Method and device for setting and adjusting load held by robotic arm

Abstract

The present invention provides a method and device for setting and adjusting the load held by a robotic arm. The setting method includes: collecting force feedback data when the robotic arm holds and moves a grasped object; predicting the sliding distance of the grasped object relative to the robotic arm based on the force feedback data at multiple moments; determining whether the sliding distance exceeds a first distance threshold; when the sliding distance exceeds the first distance threshold, determining a load increase amount according to the sliding distance; and using the load increase amount and the current load to set a target load.

Description

Technical Field

[0001] The present invention relates to the field of robot control, and particularly to a method and device for setting and adjusting the grasping load of a robot hand. Background Art

[0002] Robot hands can be applied to various industries. Currently, the automated grasping strategies of robot hands are usually designed for some fixed application scenarios so that they can grasp specific objects. In some application scenarios, the grasped objects may have strong diversity. If the adaptability of the grasping strategy is insufficient, it may result in the inability to grasp the target object or damage to the target object during grasping.

[0003] The existing adaptive grasping strategy is to monitor in real time whether the object slides after grasping the object. Once the object slides, the load is increased until the object no longer slides. It is found in actual applications that the adaptability of this strategy is insufficient. On the one hand, the timing of deciding to increase the load is unreasonable, that is, when the object has already slid and then the load is adjusted, it is impossible to effectively stop the sliding. On the other hand, it is difficult to preset the increasing rate of the load. An excessively low increasing rate cannot avoid sliding, while an excessively high increasing rate may damage the object. Summary of the Invention

[0004] In view of this, the present application provides a method for setting the grasping load of a robot hand, including:

[0005] Collecting force feedback data when the robot hand grasps and moves the grasped object;

[0006] Predicting the sliding distance of the grasped object relative to the robot hand according to the force feedback data at multiple moments;

[0007] Judging whether the sliding distance exceeds a first distance threshold;

[0008] When the sliding distance exceeds the first distance threshold, determining the load increase amount according to the sliding distance;

[0009] Setting the target load by using the load increase amount and the current load.

[0010] Optionally, the force feedback data includes pressure data collected by an electronic skin arranged at the contact part between the robot hand and the grasped object and driving force data of each driving device of the robot hand.

[0011] Optionally, the pressure data is array data collected by electronic skins respectively arranged on the palm and multiple fingers of the robot hand.

[0012] Optionally, predicting the sliding distance of the grasped object relative to the robot hand according to the force feedback data at multiple moments includes:

[0013] The time series prediction module predicts the pressure feature vector at the next moment based on the array data at multiple moments;

[0014] Extract the vertical driving force component from the driving force data, and the time series prediction module predicts the driving force feature vector at the next moment based on the vertical driving force components at multiple moments;

[0015] Connect the pressure feature vector and the driving force feature vector to obtain a global feature vector;

[0016] Obtain the sliding distance according to the global feature vector.

[0017] Optionally, determining whether the sliding distance exceeds a first distance threshold includes:

[0018] Determine whether the sliding distance exceeds a second distance threshold;

[0019] When the sliding distance does not exceed the second distance threshold, determine whether the sliding distance exceeds the first distance threshold;

[0020] When the sliding distance exceeds the second distance threshold, move the robotic arm downward, place the grasped object back on the bearing plane, and prepare to grasp again.

[0021] Optionally, setting the target load using the load increase amount and the current load includes:

[0022] Calculate the total load set by the load increase amount and the current load;

[0023] Determine whether the total load exceeds the load upper limit;

[0024] When the total load does not exceed the load upper limit, set the total load as the target load;

[0025] When the total load exceeds the load upper limit, set the load upper limit as the target load.

[0026] This application also provides a method for adjusting the load held by a robotic arm, including:

[0027] During the process of the robotic arm grasping and moving the grasped object, trigger a first adjustment operation according to a first period and trigger a second adjustment operation according to a second period. The first adjustment operation includes determining whether the grasped object slides based on the force feedback data at the current moment and the force feedback data at the previous moment, and increasing the load when the grasped object slides. The second adjustment operation includes setting the target load using the above method and increasing the load until the target load is reached.

[0028] Optionally, the second period is greater than the first period.

[0029] Optionally, after triggering the second adjustment operation and before the current load reaches the target load, triggering the first adjustment operation is paused.

[0030] Optionally, the load increase amount when triggering the first adjustment operation is a fixed amount, and the fixed amount is less than the load increase amount in the second adjustment operation.

[0031] Optionally, the force feedback data in the first adjustment operation includes the total pressure data collected by an electronic skin provided at the contact part between the robotic arm and the grasped object, and the vertical component extracted from the driving force data of a specific driving device.

[0032] Optionally, determining whether the grasped object slides according to the force feedback data at the current moment and the force feedback data at the previous moment includes:

[0033] Calculating the difference in total pressure according to the total pressure data at the current moment and the previous moment, and calculating the difference in vertical component according to the vertical driving force component at the current moment and the previous moment;

[0034] Determining whether any one of the difference in total pressure and the difference in vertical component exceeds a corresponding set threshold;

[0035] When any one of the difference in total pressure and the difference in vertical component exceeds a corresponding set threshold, it is determined that the grasped object has slid.

[0036] Correspondingly, the present application provides a robotic arm grasping control device, including: a processor and a memory connected to the processor; wherein, the memory stores instructions executable by the processor, and when the instructions are executed by the processor, the processor is enabled to execute the above method.

[0037] The robotic arm grasping load setting method and device provided by the present application can predict the sliding distance of an object relative to the robotic arm by using the feedback data at multiple moments, and can obtain a target load suitable for the current grasped object before the object actually slides, thereby effectively stopping the sliding trend. Moreover, this solution sets the load increment based on the predicted sliding distance size, making the size of the target load reasonable and avoiding damaging the object while stopping the sliding.

[0038] According to the robotic arm grasping load adjustment method and device provided by the present application, two different adjustment operations are used to adjust the load of the robotic arm. Through the first adjustment operation with a faster response speed, it is ensured that the load is increased in a timely manner when a sliding event of the grasped object occurs. Through the second adjustment operation with higher accuracy, the sliding distance is predicted and a suitable target load is set, thereby stopping the sliding trend. This solution takes into account both the response speed and accuracy, and improves the adaptability of the robotic arm to grasp objects. Description of the Drawings

[0039] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0040] Figure 1 It is a flowchart of the method for setting the load held by the robotic arm in the embodiments of the present invention;

[0041] Figure 2 It is a schematic diagram of the neural network model architecture in the embodiments of the present invention. Specific Embodiments

[0042] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the drawings. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0043] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0044] The process of the robotic arm grasping an object can be decomposed into four stages, namely the positioning stage, the loading stage, the grasping stage, and the placing stage. Among them, the positioning stage refers to the robotic arm determining the position where the object to be grasped is located, moving, approaching, and until contacting the object to be grasped; the loading stage refers to after the robotic arm contacts the object to be grasped, adjusting the load of the robotic arm to determine the load value that can grasp but not damage the object to be grasped; the grasping stage refers to grasping the object and moving (moving upward, moving horizontally above the target position); the placing stage refers to placing the object at the target position.

[0045] To avoid the object to be grasped slipping from the robotic arm during the grasping stage, the load needs to be further adjusted. For this stage, this embodiment provides a method for setting the load held by the robotic arm, which can be executed by an electronic device such as a computer or a server, such as Figure 1 shown and includes the following operations:

[0046] S31A, collect the force feedback data when the robotic arm grasps and moves the object to be grasped. The force feedback data can specifically be the pressure or pressure of the contact part between the robotic arm and the object to be grasped, as well as the force of the driving device for changing the load. This method can collect the force feedback data at a certain period, and the value of the force feedback data will change with the vertical or horizontal movement of the robotic arm.

[0047] S32A. Predict the sliding distance of the grasped object relative to the robotic arm based on the force feedback data at multiple moments. The multiple moments may include the current acquisition moment and one or more acquisition moments before it. For example, the initial moment when entering the grasping stage is t0, and the current acquisition moment is t i , then the multiple moments can be from t0 to t i (i.e., all moments from the initial moment to the current moment), or it can be t i-n、 t i (i.e., the current moment and the previous n moments), or it can be t j ~t i (t j refers to the moment when the load was last changed or set).

[0048] The sliding distance can be calculated using a pre-established mathematical model. For example, calculate the corresponding sliding distance based on the change trend, change rate, etc. of the force feedback data at multiple moments; or use a neural network model such as LSTM (Long Short-Term Memory) suitable for processing sequence data, with the force feedback data at multiple moments as the input to output the sliding distance.

[0049] It should be noted that the sliding distance obtained in step S32A refers to the distance that the grasped object is expected to slide, rather than the distance that has already slid before.

[0050] S33A. Determine whether the sliding distance exceeds the first distance threshold. When the sliding distance exceeds the first distance threshold, it indicates that the grasped object may slide more than a certain distance under the current load grasping, thus there is a risk of slipping from the robotic arm. Therefore, step S34A is executed. Otherwise, the load remains unchanged and returns to step S31A.

[0051] In an alternative embodiment, step S33A further includes determining whether the sliding distance exceeds a second distance threshold (the second distance threshold is greater than the first distance threshold. For example, the first distance threshold is 2 mm and the second distance threshold is 7 mm); when the sliding distance does not exceed the second distance threshold, determine whether the sliding distance exceeds the first distance threshold; when the sliding distance exceeds the second distance threshold, it indicates that the sliding distance of the grasped object under the current load grasping is relatively large, and even increasing the load may not be able to stop the slipping. In this case, the robotic arm can be moved downward to place the grasped object back on the bearing plane and prepare for re-grasping.

[0052] S34A. Determine the load increase amount according to the sliding distance. A corresponding relationship between the sliding distance and the load increase amount can be established in advance. For example, different load increase amounts correspond to different sliding distance intervals. If the sliding distance is within a certain interval, the load increase amount corresponding to that interval is used; or mathematical calculation models, neural network models, etc. can be used to calculate the load increase amount based on the sliding distance.

[0053] S35A. Set the target load using the load increase amount and the current load. For example, target load = current load + load increase amount. The load can be understood as a force-related eigenvalue, which can characterize the grasping force of the robotic arm on the grasped object. The eigenvalue can be one or more. Specifically, it can be part of the force feedback data or a value calculated based on the force feedback data.

[0054] The method for setting the grasping load of the robotic arm provided by the embodiments of the present invention uses the feedback data at multiple moments to predict the sliding distance of the object relative to the robotic arm, and can obtain the target load suitable for the current grasped object before the object actually slides, thereby effectively stopping the sliding trend. Moreover, this solution sets the load increment based on the predicted sliding distance size, making the size of the target load reasonable, and avoiding damaging the object while stopping the sliding.

[0055] In one embodiment, in order to further avoid possible damage to the grasped object, the sum of the current load and the load increase amount (total load) can be calculated first, and it is determined whether the total load exceeds the load upper limit. The load upper limit refers to the load above which the grasped object is likely to be damaged (which can be called the damage load). The load upper limit is a preset value, which can be specifically calculated through a set algorithm during the loading stage or can also be set manually. When the total load does not exceed the load upper limit, the total load is set as the target load; when the total load exceeds the load upper limit, the load upper limit is set as the target load.

[0056] In one embodiment, the force feedback data includes the pressure data collected by the electronic skin set at the contact part between the robotic arm and the grasped object, and the driving force data of each driving device of the robotic arm. Further, the pressure data is the array data collected by the electronic skin respectively set on the palm and multiple fingers of the robotic arm.

[0057] The contact part between the robotic arm and the grasped object and the driving device for changing the load depend on the structure of the robotic arm. Taking the humanoid robotic arm as an example, the contact part is the surfaces of 5 fingers and 1 palm, and the driving device for changing the load can be the device for driving the finger joints to move. Specifically, each movable joint can correspond to an independent driving device.

[0058] For a humanoid robotic arm, the pressure data collected by the electronic skin refers to the acting force (pressure or pressure intensity) exerted by the grasped object on the surfaces of the fingers and the palm. The driving force data refers to the force borne by the driving device.

[0059] For example, the pressure data of the th row and th column of the th piece of e-skin is denoted as ; the driving force data .

[0060] According to the above example, the sizes of the e-skins at different positions of the robotic hand or the number of measurement points therein can be different, so the number of rows and columns of the array is different.

[0061] Furthermore, regarding the loads (current load and target load), assuming that each of the 5 fingers is provided with 1 piece of e-skin and the palm is provided with 1 piece of e-skin, the load can be the maximum value among all the array data, i.e., max( ); the load can also be the relatively smaller value between the maximum value of the array data collected by the e-skin of the thumb and the maximum value of the array data collected by the e-skins of the other fingers. The maximum value of the array data of the thumb is denoted as max( ), and the maximum value of the array data of the other four fingers is denoted as max( , , , ), then the load is min(max( ), max( , , ))).

[0062] As Figure 2 shown, in one embodiment, step S32A uses a neural network model to predict the sliding distance, which specifically includes the following processing:

[0063] The time series prediction module predicts the pressure feature vector at the next moment based on the array data at multiple moments. Specifically, a number of ConvLSTM (Convolutional Long Short-Term Memory Network) are used to normalize the pressure data of each piece of e-skin and use it as the input data of a number of consecutive ConvLSTM networks. This layer extracts spatial features at each time step, captures the dynamic changes between time steps, outputs new features and normalizes them. Then, through the Flatten layer, the multi-dimensional data is flattened into a one-dimensional vector (the pressure feature vector is obtained by taking the last time step and flattening). The above layers process each piece of e-skin separately, and the parameters of the layers are shared.

[0064] Extract the vertical driving force component from the driving force data, and the time series prediction module predicts the driving force feature vector at the next moment based on the vertical driving force components at multiple moments. Specifically, a number of LSTM are used to The vertical driving force component therein is used as the input data for several LSTMs, and after being transformed by several LSTM layers, a driving force feature vector is output.

[0065] The connection pressure feature vector and the driving force feature vector are combined to obtain a global feature vector.

[0066] The sliding distance is obtained according to the global feature vector. Specifically, an MLP (Multi-Layer Perceptron) is used, and the global feature vector is used as the input data of the MLP, and the sliding distance is output.

[0067] After obtaining the target load through the above steps S31A~S35A, the operation of adjusting the current load of the robotic arm until the target load is reached is referred to as the second adjustment operation in this application; in one embodiment, it is also adjusted in cooperation with another load adjustment operation based on threshold comparison during the grasping stage.

[0068] This embodiment provides a load adjustment method based on threshold comparison (referred to as the first adjustment operation in this application), which can be executed by an electronic device such as a computer or a server, including:

[0069] S31B, judging whether the grasped object slides according to the force feedback data at the current moment and the force feedback data at the previous moment. When the grasped object slides, step S32B is executed; otherwise, continuous detection is performed. Different from S32A, S31B only gives a yes or no determination result without calculating the sliding distance.

[0070] Regarding the force feedback data of this embodiment, the same data as in the above step S31A can be used, including the pressure data collected by the electronic skin arranged at the contact part between the robotic arm and the grasped object, and the driving force data of each driving device of the robotic arm. Further, the pressure data is the array data collected by the electronic skin respectively arranged on the palm and multiple fingers of the robotic arm.

[0071] To improve the calculation efficiency, the force feedback data used in this embodiment is the sum of the pressure data collected by each electronic skin and the vertical component extracted from the driving force data of a specific driving device. As a preferred embodiment, step S31B includes: calculating the difference in total pressure according to the sum of the pressure data at the current moment and the previous moment, and calculating the difference in vertical components according to the vertical driving force components at the current moment and the previous moment; judging whether any one of the difference in total pressure and the difference in vertical components exceeds the corresponding set threshold; when any one of the difference in total pressure and the difference in vertical components exceeds the corresponding set threshold, it is determined that the grasped object has slid.

[0072] S32B, increasing the load, and the load increase amount is a fixed amount.

[0073] Based on the above first adjustment operation and second adjustment operation, this embodiment provides a method for adjusting the load held by a robotic arm, which can be executed by an electronic device such as a computer or a server, including:

[0074] During the process of the robotic arm grasping and moving the grasped object, the first adjustment operation is triggered according to the first period, and the second adjustment operation is triggered according to the second period. The details of the first adjustment operation are shown in steps S31B to S32B, and the details of the second adjustment operation are shown in steps S31A to S35A to set the target load, and then the current load is adjusted until the target load is reached.

[0075] The triggering periods of the two different adjustment operations are different, so as to ensure that they are not triggered at the same time step. Preferably, the second period is greater than the first period. The load increase amount in the first adjustment operation is a fixed amount, and the load increase amount in the second adjustment operation is an unfixed amount. In a preferred embodiment, the fixed amount is configured to be less than the minimum value of the unfixed amount, for example, different orders of magnitude can be adopted.

[0076] It can be seen from this that when it is determined that the load needs to be increased by triggering the first adjustment operation, a relatively small increase amount will be applied; when it is determined that the load needs to be increased by triggering the second adjustment operation, a relatively large and unfixed increase amount will be applied. Further, after triggering the second adjustment operation, before the current load reaches the target load, the triggering of the first adjustment operation is suspended.

[0077] As an example, the initial load is 1.0 N, the acquisition period of the force feedback data is 10 ms, the first period is 20 ms, and the second period is 100 ms. Two data sets M1 and M2 are set up. M1 is used to store the force feedback data B (the pressure of each electronic skin, the vertical component of the driving force of each driving device) required for the second adjustment operation, and M2 is used to store the force feedback data A (the total pressure, the vertical component of a specific driving device) required for the first adjustment operation. The initial moment when entering the grasping stage is recorded as t0. The example of adjusting the load is as follows:

[0078] t0 + 10 ms: The force feedback data A of the first time step is added to the data set M1.

[0079] t0 + 20 ms: The first adjustment operation is triggered. At this time, the data set M2 is empty, the load is not adjusted, and the force feedback data B is calculated based on the force feedback data A of the first time step and stored in M2.

[0080] t0 + 30 ms: The force feedback data A of the second time step is added to M1.

[0081] t0 + 40 ms: Trigger the first adjustment operation. Calculate the difference in the total pressure and the vertical component by computing the force feedback data B stored in M2 and the force feedback data B calculated from the data of the second time step, and determine whether sliding has occurred (Steps S31B - S32B). Assume that it is determined that no sliding has occurred, do not adjust the load, and the force feedback data B calculated from the data of the second time step overwrites M2. The above calculations can be completed within a few milliseconds.

[0082] ……

[0083] t0 + 80 ms: Trigger the first adjustment operation. Calculate the difference in the total pressure and the vertical component by computing the force feedback data B stored in M2 and the force feedback data B calculated from the data of the fourth time step, and determine whether sliding has occurred (Steps S31B - S32B). Assume that it is determined that sliding has occurred, increase the load to 1.02 N (the increase amount is a fixed value of 0.02 N). The force feedback data B calculated from the data of the fourth time step overwrites M2.

[0084] t0 + 90 ms: Add the force feedback data A of the fifth time step to M1.

[0085] t0 + 100 ms: Trigger the first adjustment operation. Assume that the load is increased to 1.04 N. The force feedback data B calculated from the data of the fifth time step overwrites M2.

[0086] t0 + 105 ms: Trigger the second adjustment operation. In an independent thread, set the target load according to the force feedback data A from t0 to the trigger moment. At this time, start to determine the target load using Steps S31A - S35A. In actual applications, since this calculation process takes a certain amount of time, the first adjustment operation continues to be triggered in the next several time steps until the result is obtained. It should be noted that the current load in Step S35A refers to the load at the time when the second adjustment operation is triggered, that is, 1.04 N.

[0087] t0 + 110 ms: Add the force feedback data A of the sixth time step to M1.

[0088] t0 + 120 ms: Trigger the first adjustment operation. Assume that the load is increased to 1.06 N. The force feedback data B calculated from the data of the sixth time step overwrites M2.

[0089] t0 + 130 ms: Add the force feedback data A of the seventh time step to M1.

[0090] t0 + 140 ms: Trigger the first adjustment operation. Assume that the load is increased to 1.08 N. The force feedback data B calculated from the data of the seventh time step overwrites M2.

[0091] t0 + 145 ms: Steps S31A - S35A are completed, and the sliding distance of 3 mm is output. This distance is greater than the first threshold of 2 mm but less than the second threshold of 7 mm. The load increase is determined to be 0.2 N. The target load = the load at the time of triggering the second adjustment operation (t0 + 105 ms), which is 1.04 N + 0.2 N = 1.24 N. The current load is 1.08 N, and the target load is 1.24 N. Increase the load until the target load is reached. Before the adjustment is completed, the first adjustment operation is no longer triggered.

[0092] t0 + 150 ms: The force feedback data A of the 8th time step is added to M1.

[0093] t0 + 170 ms: The force feedback data A of the 9th time step is added to M1.

[0094] The above examples are only for illustrating how to control the load by simultaneously using two adjustment operations. The specific values of load, cycle, etc. used in the examples are for reference only and do not limit the values of the relevant data.

[0095] According to the robot hand grasping load adjustment method provided by the embodiments of the present invention, two different adjustment operations are used to adjust the robot hand load. Through the first adjustment operation with a faster response speed, it is ensured that the load is increased in time when a sliding event of the grasped object occurs. Through the second adjustment operation with higher accuracy, the sliding distance is predicted and a suitable target load is set, thereby stopping the sliding trend. This solution takes into account both the response speed and accuracy, and improves the adaptability of the robot hand to grasp objects.

[0096] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can be implemented in the form of a computer program product implemented on one or more computer - usable storage media (including but not limited to disk memory, CD - ROM, optical memory, etc.) containing computer - usable program code.

[0097] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general - purpose computer, a special - purpose computer, an embedded processor, or other programmable data - processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data - processing devices generate for implementing in the process Figure 1 a process or multiple processes and / or blocks Figure 1A device for the functions specified in one or more boxes.

[0098] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, and the instruction device implements the functions specified in one Figure 1 process or multiple processes and / or boxes Figure 1 a box or multiple boxes.

[0099] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 process or multiple processes and / or boxes Figure 1 a box or multiple boxes.

[0100] Obviously, the above embodiments are only examples for clear illustration and are not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to enumerate all the implementation manners here. And the obvious changes or modifications derived therefrom are still within the protection scope of the present invention.

Claims

1. A method for setting the load held by a robotic arm, characterized in that, Including: Collecting force feedback data when a robotic arm holds and moves an object to be grasped, where the force feedback data includes pressure data collected by an electronic skin disposed at the contact part between the robotic arm and the object to be grasped and driving force data of each driving device of the robotic arm, and the pressure data is array data collected by electronic skins respectively disposed on the palm and multiple fingers of the robotic arm; Predicting the sliding distance of the object to be grasped relative to the robotic arm based on the force feedback data at multiple moments, further including: predicting a pressure feature vector at the next moment based on the array data at multiple moments by a time series prediction module; extracting a vertical driving force component from the driving force data, and predicting a driving force feature vector at the next moment based on the vertical driving force components at multiple moments by the time series prediction module; connecting the pressure feature vector and the driving force feature vector to obtain a global feature vector; obtaining the sliding distance according to the global feature vector; Judging whether the sliding distance exceeds a first distance threshold; When the sliding distance exceeds the first distance threshold, determining a load increase amount according to the sliding distance; Setting a target load by using the load increase amount and the current load.

2. The method according to claim 1, wherein Judging whether the sliding distance exceeds the first distance threshold includes: Judging whether the sliding distance exceeds a second distance threshold; When the sliding distance does not exceed the second distance threshold, judging whether the sliding distance exceeds the first distance threshold; When the sliding distance exceeds the second distance threshold, moving the robotic arm downward to place the object to be grasped back on the bearing plane and preparing for re-grasping.

3. The method according to claim 1, characterized in that Setting a target load by using the load increase amount and the current load includes: Calculating the total load set by the load increase amount and the current load; Judging whether the total load exceeds a load upper limit; When the total load does not exceed the load upper limit, setting the total load as the target load; When the total load exceeds the load upper limit, setting the load upper limit as the target load.

4. A method for adjusting the load held by a robotic arm, characterized in that, Including: During the process of the robotic arm holding and moving the object to be grasped, triggering a first adjustment operation according to a first period and triggering a second adjustment operation according to a second period. The first adjustment operation includes judging whether the object to be grasped slides according to the force feedback data at the current moment and the force feedback data at the previous moment, and increasing the load when the object to be grasped slides. The second adjustment operation includes setting a target load by using the method according to any one of claims 1-3 and increasing the load until the target load is reached.

5. The method according to claim 4, characterized in that, The second period is greater than the first period.

6. The method according to claim 5, characterized in that After triggering the second adjustment operation and before the current load reaches the target load, pausing the triggering of the first adjustment operation.

7. The method according to claim 4, wherein The load increase amount when triggering the first adjustment operation is a fixed amount, and the fixed amount is less than the load increase amount in the second adjustment operation.

8. The method according to claim 4, characterized in that, The force feedback data in the first adjustment operation includes the total pressure data collected by the electronic skin disposed at the contact part between the robotic arm and the object to be grasped and the vertical component extracted from the driving force data of a specific driving device.

9. The method according to claim 8, wherein Judging whether the object to be grasped slides according to the force feedback data at the current moment and the force feedback data at the previous moment includes: Calculate the difference in total pressure based on the sum of the pressure data at the current moment and the previous moment, and calculate the difference in vertical components based on the vertical driving force components at the current moment and the previous moment; Determine whether any one of the difference in total pressure and the difference in vertical components exceeds a corresponding set threshold; When any one of the difference in total pressure and the difference in vertical components exceeds a corresponding set threshold, it is determined that the gripper has slipped.

10. A robotic hand gripping control device, characterized in that, Comprising: A processor and a memory connected to the processor; wherein, the memory stores instructions executable by the processor, and the instructions are executed by the processor to cause the processor to execute the method according to any one of claims 1-9.

Images