A high-precision flexible jaw closed-loop force control system and a grabbing method

Abstract

The application discloses a high-precision flexible gripper closed-loop force control system and a grabbing method, and relates to the technical field of automatic control.The method comprises the following steps: dividing a contact subarea on a flexible gripper finger surface, combining a target workpiece continuous contour to build a contact baseline data set, and the contact baseline data set comprises a theoretical contact subarea sequence, a first theoretical supporting center position, a second theoretical supporting center position, a theoretical synchronization difference sequence, a pressure noise threshold and a deformation noise threshold; identifying an actual contact subarea and a first contact time point in a gripper closing stage, building an actual contact subarea sequence, comparing the actual contact sequence with the theoretical contact sequence, simultaneously analyzing the left and right first contact synchronization deviation degree, continuously monitoring the offset change of the left and right supporting center positions in a grabbing stage, generating a grabbing level control instruction, and adjusting the gripper clamping torque.The application improves the control precision of the gripper grabbing through contact identification, synchronization determination and closed-loop control of the grabbing process.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, specifically to a closed-loop force control system and gripping method for a high-precision flexible gripper. Background Technology

[0002] With the development of intelligent manufacturing systems and the gradual promotion of flexible production lines, the application scope of industrial robots in sorting, assembly, and transportation continues to expand. Especially in the processing of electronic components, precision structural parts, and irregularly shaped workpieces, higher requirements are placed on the adaptability of end effectors. Against this backdrop, flexible grippers, as an execution structure that takes into account both deformation adaptation and contact buffering capabilities, have gradually become an important technical path for gripping operations under complex working conditions. Their application has expanded from handling regular workpieces to irregular contour recognition, multi-point contact collaborative control, and dynamic gripping process adjustment. The control methods surrounding flexible grippers have also gradually shifted from traditional open-loop drive to closed-loop force control mechanisms based on contact state perception and process feedback.

[0003] However, in the gripping process of flexible grippers, existing technologies usually rely on overall force feedback or single-point pressure detection to adjust the gripping process. They lack analysis of the collaborative relationship between the multiple contact zones of the gripper finger surfaces, making it difficult to describe in detail the contact sequence, the synchronous relationship of left and right bearing, and the changes in the local bearing center. At the same time, for workpieces with irregular contours, existing methods are unable to establish a reference for the contact distribution before gripping, and they also lack a mechanism to identify the deviation between the actual contact sequence and the theoretical contact sequence during gripping. As a result, gripping control remains at the overall adjustment level, making it difficult to make targeted adjustments for local imbalances. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a closed-loop force control system and gripping method for a high-precision flexible gripper, solving the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] In a first aspect, the present invention provides a gripping method for a high-precision flexible gripper, comprising the following steps:

[0007] S1. Obtain several contact zones on the finger surfaces of the flexible gripper and deploy multiple sets of sensors. During the gripper's stationary phase, analyze the proximity of the corresponding contact zones to the boundary points of each contour within the continuous contour and construct a contact baseline dataset. The contact baseline dataset includes the theoretical contact zone sequence, the first theoretical bearing center position, the second theoretical bearing center position, the theoretical synchronization difference sequence, the pressure noise threshold, and the deformation noise threshold.

[0008] S2. During the gripper closing phase, based on the contact baseline dataset, contact state partitions are selected, and the first contact time point of each contact state partition is determined. The actual contact partition sequence is constructed, and a synchronization deviation analysis command is issued after comparative analysis.

[0009] S3. After receiving the synchronization deviation analysis command, analyze the degree of left and right first touch synchronization deviation according to the theoretical synchronization difference sequence, and issue the load offset analysis command.

[0010] S4. After receiving the load-bearing offset analysis command, during the gripper grasping stage, based on the first theoretical load-bearing center position and the second theoretical load-bearing center position, analyze the level of change of the left center offset unit and the right center offset unit of the corresponding monitoring point compared with the previous monitoring point, and generate the corresponding gripping level control command for the monitoring point.

[0011] S5. Execute the corresponding capture control operation according to the capture level control command generated for the corresponding monitoring point.

[0012] Preferably, the flexible gripper finger surface is uniformly divided to obtain several contact zones, and multiple sets of sensors are deployed, specifically including:

[0013] An industrial camera is set above the gripping station and a side camera is set in front of the gripper. A displacement encoder is set on the gripper drive slider. Based on the flexible contact layer of the left and right finger surfaces of the gripper, the flexible contact layer is evenly divided along the length and width directions of the finger surfaces to obtain several contact partitions. Thin film pressure sensors are embedded on the surface of each contact partition and flexible strain sensors are attached to the corresponding bottom layer.

[0014] Preferably, the proximity of the corresponding contact zone to the boundary points of each contour within the continuous contour is analyzed, specifically including:

[0015] During the gripper stationary phase, top view and side view contour images of the target workpiece to be gripped are acquired using an industrial camera and a side camera. After continuous boundary recognition, the continuous contour of the target workpiece to be gripped is determined.

[0016] Using the midpoint of the line connecting the geometric center points of the left and right finger surfaces of the gripper as the origin, a three-dimensional coordinate system is established. After the gripper assembly calibration process, the center coordinate positions of each contact zone are obtained. Each contact zone is assigned a unique direction label number, which includes the left side number and the right side number.

[0017] Based on the established three-dimensional coordinate system, the point coordinates of each boundary point within the continuous contour are extracted. Based on the projection area of ​​the continuous contour vertically projected onto the left and right finger surfaces of the gripper, and combined with the center coordinates of each contact zone, the center coordinates of each contact zone within the projection area are extracted.

[0018] For any contact zone within the projection area, analyze the proximity of the corresponding contact zone to the boundary points of each contour within the continuous contour, and determine the pre-contact distance of the corresponding contact zone.

[0019] The pre-contact distances of each contact zone are matched with the directional label numbers of the corresponding contact zones. Based on the magnitude of the pre-contact distances, the contact zones are sorted in ascending order to construct a theoretical contact zone sequence. Each contact zone in the theoretical contact zone sequence corresponds to a unique directional label number and a pre-contact distance.

[0020] Preferably, the first theoretical bearing center location and the second theoretical bearing center location specifically include:

[0021] Based on the direction label numbers of each contact partition in the theoretical contact partition sequence, the left contact partition is paired with the corresponding right contact partition to obtain several partition alignment groups. The left pre-contact distance and the right pre-contact distance in each partition alignment group are summed to obtain the unit pre-contact distance of each partition alignment group. Combined with the statistical mean calculation algorithm, the average pre-contact distance is determined. The difference between the unit pre-contact distance and the average pre-contact distance of each partition alignment group is calculated. The center coordinate positions of the left contact partition and the right contact partition in the partition alignment group with the smallest difference are extracted and used as the first theoretical bearing center position and the second theoretical bearing center position, respectively.

[0022] Preferably, the theoretical synchronization difference sequence, pressure noise threshold, and deformation noise threshold specifically include:

[0023] For any partition alignment group, extract the left pre-contact distance and the right pre-contact distance in the corresponding partition alignment group, analyze the pre-contact synchronization level of the corresponding partition alignment group, and determine the theoretical synchronization difference of the corresponding partition alignment group;

[0024] Based on the determined theoretical synchronization difference values ​​of each partition pair, the corresponding theoretical synchronization differences are sorted in ascending order to construct a theoretical synchronization difference sequence.

[0025] By embedding thin-film pressure sensors on the surface of each contact zone and attaching flexible strain sensors in the corresponding underlying layer, the unloaded pressure value and unloaded compression deformation value of each contact zone are collected respectively, and used as the pressure noise threshold and deformation noise threshold of the corresponding contact zone.

[0026] A contact baseline dataset is constructed based on the theoretical contact partition sequence, the first theoretical contact center location, the second theoretical contact center location, the theoretical synchronization difference sequence, the pressure noise threshold, and the deformation noise threshold.

[0027] Preferably, the first contact time point of each contact state zone is determined, specifically including:

[0028] During the gripper closing phase, sampling points are set with a preset first time step, and the pressure and deformation values ​​of each contact zone at different sampling points are collected in real time through an embedded thin-film pressure sensor and a attached flexible strain sensor.

[0029] According to the time sequence of the sampling points, the unit change zone pressure value and unit change zone deformation value of the corresponding contact zone at the corresponding sampling point are compared and analyzed with the pressure noise threshold and deformation noise threshold of the corresponding contact zone. If the unit change zone pressure value at the corresponding sampling point is greater than the pressure noise threshold and the unit change zone deformation value is greater than the deformation noise threshold, then the corresponding contact zone is recorded as a contact state zone, and the first time point that meets the comparison conditions is recorded as the first contact time point of the corresponding contact state zone. Otherwise, the corresponding contact zone is recorded as a contact idle zone.

[0030] Preferably, after comparative analysis, a synchronization deviation analysis command is issued, specifically including:

[0031] Based on the first contact time of each contact state partition, the contact state partitions are sorted in chronological order to construct the actual contact partition sequence. The actual contact sequence value of each contact state partition in the actual contact partition sequence is extracted. Feature recognition is performed on the theoretical contact partition sequence to extract the theoretical contact sequence value of each contact state partition.

[0032] The theoretical contact sequence and actual contact sequence of each contact state partition are compared and analyzed. If the actual contact sequence value of the corresponding contact state partition exceeds the theoretical contact sequence value, the corresponding contact state partition is recorded as a misaligned contact state partition, and the number of misaligned contact state partitions is counted. If the actual contact sequence value of the corresponding contact state partition does not exceed the theoretical contact sequence value, the corresponding contact state partition is recorded as a normal contact state partition, and the number of normal contact state partitions is counted. The ratio of the number of misaligned contact state partitions to the total number of contact state partitions is calculated to obtain the misalignment ratio. When the misalignment ratio exceeds 5%, a synchronization deviation analysis command is issued; otherwise, no additional synchronization deviation analysis command is issued.

[0033] Preferably, the analysis includes the degree of deviation in synchronization between the left and right first touches, specifically including:

[0034] After receiving the synchronization deviation analysis command, the partition alignment group where the misaligned contact state partition is located is recorded as the partition comparison alignment group. Based on the first contact time point of each contact state partition, the first contact time point of the left contact partition and the first contact time point of the right contact partition in each partition comparison alignment group are extracted. The degree of left and right first contact synchronization deviation in the corresponding partition comparison alignment group is analyzed to determine the actual first contact time difference of the corresponding partition comparison alignment group.

[0035] Feature recognition is performed on the theoretical synchronization difference sequence to extract the theoretical synchronization difference value of each partition comparison alignment group. Combined with the average advancing speed of the gripper before the first contact, the theoretical first contact time difference of each partition comparison alignment group is determined. The average advancing speed is obtained by reading the displacement change of the displacement encoder at several consecutive sampling points before the first contact and dividing it by the corresponding time change.

[0036] The actual first contact time difference of each partition comparison alignment group is compared and analyzed with the corresponding theoretical first contact time difference, specifically including:

[0037] If the actual first contact time difference of the corresponding partition comparison alignment group is equal to the corresponding theoretical first contact time difference, the synchronization status of the corresponding partition comparison alignment group is recorded as normal synchronization status. If the actual first contact time difference of the corresponding partition comparison alignment group is not equal to the corresponding theoretical first contact time difference, the synchronization status of the corresponding partition comparison alignment group is recorded as abnormal synchronization status. The synchronization status of all partition comparison alignment groups is summarized. If there is an abnormal synchronization status among the synchronization statuses of all partition comparison alignment groups, a load-bearing offset analysis instruction is generated. If there is no abnormal synchronization status among the synchronization statuses of all partition comparison alignment groups, no additional load-bearing offset analysis instruction is generated.

[0038] Preferably, the analysis includes the level of change in left center offset and right center offset of the corresponding monitoring point compared to the previous monitoring point, specifically including:

[0039] Upon receiving the load-bearing offset analysis command, during the gripper grasping stage, monitoring points are set with a preset second time step. Through embedded thin-film pressure sensors and attached flexible strain sensors, the load-bearing status of the left and right contact zones within each partition comparison alignment group is monitored in real time. The partition clamping pressure value and partition clamping deformation value of the left and right contact zones within the corresponding partition comparison alignment group at each monitoring point are obtained. The partition clamping pressure value and partition clamping deformation value are weighted and summed to obtain the load-bearing degree coefficient of the left and right contact zones within the corresponding partition comparison alignment group at each monitoring point. After feature recognition, the maximum value of the load-bearing degree coefficient of the left contact zone and the maximum value of the load-bearing degree coefficient of the right contact zone at each monitoring point are extracted respectively.

[0040] The center coordinates of the contact zones corresponding to the maximum bearing capacity coefficient of the left and right contact zones of each monitoring point are taken as the actual bearing center positions of the left and right contact zones of each monitoring point.

[0041] The actual receiving center position of the left partition of each monitoring point is associated with the corresponding first theoretical receiving center position, and the actual receiving center position of the right partition is associated with the corresponding second theoretical receiving center position. Combined with the Euclidean distance algorithm, the left center offset distance and right center offset distance of each monitoring point are determined respectively.

[0042] During the gripper grasping phase, for any monitoring point other than the first monitoring point, the left and right center offset distances of the corresponding monitoring point are subtracted from the left and right center offset distances of the previous monitoring point. These are used as the left and right center offset unit change distances of the corresponding monitoring point relative to the previous monitoring point. The left and right center offset unit change distances are used to characterize the left and right center offset levels of the corresponding monitoring point relative to the previous monitoring point during the gripper grasping phase.

[0043] Preferably, the corresponding capture level control instruction for the monitoring point is generated, specifically including:

[0044] Based on the magnitudes of the changes in left and right center offsets of the corresponding monitoring point compared to the previous monitoring point, a capture level control instruction is generated for the corresponding monitoring point, specifically including:

[0045] If the left center offset unit distance and the right center offset unit distance of the corresponding monitoring point are both not equal to zero compared to the previous monitoring point, it indicates that the gripper is in a bidirectional unstable state when grasping the target workpiece at the corresponding monitoring point, and a first-level gripping control command is generated for the corresponding monitoring point. If the left center offset unit distance or the right center offset unit distance of the corresponding monitoring point is equal to zero compared to the previous monitoring point, it indicates that the gripper is in a unidirectional unstable state when grasping the target workpiece at the corresponding monitoring point, and a second-level gripping control command is generated for the corresponding monitoring point. If the left center offset unit distance and the right center offset unit distance of the corresponding monitoring point are both equal to zero compared to the previous monitoring point, it indicates that the gripper is in a stable gripping state when grasping the target workpiece at the corresponding monitoring point, and a third-level gripping control command is generated for the corresponding monitoring point.

[0046] Preferably, the corresponding capture control operation is executed, specifically including:

[0047] If the gripping level control instruction for the corresponding monitoring point is a Level 1 gripping control instruction, the execution content is: increase the clamping torque of the gripper at the next monitoring point to the target workpiece by two levels; if the gripping level control instruction for the corresponding monitoring point is a Level 2 gripping control instruction, the execution content is: increase the clamping torque of the gripper at the next monitoring point to the target workpiece by one level; if the gripping level control instruction for the corresponding monitoring point is a Level 3 gripping control instruction, the execution content is: continue to maintain the clamping torque level of the gripper at the current monitoring point to the target workpiece, and continuously monitor the offset of the left and right bearing center positions during the movement of the gripper to the target workpiece.

[0048] In a second aspect, the present invention provides a closed-loop force control system for a high-precision flexible gripper, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the gripping method described in the first aspect above.

[0049] This invention provides a closed-loop force control system and gripping method for a high-precision flexible gripper, which has the following advantages:

[0050] (1) By constructing a contact baseline dataset before gripping, the spatial relationship between the contact partition and the continuous contour of the target workpiece is transformed into a theoretical contact partition sequence with sequence characteristics. On this basis, the theoretical bearing center position and theoretical synchronization difference sequence are introduced to uniformly characterize the spatial distribution and temporal evolution of contact behavior. During the gripping process, the structured identification of the contact sequence relationship is realized by constructing the actual contact partition sequence and its item-by-item alignment analysis with the theoretical contact partition sequence. Furthermore, the contact synchronization state is determined by the correspondence between the left and right first contact time difference and the theoretical synchronization difference. In the load analysis stage, by continuously monitoring the load state changes of each partition alignment group and combining the offset relationship between the actual bearing center position and the theoretical bearing center position, the local force evolution trajectory during the gripping process is gradually analyzed. The three types of information, contact sequence deviation, synchronization deviation and load offset, are transformed into hierarchical control commands and applied to the clamping torque adjustment process. This transforms the entire gripping process from a single overall feedback mode into a closed-loop control path with multi-partition and multi-stage linkage characteristics, forming a continuous control process from contact prediction, contact identification to load correction.

[0051] (2) To address the lack of a reference for the pre-grabbing contact distribution in the existing technology, the continuous contour of the target workpiece is spatially mapped, and the pre-contact distance corresponding to each contact zone is constructed by combining the center coordinate position of the contact zone. The zone is then sorted according to the distance to form a theoretical contact zone sequence with spatial sequence relationship, so that the contact behavior has a quantifiable sequence basis before grasping. On this basis, the left and right contact zones are numbered and paired, and the theoretical synchronization difference of each zone alignment group is further calculated and a theoretical synchronization difference sequence is constructed, expanding the left and right contact relationship from a single time event to a synchronization reference system with a sequence structure. In the actual grasping process, the first contact time point is collected and the actual contact zone sequence is formed. The actual contact sequence is compared with the theoretical contact sequence item by item to identify the misaligned contact state zone, and the synchronization deviation analysis process is triggered when the misalignment ratio reaches the set condition. Furthermore, the correspondence between the actual first contact time difference and the theoretical first contact time difference is combined to determine the left and right contact synchronization state group by group, thereby transforming the time deviation problem in the contact process into a structured sequence comparison problem, so that the contact synchronization state no longer depends on a single point or overall judgment, but is identified item by item based on the partition alignment relationship.

[0052] (3) Based on synchronous deviation identification, a continuous monitoring mechanism for load-bearing state is introduced. By setting monitoring points in the grasping stage, the clamping pressure value and clamping deformation value of each partition in the alignment group are collected and the corresponding load-bearing degree coefficient is formed. Then, the contact partition center position corresponding to the maximum value of the load-bearing degree coefficient on the left and right sides is extracted as the actual load-bearing center position of each monitoring point. By associating the actual load-bearing center position with the theoretical load-bearing center position and calculating its spatial offset distance, the offset change is further calculated between continuous monitoring points, so that the load-bearing change is transformed from a static offset to a dynamic change sequence. On this basis, the grasping process is graded according to the combination state of the unit change distance of the left and right center offsets, and the corresponding grasping level control command is generated, so that the control process can be segmented according to the local load-bearing change trend. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the gripping method of a high-precision flexible gripper according to the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Example

[0056] Please see Figure 1 This invention provides a gripping method for a high-precision flexible gripper, comprising the following steps:

[0057] S1. Obtain several contact zones on the flexible gripper finger surfaces and deploy multiple sets of sensors. During the gripper stationary stage, analyze the proximity of the corresponding contact zones to the boundary points of each contour within the continuous contour and construct a contact baseline dataset. The contact baseline dataset includes a theoretical contact zone sequence, a first theoretical bearing center position, a second theoretical bearing center position, a theoretical synchronization difference sequence, a pressure noise threshold, and a deformation noise threshold. The deployment of multiple sets of sensors includes an industrial camera set above the gripping station and a lateral camera set in front of the gripper side, and a displacement encoder set on the gripper drive slider. Based on the flexible contact layer of the left and right finger surfaces of the gripper, the flexible contact layer is uniformly divided along the length and width directions of the finger surfaces to obtain several contact zones. Thin film pressure sensors are embedded on the surface of each contact zone, and flexible strain sensors are attached to the corresponding bottom layer.

[0058] Specifically, the analysis examines the proximity of the corresponding contact zone to the boundary points of each contour within the continuous contour, including:

[0059] During the gripper stationary phase, top view and side view contour images of the target workpiece to be gripped are acquired using an industrial camera and a side camera. After continuous boundary recognition, the continuous contour of the target workpiece to be gripped is determined.

[0060] It should be noted that the continuous boundary recognition process includes: performing grayscale normalization on the top-view and side-view contour images acquired by the industrial camera to eliminate pixel fluctuations caused by lighting differences; performing edge gradient extraction on the normalized images to obtain the grayscale change amplitude of each pixel position and selecting a set of candidate edge pixels that meet the condition of continuous gradient change; connecting the candidate edge pixels point by point according to spatial adjacency to construct initial boundary segments, and performing neighborhood search and direction consistency verification on the boundary segments with breaks, splicing the segments that meet the extension direction consistency to form a continuous boundary chain; removing discrete noise points in the continuous boundary chain, and smoothing and correcting local abrupt points according to the boundary curvature change to obtain stable boundary curves in the top-view and side-view directions; and performing spatial alignment and fusion processing on the top-view boundary curve and the side-view boundary curve in a unified coordinate system to determine the continuous contour of the target workpiece to be grasped.

[0061] Using the midpoint of the line connecting the geometric center points of the left and right finger surfaces of the gripper as the origin, a three-dimensional coordinate system is established. After the gripper assembly calibration process, the center coordinate positions of each contact zone are obtained. Each contact zone is assigned a unique direction label number, which includes the left side number and the right side number.

[0062] Based on the established three-dimensional coordinate system, the point coordinates of each boundary point within the continuous contour are extracted. Based on the projection area of ​​the continuous contour vertically projected onto the left and right finger surfaces of the gripper, and combined with the center coordinates of each contact zone, the center coordinates of each contact zone within the projection area are extracted.

[0063] It should be noted that in a specific precision assembly application scenario for electronic connectors, the grippers are used to grasp miniature housings with asymmetrical shapes. In a workpiece-free state, the spatial distribution of each contact zone on the left and right finger surfaces is read. A calibration plate, in conjunction with an industrial camera, acquires the boundary contours and spatial orientation information of the left and right finger surfaces, determining the geometric center points of the left and right finger surfaces. The midpoint of the line connecting the two geometric center points is used as the origin of the coordinate system. The direction along the connecting line is defined as the X-axis, the finger surface normal direction as the Z-axis, and the direction perpendicular to the X-axis and located in the finger surface unfolding direction as the Y-axis, establishing a three-dimensional coordinate system. Through a segment-by-segment scanning method, the spatial position of each contact zone in this coordinate system is read, mapping the center point of each zone surface to the corresponding three-dimensional coordinate value. During this process, according to the distribution order of the contact zones on the left and right finger surfaces, the left finger surface zones are assigned progressively increasing left-side numbers from left to right, and the right finger surface zones are assigned corresponding sequential right-side numbers. Zones with symmetrical or relative positions are also associated with numbers, forming unique directional label numbers with directional distinctions.

[0064] For any contact zone within the projection area, analyze the proximity of the corresponding contact zone to the boundary points of each contour within the continuous contour, and determine the pre-contact distance of the corresponding contact zone.

[0065] The mathematical expression for the pre-contact distance of the corresponding contact zone is as follows:

[0066]

[0067] In the formula, Let i be the center coordinates of the i-th contact partition. Let p be the coordinates of the p-th boundary point in the continuous contour. It is a local minimum function;

[0068] The pre-contact distances of each contact zone are matched with the directional label numbers of the corresponding contact zones. Based on the magnitude of the pre-contact distances, the contact zones are sorted in ascending order to construct a theoretical contact zone sequence. Each contact zone in the theoretical contact zone sequence corresponds to a unique directional label number and a pre-contact distance.

[0069] Specifically, the locations of the first and second theoretical support centers include:

[0070] Based on the direction label numbers of each contact partition in the theoretical contact partition sequence, the left contact partition is paired with the corresponding right contact partition to obtain several partition alignment groups. The left pre-contact distance and the right pre-contact distance in each partition alignment group are summed to obtain the unit pre-contact distance of each partition alignment group. Combined with the statistical mean calculation algorithm, the average pre-contact distance is determined. The difference between the unit pre-contact distance and the average pre-contact distance of each partition alignment group is calculated. The center coordinate positions of the left contact partition and the right contact partition in the partition alignment group with the smallest difference are extracted and used as the first theoretical bearing center position and the second theoretical bearing center position, respectively.

[0071] It should be noted that the first theoretical bearing center position corresponds to the theoretical force concentration point on the left finger side, and the second theoretical bearing center position corresponds to the theoretical force concentration point on the right finger side. These two positions represent the spatial positions where the left and right finger surfaces should preferentially participate in contact and bear the main bearing function under ideal gripping conditions. In the subsequent gripping process, the actual bearing center position obtained by real-time monitoring is compared with these two theoretical bearing center positions to determine whether the force distribution has shifted during the gripping process, as well as the direction and degree of the shift, and serves as a reference benchmark for force shift analysis and the generation of gripping level control commands.

[0072] Among them, the theoretical synchronization difference sequence, pressure noise threshold, and deformation noise threshold specifically include:

[0073] For any partition alignment group, extract the left pre-contact distance and the right pre-contact distance in the corresponding partition alignment group, analyze the pre-contact synchronization level of the corresponding partition alignment group, and determine the theoretical synchronization difference of the corresponding partition alignment group;

[0074] The mathematical expression for the theoretical synchronization difference of the corresponding partition pair is as follows:

[0075]

[0076] In the formula, This represents the theoretical synchronization difference for the m-th partition pair group. and These are the pre-contact distances of the left and right contact partitions in the m-th partition alignment group, respectively;

[0077] Based on the determined theoretical synchronization difference values ​​of each partition pair, the corresponding theoretical synchronization differences are sorted in ascending order to construct a theoretical synchronization difference sequence.

[0078] By embedding thin-film pressure sensors on the surface of each contact zone and attaching flexible strain sensors in the corresponding underlying layer, the unloaded pressure value and unloaded compression deformation value of each contact zone are collected respectively, and used as the pressure noise threshold and deformation noise threshold of the corresponding contact zone.

[0079] A contact baseline dataset is constructed based on the theoretical contact partition sequence, the first theoretical contact center location, the second theoretical contact center location, the theoretical synchronization difference sequence, the pressure noise threshold, and the deformation noise threshold.

[0080] S2. During the gripper closing phase, based on the contact baseline dataset, contact state partitions are selected, and the first contact time point of each contact state partition is determined. The actual contact partition sequence is constructed, and a synchronization deviation analysis command is issued after comparative analysis.

[0081] Specifically, determining the first contact time point for each contact state zone includes:

[0082] During the gripper closing phase, sampling points are set with a preset first time step, and the pressure and deformation values ​​of each contact zone at different sampling points are collected in real time through an embedded thin-film pressure sensor and a attached flexible strain sensor.

[0083] The first time step is set according to the response change rate during the initial contact establishment stage of the gripper closure, so that adjacent sampling points can capture the pressure and deformation change process at the moment of first contact. Specifically, it can be set to a fixed sampling interval in the range of 1ms to 5ms.

[0084] According to the time sequence of the sampling points, the unit change zone pressure value and unit change zone deformation value of the corresponding contact zone at the corresponding sampling point are compared and analyzed with the pressure noise threshold and deformation noise threshold of the corresponding contact zone. If the unit change zone pressure value at the corresponding sampling point is greater than the pressure noise threshold and the unit change zone deformation value is greater than the deformation noise threshold, then the corresponding contact zone is recorded as a contact state zone, and the first time point that meets the comparison conditions is recorded as the first contact time point of the corresponding contact state zone. Otherwise, the corresponding contact zone is recorded as a contact idle zone.

[0085] The process of issuing a synchronization deviation analysis command after comparative analysis includes:

[0086] Based on the first contact time of each contact state partition, the contact state partitions are sorted in chronological order to construct the actual contact partition sequence. The actual contact sequence value of each contact state partition in the actual contact partition sequence is extracted. Feature recognition is performed on the theoretical contact partition sequence to extract the theoretical contact sequence value of each contact state partition.

[0087] The theoretical contact sequence and actual contact sequence of each contact state partition are compared and analyzed. If the actual contact sequence value of the corresponding contact state partition exceeds the theoretical contact sequence value, the corresponding contact state partition is recorded as a misaligned contact state partition, and the number of misaligned contact state partitions is counted. If the actual contact sequence value of the corresponding contact state partition does not exceed the theoretical contact sequence value, the corresponding contact state partition is recorded as a normal contact state partition, and the number of normal contact state partitions is counted. The ratio of the number of misaligned contact state partitions to the total number of contact state partitions is calculated to obtain the misalignment ratio. When the misalignment ratio exceeds 5%, a synchronization deviation analysis command is issued; otherwise, no additional synchronization deviation analysis command is issued.

[0088] S3. After receiving the synchronization deviation analysis command, analyze the degree of left and right first touch synchronization deviation according to the theoretical synchronization difference sequence, and issue the load offset analysis command.

[0089] The analysis of the degree of synchronization deviation between the left and right first touches specifically includes:

[0090] After receiving the synchronization deviation analysis command, the partition alignment group where the misaligned contact state partition is located is recorded as the partition comparison alignment group. Based on the first contact time point of each contact state partition, the first contact time point of the left contact partition and the first contact time point of the right contact partition in each partition comparison alignment group are extracted. The degree of left and right first contact synchronization deviation in the corresponding partition comparison alignment group is analyzed to determine the actual first contact time difference of the corresponding partition comparison alignment group.

[0091] The mathematical expression for the actual first contact time difference of the corresponding partition comparison group is as follows:

[0092]

[0093] In the formula, For the k-th partition, compare the actual first contact time difference of the corresponding group. and These are the first contact time points of the left contact partition and the right contact partition within the k-th partition comparison group, respectively.

[0094] Feature recognition is performed on the theoretical synchronization difference sequence to extract the theoretical synchronization difference value of each partition comparison alignment group. Combined with the average advancing speed of the gripper before the first contact, the theoretical first contact time difference of each partition comparison alignment group is determined. The average advancing speed is obtained by reading the displacement change of the displacement encoder at several consecutive sampling points before the first contact and dividing it by the corresponding time change.

[0095] The actual first contact time difference of each partition comparison alignment group is compared and analyzed with the corresponding theoretical first contact time difference, specifically including:

[0096] If the actual first contact time difference of the corresponding partition comparison alignment group is equal to the corresponding theoretical first contact time difference, the synchronization status of the corresponding partition comparison alignment group is recorded as normal synchronization status. If the actual first contact time difference of the corresponding partition comparison alignment group is not equal to the corresponding theoretical first contact time difference, the synchronization status of the corresponding partition comparison alignment group is recorded as abnormal synchronization status. The synchronization status of all partition comparison alignment groups is summarized. If there is an abnormal synchronization status among the synchronization statuses of all partition comparison alignment groups, a load-bearing offset analysis instruction is generated. If there is no abnormal synchronization status among the synchronization statuses of all partition comparison alignment groups, no additional load-bearing offset analysis instruction is generated.

[0097] S4. After receiving the load-bearing offset analysis command, during the gripper grasping stage, based on the first theoretical load-bearing center position and the second theoretical load-bearing center position, analyze the level of change of the left center offset unit and the right center offset unit of the corresponding monitoring point compared with the previous monitoring point, and generate the corresponding gripping level control command for the monitoring point.

[0098] The analysis includes the level of change in left and right center offset units at the corresponding monitoring points compared to the previous monitoring point, specifically including:

[0099] Upon receiving the load-bearing offset analysis command, during the gripper grasping stage, monitoring points are set with a preset second time step. Through embedded thin-film pressure sensors and attached flexible strain sensors, the load-bearing status of the left and right contact zones within each partition comparison alignment group is monitored in real time. The partition clamping pressure value and partition clamping deformation value of the left and right contact zones within the corresponding partition comparison alignment group at each monitoring point are obtained. The partition clamping pressure value and partition clamping deformation value are weighted and summed to obtain the load-bearing degree coefficient of the left and right contact zones within the corresponding partition comparison alignment group at each monitoring point. After feature recognition, the maximum value of the load-bearing degree coefficient of the left and right contact zones at each monitoring point are extracted respectively. The second time step is set according to the gripper closing and advancing speed and the sensor sampling response period to ensure that the displacement change between adjacent monitoring points does not exceed 1 / 5 of the scale of a single contact zone. Specifically, it can be set to a fixed sampling interval within the range of 5ms to 20ms.

[0100] The center coordinates of the contact zones corresponding to the maximum bearing capacity coefficient of the left and right contact zones of each monitoring point are taken as the actual bearing center positions of the left and right contact zones of each monitoring point.

[0101] It should be noted that the actual bearing center positions of the left and right zones at each monitoring point are determined based on the force distribution of the contact zones within the alignment group at the current monitoring point. Specifically, at the same monitoring point, the bearing degree coefficient for each contact zone is obtained by combining the zone clamping pressure and zone clamping deformation values ​​of each contact zone on the left and right sides. Within the same side zone range, the contact zone with the largest bearing degree coefficient is selected, and the center coordinate position of that contact zone is taken as the actual bearing center position at the current monitoring point. The basis for this determination is that the bearing degree coefficient reflects the concentration and contribution level of the contact zone in the clamping force at the current moment, and the zone with the largest value corresponds to the position that bears the main clamping role in the current contact process.

[0102] The actual receiving center position of the left partition of each monitoring point is associated with the corresponding first theoretical receiving center position, and the actual receiving center position of the right partition is associated with the corresponding second theoretical receiving center position. Combined with the Euclidean distance algorithm, the left center offset distance and right center offset distance of each monitoring point are determined respectively.

[0103] During the gripper grasping phase, for any monitoring point other than the first monitoring point, the left and right center offset distances of the corresponding monitoring point are subtracted from the left and right center offset distances of the previous monitoring point. These are used as the left and right center offset unit change distances of the corresponding monitoring point relative to the previous monitoring point. The left and right center offset unit change distances are used to characterize the left and right center offset levels of the corresponding monitoring point relative to the previous monitoring point during the gripper grasping phase.

[0104] Specifically, the generation of capture level control instructions for corresponding monitoring points includes:

[0105] Based on the magnitudes of the changes in left and right center offsets of the corresponding monitoring point compared to the previous monitoring point, a capture level control instruction is generated for the corresponding monitoring point, specifically including:

[0106] If the left center offset unit distance and the right center offset unit distance of the corresponding monitoring point are both not equal to zero compared to the previous monitoring point, it indicates that the gripper is in a bidirectional unstable state when grasping the target workpiece at the corresponding monitoring point, and a first-level gripping control command is generated for the corresponding monitoring point. If the left center offset unit distance or the right center offset unit distance of the corresponding monitoring point is equal to zero compared to the previous monitoring point, it indicates that the gripper is in a unidirectional unstable state when grasping the target workpiece at the corresponding monitoring point, and a second-level gripping control command is generated for the corresponding monitoring point. If the left center offset unit distance and the right center offset unit distance of the corresponding monitoring point are both equal to zero compared to the previous monitoring point, it indicates that the gripper is in a stable gripping state when grasping the target workpiece at the corresponding monitoring point, and a third-level gripping control command is generated for the corresponding monitoring point.

[0107] S5. Based on the generated capture level control command for the corresponding monitoring point, execute the corresponding capture control operation, specifically including:

[0108] If the gripping level control instruction for the corresponding monitoring point is a Level 1 gripping control instruction, the execution content is: increase the clamping torque of the gripper at the next monitoring point to the target workpiece by two levels; if the gripping level control instruction for the corresponding monitoring point is a Level 2 gripping control instruction, the execution content is: increase the clamping torque of the gripper at the next monitoring point to the target workpiece by one level; if the gripping level control instruction for the corresponding monitoring point is a Level 3 gripping control instruction, the execution content is: continue to maintain the clamping torque level of the gripper at the current monitoring point to the target workpiece, and continuously monitor the offset of the left and right bearing center positions during the movement of the gripper to the target workpiece.

[0109] It should be noted that in precision electronic component handling scenarios, the grippers are driven by servo motors or electric push rods, and their output torque corresponds to the drive current or displacement control quantity. During the initialization phase, the gripping torque is pre-divided into several discrete levels, and a mapping table between level numbers and drive control quantities is established. When a gripping level control command for the corresponding monitoring point is generated, the controller reads the corresponding control quantity from the mapping table based on the current level number and switches levels according to the command requirements. For example, a level one gripping control command corresponds to switching up two levels from the current level, that is, increasing the drive current or push rod target displacement to the corresponding level value, thereby causing the gripper to output a larger gripping torque. The level two gripping control command executes a single-level increment. The level three gripping control command keeps the current level unchanged and only continuously outputs the current drive control quantity. During execution, the controller performs closed-loop verification of the actual output state through displacement encoders and current feedback signals, ensuring that the adjustment of the gripping torque is consistent with the control command of the monitoring point, and continues to participate in tracking the center offset state at the next monitoring point.

[0110] This application also provides a closed-loop force control system for a high-precision flexible gripper, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the above-described high-precision flexible gripper grasping method.

[0111] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0112] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0113] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0114] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0115] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0116] In conclusion, 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 gripping method using a high-precision flexible gripper, characterized in that: Includes the following steps: S1. Obtain several contact zones on the finger surfaces of the flexible gripper and deploy multiple sets of sensors. During the gripper's stationary phase, analyze the proximity of the corresponding contact zones to the boundary points of each contour within the continuous contour and construct a contact baseline dataset. The contact baseline dataset includes the theoretical contact zone sequence, the first theoretical bearing center position, the second theoretical bearing center position, the theoretical synchronization difference sequence, the pressure noise threshold, and the deformation noise threshold. S2. During the gripper closing phase, based on the contact baseline dataset, contact state partitions are selected, and the first contact time point of each contact state partition is determined. The actual contact partition sequence is constructed, and a synchronization deviation analysis command is issued after comparative analysis. S3. After receiving the synchronization deviation analysis command, analyze the degree of left and right first touch synchronization deviation according to the theoretical synchronization difference sequence, and issue the load offset analysis command. S4. After receiving the load-bearing offset analysis command, during the gripper grasping stage, based on the first theoretical load-bearing center position and the second theoretical load-bearing center position, analyze the level of change of the left center offset unit and the right center offset unit of the corresponding monitoring point compared with the previous monitoring point, and generate the corresponding gripping level control command for the monitoring point. S5. Execute the corresponding capture control operation according to the capture level control command generated for the corresponding monitoring point.

2. The gripping method of a high-precision flexible gripper according to claim 1, characterized in that: The flexible gripper's finger surfaces are uniformly divided to obtain several contact zones, and multiple sets of sensors are deployed, specifically including: An industrial camera is set above the gripping station and a side camera is set in front of the gripper. A displacement encoder is set on the gripper drive slider. Based on the flexible contact layer of the left and right finger surfaces of the gripper, the flexible contact layer is evenly divided along the length and width directions of the finger surfaces to obtain several contact partitions. Thin film pressure sensors are embedded on the surface of each contact partition and flexible strain sensors are attached to the corresponding bottom layer.

3. The gripping method of a high-precision flexible gripper according to claim 1, characterized in that: Analyze the proximity of the corresponding contact zone to the boundary points of each contour within the continuous contour, specifically including: During the gripper stationary phase, top view and side view contour images of the target workpiece to be gripped are acquired using an industrial camera and a side camera. After continuous boundary recognition, the continuous contour of the target workpiece to be gripped is determined. Using the midpoint of the line connecting the geometric center points of the left and right finger surfaces of the gripper as the origin, a three-dimensional coordinate system is established. After the gripper assembly calibration process, the center coordinate positions of each contact zone are obtained. Each contact zone is assigned a unique direction label number, which includes the left side number and the right side number. Based on the established three-dimensional coordinate system, the point coordinates of each boundary point within the continuous contour are extracted. Based on the projection area of ​​the continuous contour vertically projected onto the left and right finger surfaces of the gripper, and combined with the center coordinates of each contact zone, the center coordinates of each contact zone within the projection area are extracted. For any contact zone within the projection area, analyze the proximity of the corresponding contact zone to the boundary points of each contour within the continuous contour, and determine the pre-contact distance of the corresponding contact zone. The pre-contact distances of each contact zone are matched with the directional label numbers of the corresponding contact zones. Based on the magnitude of the pre-contact distances, the contact zones are sorted in ascending order to construct a theoretical contact zone sequence. Each contact zone in the theoretical contact zone sequence corresponds to a unique directional label number and a pre-contact distance.

4. The gripping method of a high-precision flexible gripper according to claim 1, characterized in that: The locations of the first and second theoretical centers are specifically included as follows: Based on the direction label numbers of each contact partition in the theoretical contact partition sequence, the left contact partition is paired with the corresponding right contact partition to obtain several partition alignment groups. The left pre-contact distance and the right pre-contact distance in each partition alignment group are summed to obtain the unit pre-contact distance of each partition alignment group. Combined with the statistical mean calculation algorithm, the average pre-contact distance is determined. The difference between the unit pre-contact distance and the average pre-contact distance of each partition alignment group is calculated. The center coordinate positions of the left contact partition and the right contact partition in the partition alignment group with the smallest difference are extracted and used as the first theoretical bearing center position and the second theoretical bearing center position, respectively.

5. The gripping method of a high-precision flexible gripper according to claim 1, characterized in that: Theoretical synchronization difference sequence, pressure noise threshold, and deformation noise threshold, specifically including: For any partition alignment group, extract the left pre-contact distance and the right pre-contact distance in the corresponding partition alignment group, analyze the pre-contact synchronization level of the corresponding partition alignment group, and determine the theoretical synchronization difference of the corresponding partition alignment group; Based on the determined theoretical synchronization difference values ​​of each partition pair, the corresponding theoretical synchronization differences are sorted in ascending order to construct a theoretical synchronization difference sequence. By embedding thin-film pressure sensors on the surface of each contact zone and attaching flexible strain sensors in the corresponding underlying layer, the unloaded pressure value and unloaded compression deformation value of each contact zone are collected respectively, and used as the pressure noise threshold and deformation noise threshold of the corresponding contact zone. A contact baseline dataset is constructed based on the theoretical contact partition sequence, the first theoretical contact center location, the second theoretical contact center location, the theoretical synchronization difference sequence, the pressure noise threshold, and the deformation noise threshold.

6. The gripping method of a high-precision flexible gripper according to claim 1, characterized in that: Determine the first contact time point for each contact state zone, specifically including: During the gripper closing phase, sampling points are set with a preset first time step, and the pressure and deformation values ​​of each contact zone at different sampling points are collected in real time through an embedded thin-film pressure sensor and a attached flexible strain sensor. According to the time sequence of the sampling points, the unit change zone pressure value and unit change zone deformation value of the corresponding contact zone at the corresponding sampling point are compared and analyzed with the pressure noise threshold and deformation noise threshold of the corresponding contact zone. If the unit change zone pressure value at the corresponding sampling point is greater than the pressure noise threshold and the unit change zone deformation value is greater than the deformation noise threshold, then the corresponding contact zone is recorded as a contact state zone, and the first time point that meets the comparison conditions is recorded as the first contact time point of the corresponding contact state zone. Otherwise, the corresponding contact zone is recorded as a contact idle zone.

7. The gripping method of a high-precision flexible gripper according to claim 6, characterized in that: Following the comparative analysis, a synchronization deviation analysis command is issued, which specifically includes: Based on the first contact time of each contact state partition, the contact state partitions are sorted in chronological order to construct the actual contact partition sequence. The actual contact sequence value of each contact state partition in the actual contact partition sequence is extracted. Feature recognition is performed on the theoretical contact partition sequence to extract the theoretical contact sequence value of each contact state partition. The theoretical contact sequence and actual contact sequence of each contact state partition are compared and analyzed. If the actual contact sequence value of the corresponding contact state partition exceeds the theoretical contact sequence value, the corresponding contact state partition is recorded as a misaligned contact state partition, and the number of misaligned contact state partitions is counted. If the actual contact sequence value of the corresponding contact state partition does not exceed the theoretical contact sequence value, the corresponding contact state partition is recorded as a normal contact state partition, and the number of normal contact state partitions is counted. The ratio of the number of misaligned contact state partitions to the total number of contact state partitions is calculated to obtain the misalignment ratio. When the misalignment ratio exceeds 5%, a synchronization deviation analysis command is issued; otherwise, no additional synchronization deviation analysis command is issued.

8. The gripping method of a high-precision flexible gripper according to claim 7, characterized in that: Analyze the degree of left and right first touch synchronization deviation, specifically including: After receiving the synchronization deviation analysis command, the partition alignment group where the misaligned contact state partition is located is recorded as the partition comparison alignment group. Based on the first contact time point of each contact state partition, the first contact time point of the left contact partition and the first contact time point of the right contact partition in each partition comparison alignment group are extracted. The degree of left and right first contact synchronization deviation in the corresponding partition comparison alignment group is analyzed to determine the actual first contact time difference of the corresponding partition comparison alignment group. Feature identification is performed on the theoretical synchronization difference sequence to extract the theoretical synchronization difference value of each partition comparison alignment group. Combined with the average advancing speed of the gripper before the first contact, the theoretical first contact time difference of each partition comparison alignment group is determined. The actual first contact time difference of each partition comparison alignment group is compared and analyzed with the corresponding theoretical first contact time difference, specifically including: If the actual first contact time difference of the corresponding partition comparison alignment group is equal to the corresponding theoretical first contact time difference, the synchronization status of the corresponding partition comparison alignment group is recorded as normal synchronization status. If the actual first contact time difference of the corresponding partition comparison alignment group is not equal to the corresponding theoretical first contact time difference, the synchronization status of the corresponding partition comparison alignment group is recorded as abnormal synchronization status. The synchronization status of all partition comparison alignment groups is summarized. If there is an abnormal synchronization status among the synchronization statuses of all partition comparison alignment groups, a load-bearing offset analysis instruction is generated. If there is no abnormal synchronization status among the synchronization statuses of all partition comparison alignment groups, no additional load-bearing offset analysis instruction is generated.

9. The gripping method of a high-precision flexible gripper according to claim 8, characterized in that: Analyze the changes in left and right center offset units at the corresponding monitoring points compared to the previous monitoring point, specifically including: Upon receiving the load-bearing offset analysis command, during the gripper grasping stage, monitoring points are set with a preset second time step. Through embedded thin-film pressure sensors and attached flexible strain sensors, the load-bearing status of the left and right contact zones within each partition comparison alignment group is monitored in real time. The partition clamping pressure value and partition clamping deformation value of the left and right contact zones within the corresponding partition comparison alignment group at each monitoring point are obtained. The partition clamping pressure value and partition clamping deformation value are weighted and summed to obtain the load-bearing degree coefficient of the left and right contact zones within the corresponding partition comparison alignment group at each monitoring point. After feature recognition, the maximum value of the load-bearing degree coefficient of the left contact zone and the maximum value of the load-bearing degree coefficient of the right contact zone at each monitoring point are extracted respectively. The center coordinates of the contact zones corresponding to the maximum bearing capacity coefficient of the left and right contact zones of each monitoring point are taken as the actual bearing center positions of the left and right contact zones of each monitoring point. The actual receiving center position of the left partition of each monitoring point is associated with the corresponding first theoretical receiving center position, and the actual receiving center position of the right partition is associated with the corresponding second theoretical receiving center position. Combined with the Euclidean distance algorithm, the left center offset distance and right center offset distance of each monitoring point are determined respectively. During the gripper grasping phase, for any monitoring point other than the first monitoring point, the left and right center offset distances of the corresponding monitoring point are subtracted from the left and right center offset distances of the previous monitoring point. These are used as the left and right center offset unit change distances of the corresponding monitoring point relative to the previous monitoring point. The left and right center offset unit change distances are used to characterize the left and right center offset levels of the corresponding monitoring point relative to the previous monitoring point during the gripper grasping phase.

10. A closed-loop force control system for a high-precision flexible gripper, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the grasping method according to any one of claims 1 to 9.

Images