that monitoring location to obtain the first average difference. The absolute value of the difference between the motor speeds of that moment and its neighboring moments is weighted and averaged using the time distance weights of the neighborhood moments of that monitoring location to obtain the second average difference. A negative correlation mapping is performed on the ratio of the first average difference to the product of the second average difference and the motor speed at that moment using an exponential function with a base of the natural constant to obtain the rigidity estimation reliability of that monitoring location at that moment.

5. The compliant control method for a robotic arm based on sensor monitoring according to claim 1, characterized in that, The step of obtaining the final decomposition weights for each time step based on the rigid estimation reliability of a monitoring location at each time step and the initial decomposition weights of the STL algorithm includes: The reliability of the rigidity estimate at a monitoring location at a given time is compared with the sum of the reliability of the rigidity estimates at all times within the local range when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm. This sum is then multiplied by the initial decomposition weights when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm to obtain the weight factor at that time. The final decomposition weight at that monitoring location at that time is then compared with the sum of the weight factors at all times within the local range when the relative rigidity estimate coefficient at that time is used for local weighted decomposition using the STL algorithm to obtain the final decomposition weight at that monitoring location at that time.

6. The compliant control method for a robotic arm based on sensor monitoring according to claim 1, characterized in that, The predicted value of the relative stiffness estimation coefficient for the next moment based on the trend term and period term corresponding to a monitoring location includes: The GRU algorithm is used to predict the trend and periodic terms corresponding to a monitoring location to obtain the predicted trend and periodic values ​​for the next moment. The predicted trend and periodic values ​​for the next moment are added together to obtain the predicted relative stiffness estimation coefficient for the next moment.

7. The compliant control method for a robotic arm based on sensor monitoring according to claim 1, characterized in that, The stiffness matrix adjustment coefficient for the next moment is obtained based on the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment, the relative stiffness estimation coefficients at each historical moment, and the predicted value of the relative stiffness estimation coefficient. This includes: If the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment is greater than or equal to the estimated value of the relative stiffness at the current moment; the time distance weight of the historical moment is obtained by negatively mapping the time interval between the historical moment and the current moment using an exponential function with the natural constant as the base; the third average difference is obtained by weighting and averaging the absolute values ​​of the differences between the predicted value of the relative stiffness estimation coefficient at each historical moment and the relative stiffness estimation coefficient based on the time distance weight of each historical moment, and the third average difference is obtained by negatively mapping the third average difference using an exponential function with the natural constant as the base; the relative change degree is obtained by subtracting the ratio of the predicted value of the relative stiffness estimation coefficient at the next moment to the relative stiffness estimation coefficient at the current moment from the first preset value and taking the absolute value; the adjustment range is obtained by multiplying the stiffness matrix adjustment coefficient at the current moment, the relative change degree, and the third mapping value; the stiffness matrix adjustment coefficient at the current moment is added to the adjustment range to obtain the stiffness matrix adjustment coefficient at the next moment. If the predicted value of the relative stiffness estimation coefficient at a monitoring location at the next moment is less than the relative stiffness estimation coefficient at the current moment, the stiffness matrix adjustment coefficient at the current moment is subtracted from the adjustment amplitude to obtain the stiffness matrix adjustment coefficient at the next moment.

Images