Vacuum suction cup grabbing control method and device

Abstract

The application discloses a kind of vacuum suction head grabbing control method and device, the method includes: real-time acquisition vacuum pressure data of vacuum suction head, and simultaneously acquisition touch data;When touch data variation exceeds contact threshold, it is judged that vacuum suction head contacts object;In response to contact, vacuum pressure data and touch data are fused and analyzed, to determine the current grabbing state;According to current grabbing state, control instruction is generated, for adjusting adsorption parameter and / or motion parameter.The device includes vacuum suction head, touch sensor integrated on the suction head, vacuum pressure sensor and controller.The application solves the problem that current vacuum suction head cannot perceive contact state, is easy to misjudge, and is easy to damage flexible object, realizes the real-time perception and closed-loop control of contact state, collapse, wrinkle, leakage and other abnormalities in grabbing process, significantly improves the success rate and safety of grabbing.

Description

Technical Field

[0001] This invention relates to the field of vacuum adsorption technology, and in particular to a vacuum suction head gripping control method and a vacuum suction head gripping device. Background Technology

[0002] Vacuum adsorption technology is a commonly used technique in automation for object gripping and handling. It creates negative pressure inside the suction head, causing it to adhere to the surface of the target object, thus achieving a gripping operation. In industries such as electronics manufacturing, food packaging, logistics sorting, and textile processing, vacuum adsorption is often the preferred gripping method when handling flexible objects such as films, flexible packaging, fabrics, and sponges.

[0003] Currently, vacuum adsorption technology commonly uses vacuum suction heads to grasp flexible objects. A typical structure includes a flexible suction cup body, a vacuum generator, and a pressure sensor in the tubing. During operation, the vacuum generator creates negative pressure, causing the suction cup to adsorb the object. The pressure sensor in the tubing monitors the vacuum level inside the suction head. When the negative pressure reaches a set threshold, the system determines that adsorption is successful and performs the transport action.

[0004] However, the above-mentioned solutions have the following limitations in practical applications: First, the failure rate of adsorption is high; flexible objects are prone to local collapse, air leakage, and poor sealing during adsorption, resulting in unstable adsorption. Second, the contact state cannot be sensed; pure vacuum suction heads can only infer whether contact has occurred through vacuum pressure, but cannot sense in real time whether the contact position is correct, whether the flexible object has collapsed, whether there are wrinkles or foreign objects, or whether secondary adhesion or slippage has occurred. Third, the target object is easily damaged; flexible objects are prone to wrinkles, deformation, or even breakage due to over-adsorption. Fourth, there is a lack of closed-loop force control capability; vacuum suction heads lack tactile feedback and cannot adjust the suction force or contact strategy based on the contact force. Therefore, current vacuum adsorption technology struggles to simultaneously meet the requirements of high success rate, adaptability to flexible shapes, stability, and safety. Summary of the Invention

[0005] This invention provides a vacuum suction head gripping control method and a vacuum suction head gripping device, which solves the problems of current vacuum suction heads being unable to truly sense the contact state, being prone to misjudgment, and being prone to damaging flexible objects. It realizes real-time perception and closed-loop control of abnormalities such as contact state, collapse, wrinkles, and leakage during the gripping process, which significantly improves the gripping success rate and safety.

[0006] This invention provides a vacuum suction head gripping control method, the method comprising: Real-time acquisition of vacuum pressure data from the vacuum suction head, and simultaneous acquisition of tactile data; When the change in the tactile data exceeds a preset contact threshold, it is determined that the vacuum nozzle is in contact with the object; In response to the contact, the vacuum pressure data and the tactile data are fused and analyzed to determine the current gripping state of the vacuum suction head; Based on the current grasping state, a control command is generated, which is used to adjust the adsorption parameters of the vacuum head and / or the motion parameters of the vacuum head.

[0007] Optionally, the step of fusing and analyzing the vacuum pressure data and the tactile data to determine the current gripping state of the vacuum suction head includes: When the vacuum pressure data does not reach the preset negative pressure threshold and the tactile data shows that contact force has been generated, the current grasping state is determined to be poor sealing or leakage. When the vacuum pressure data reaches the preset negative pressure threshold, and the tactile data shows that the pressure distribution is uneven or there is a sudden change in local pressure, the current grasping state is determined to be a local collapse or a wrinkled surface. When the vacuum pressure data is stable within a preset range and the tactile data shows a uniform pressure distribution, the current grasping state is determined to be complete contact.

[0008] Optionally, the step of fusing and analyzing the vacuum pressure data and the tactile data to determine the current gripping state of the vacuum suction head further includes: When a local collapse or surface wrinkling is determined, a real-time pressure distribution map is constructed based on the pressure values ​​collected by multiple sensing units. Calculate the offset of the pressure center or the variance of the pressure values ​​in the pressure distribution map; Based on the offset or the variance, the degree of collapse or the uniformity of fit of the object is quantified, and the degree of collapse or the uniformity of fit is used as the deformation parameter of the current gripping state.

[0009] Optionally, the step of generating control commands based on the current grasping state includes: When the current gripping state is a partial collapse or the surface has wrinkles, reduce the rate of increase of vacuum suction, pause adsorption, or adjust the suction head posture; When the current gripping state is poor sealing or leakage, control the vacuum suction head to re-perform the contact action or adjust the contact position; Once the current grasping state is complete and the grasping is successful, the current adsorption parameters are maintained and the object is transported, while the changes in tactile data are continuously monitored during the transport process.

[0010] Optionally, the step of continuously monitoring changes in tactile data during handling includes: When a continuous moving trajectory is detected at the center of contact force, the sliding direction and sliding speed of the object are predicted. Based on the sliding direction and sliding speed, a motion compensation command is generated to adjust the movement speed and / or movement direction of the vacuum suction head.

[0011] Optionally, the step of adjusting the adsorption parameters and / or motion parameters of the vacuum head includes: In the initial stage of the contact action, the actual contact position coordinates of the vacuum nozzle are corrected based on the moment when the contact force signal is first detected. During the adsorption process, the output power of the vacuum generator in the vacuum head is adaptively adjusted based on the trend of contact force changes fed back by tactile data, so that the contact force is maintained within the preset safety threshold range.

[0012] Optionally, the method further includes: Record the vacuum pressure data and tactile data during each grasping process, as well as the corresponding grasping results; Based on the grasping results, the grasping parameters for similar objects are optimized and updated. The grasping parameters include contact speed, suction growth rate, and / or lifting speed.

[0013] Furthermore, embodiments of the present invention also provide a vacuum suction head gripping device, the device comprising: A vacuum suction head, located at the front end of the end effector, is used to apply negative pressure to an object to generate suction force; A tactile sensor is integrated into the adsorption working surface of the vacuum head or the peripheral structure adjacent to the adsorption working surface, and is used to collect tactile signals at the interface between the object and the vacuum head in real time. A vacuum pressure sensor is connected to the air path of the vacuum nozzle and is used to monitor the vacuum pressure data in the air path in real time. The controller is electrically connected to both the tactile sensor and the vacuum pressure sensor. The controller is configured to perform the following operations: Receive the tactile signal and the vacuum pressure data; Based on the fusion result of the tactile signal and the vacuum pressure data, the current grasping state is determined; Based on the current grasping state, control commands are generated to adjust the adsorption parameters and / or motion parameters of the vacuum head.

[0014] Optionally, the tactile sensing module is one of a flexible pressure sensing array, a capacitive tactile sensing membrane, or a distributed pressure sensor. The sensing area of ​​the tactile sensor covers the central and edge areas of the working surface of the vacuum suction head; The vacuum suction head has a deformable structure and includes a silicone suction cup, a flexible diaphragm suction head, or a multi-chamber suction head. The tactile sensor is attached to the inner or outer surface of the deformable structure and deforms synchronously with the deformation of the vacuum suction head.

[0015] Optionally, the device further includes a vacuum generator, and the controller is connected to the vacuum generator; The controller is also configured to: When the current gripping state is detected to include partial collapse or surface wrinkles, an output command is sent to control the vacuum generator to reduce the pumping rate or stop pumping. When the current gripping state is detected to include poor sealing or leakage, an instruction is output to control the vacuum head to re-perform the contact action or adjust the contact position.

[0016] One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages: 1. By collecting tactile data in real time and using its changes as the basis for contact judgment, the problem that related technologies cannot accurately perceive the actual contact moment by relying solely on the robot's preset position is solved; at the same time, by integrating vacuum pressure and tactile distribution data, different abnormal states such as poor sealing, local collapse, and surface wrinkles can be accurately distinguished, thus solving the defect of related technologies that cannot truly perceive contact quality by relying solely on vacuum pressure.

[0017] 2. By constructing a pressure distribution map and calculating the pressure center offset or pressure value variance, qualitative phenomena such as the degree of collapse and the uniformity of adhesion are quantified into calculable deformation parameters. Differentiated control commands are generated for different gripping states. When there is collapse or wrinkling, the suction growth rate is reduced or the suction head posture is adjusted. When there is leakage, the contact action is re-executed or the contact position is adjusted, thus avoiding efficiency loss and object damage caused by blind operation.

[0018] 3. By continuously monitoring tactile data during the handling process, the sliding direction and speed are predicted and motion compensation commands are generated when a continuous moving trajectory is detected at the center of contact force. This solves the problem of the inability to correct object slippage in real time during handling. At the same time, the actual contact position coordinates are corrected based on the initial contact force signal at the beginning of contact. During the adsorption process, the output power of the vacuum generator is adaptively adjusted based on the trend of contact force changes to keep the contact force within a safe threshold range. This achieves closed-loop control of the entire process from contact position correction to dynamic adjustment of adsorption force, effectively avoiding damage to objects or failure to grasp due to excessive or insufficient suction force.

[0019] 4. By recording the data and results of each grasping process, the grasping parameters for similar objects are optimized and updated, enabling the system to have self-learning capabilities. At the same time, the tactile sensor is integrated into the inner or outer surface of the deformable suction head and deforms synchronously with it, realizing real-time capture of micro-deformation during the adsorption of flexible objects. Both hardware and control aspects ensure the continuous improvement of the grasping success rate. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the framework of the vacuum suction head gripping control method of the present invention; Figure 2 This is a flowchart illustrating the vacuum suction head gripping control method of the present invention; Figure 3 This is a schematic diagram of the overall structure of the vacuum suction head gripping control device of the present invention; Figure 4 This is a partial schematic diagram of the vacuum suction head gripping control device of the present invention; Figure 5 This is a schematic diagram of the terminal structure of the hardware operating environment involved in an embodiment of the present invention. Detailed Implementation

[0021] To address the problem that vacuum suction heads, relying solely on vacuum pressure monitoring, cannot accurately perceive the contact state with objects, leading to difficulties in identifying abnormal conditions such as collapse, wrinkles, and leaks when grasping flexible objects, and resulting in misjudgments, grasping failures, or object damage, this invention provides a vacuum suction head grasping control method and device. By integrating a tactile sensor into the vacuum suction head, tactile data is collected in real time and fused with vacuum pressure data for analysis. Based on this, the current grasping state is determined, and corresponding control commands are generated to dynamically adjust adsorption or motion parameters. This invention achieves real-time perception and closed-loop control of information such as contact state, degree of collapse, and slippage trend during the grasping process, accurately identifying and responding to different abnormal conditions, significantly improving the success rate, safety, and adaptability of grasping flexible objects.

[0022] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the invention to those skilled in the art.

[0023] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0024] Example 1 In this embodiment, a vacuum suction head gripping control method is provided.

[0025] Reference Figure 1 and Figure 2 The vacuum suction head gripping control method of this embodiment includes the following steps: Step S1: Collect vacuum pressure data of the vacuum nozzle in real time, and collect tactile data at the same time; In this embodiment, vacuum pressure data refers to the negative pressure value in the air path inside the vacuum nozzle, which is obtained in real time by a vacuum pressure sensor connected to the air path and is used to determine the degree of sealing between the nozzle and the object and the magnitude of the adsorption force.

[0026] Tactile data refers to contact information collected by tactile sensors integrated into the vacuum suction head, including at least the magnitude of contact force and pressure distribution. In this embodiment, the tactile sensor may be a flexible pressure sensing array, a capacitive tactile sensing membrane, or a distributed pressure sensor, and its sensing area covers the central and edge areas of the suction head's adsorption working surface.

[0027] As the vacuum nozzle moves toward the object, the control system continuously collects vacuum pressure and tactile data. Vacuum pressure data reflects the airflow status, while tactile data reflects the contact status between the nozzle and the object. These two types of data are collected simultaneously, providing real-time input for subsequent determination of the contact moment and fusion analysis.

[0028] Optionally, the control system can simultaneously acquire two types of data at a fixed sampling frequency to ensure accurate correspondence between them in the time dimension. By simultaneously acquiring vacuum pressure and tactile data, a data foundation is provided for subsequent accurate perception of the contact state.

[0029] Step S2: When the change in the tactile data exceeds a preset contact threshold, it is determined that the vacuum nozzle is in contact with the object; In this embodiment, the change in tactile data refers to the amplitude of change in the signal output by the tactile sensor per unit time. When the suction head is not in contact with an object, the tactile data is usually zero or close to zero noise value; when the suction head contacts the object, the tactile signal will change from zero to a step or a rapid increase.

[0030] The preset contact threshold refers to a pre-set critical value used to distinguish between contact and non-contact states. This threshold can be calibrated according to the sensitivity of the tactile sensor and the application scenario, and is usually slightly higher than the noise level to avoid false triggering.

[0031] This embodiment utilizes tactile data to directly perceive contact events, accurately obtaining the time and position of contact between the vacuum nozzle and the object without the need for robot positioning. This solves the problem of misjudgment of contact caused by robot positioning errors or fluctuations in object position, enabling the vacuum nozzle to autonomously perceive the moment of contact and providing precise triggering timing for subsequent fusion analysis.

[0032] Step S3: In response to the contact, the vacuum pressure data and the tactile data are fused and analyzed to determine the current gripping state of the vacuum suction head; In this embodiment, vacuum pressure reflects the airtightness and adsorption force of the gas path, while tactile data reflects the contact position, pressure distribution, and contact force magnitude. The synchronously collected vacuum pressure and tactile data are fused together to analyze whether the vacuum pressure reaches a preset negative pressure threshold and whether it is stable; and to analyze whether the pressure distribution in the tactile data is uniform and whether there are local pressure abrupt changes. By cross-referencing these two types of data, complex operating conditions that cannot be determined by a single data point are identified.

[0033] As an optional implementation, by fusing tactile feedback with vacuum pressure analysis, accurate identification of various abnormal conditions during the grasping process is achieved, solving the problem that related technologies cannot distinguish between different states such as poor sealing, local collapse, and surface wrinkles using only vacuum pressure. Specifically, the following steps are included: Step S301: When the vacuum pressure data does not reach the preset negative pressure threshold and the tactile data shows that contact force has been generated, it is determined that the current grasping state is poorly sealed or there is a leak. In this embodiment, the preset negative pressure threshold refers to a pre-set minimum negative pressure value used to determine whether adsorption is successful. When the vacuum pressure reaches this threshold, it indicates that an effective seal has been formed between the suction head and the object, and sufficient adsorption force is available. When the tactile sensor detects contact force, but the vacuum pressure does not reach the preset negative pressure threshold, it indicates that an effective seal has not been formed between the vacuum suction head and the object.

[0034] For example, when grasping a soft package with a slightly curved surface, after the vacuum suction head descends and makes contact, the tactile sensor shows that there is contact force in the central area, but the pressure value in the edge area is low; at the same time, the vacuum pressure sensor shows that the negative pressure value reaches -15kPa, which is far below the preset -40kPa threshold. Therefore, it is determined that the seal is not good, and the subsequent execution of the control command in step S4 can generate a command to re-execute the contact action or adjust the contact position.

[0035] Step S302: When the vacuum pressure data reaches the preset negative pressure threshold and the tactile data shows that the pressure distribution is uneven or there is a sudden change in local pressure, it is determined that the current grasping state is a local collapse or that there are wrinkles on the surface. In this embodiment, the preset negative pressure range refers to a pre-defined vacuum pressure fluctuation range used to determine whether adsorption is successful. When the vacuum pressure remains stable within the preset range, it indicates good adsorption; if the pressure exceeds the range, an anomaly may exist. Local pressure abrupt changes refer to a situation where the pressure value in a certain area of ​​the tactile data is significantly lower than that of the surrounding areas. For example, the pressure value output by the sensing unit corresponding to a depression is close to zero, while the pressure in the surrounding area is normal. This pressure distribution characteristic can accurately locate the location of depressions or wrinkles on the surface of an object.

[0036] Optionally, if the vacuum pressure data reaches the preset negative pressure threshold, it indicates that the air path is well sealed. If the tactile data shows abnormal pressure distribution, it indicates that there is a problem with the contact interface between the vacuum nozzle and the object. If there is a high-pressure area with sudden local pressure changes, it is usually because the object is sucked into the suction port, causing local collapse. If there is a low-pressure area with a pressure value close to zero, it is usually because there are wrinkles or foreign objects on the surface of the object, causing local suspension.

[0037] For example, when grasping a thin-walled sponge, after the vacuum head contacts and initiates adsorption, the vacuum pressure quickly reaches -45 kPa and stabilizes within the preset range of -42 kPa to -48 kPa, indicating good air path sealing. However, the tactile sensor shows that the output pressure values ​​of the three sensing units located slightly to the left of the adsorption center are close to 0, while the output pressure values ​​of the surrounding units are between 0.6 N and 0.8 N. The pressure distribution map shows a clear zero-value depression area. Therefore, it is determined to be a local collapse, and the subsequent execution of the quality control generation in step S4 can generate instructions to reduce the growth rate of vacuum suction, pause adsorption, or adjust the attitude of the suction head.

[0038] Step S303: When the vacuum pressure data is stable within a preset range and the tactile data shows that the pressure distribution is uniform, the current grasping state is determined to be complete contact.

[0039] In this embodiment, when the vacuum pressure data is stable within a preset range, it indicates good sealing and adsorption force recombination. Uniform pressure distribution means that the pressure values ​​in different areas of the tactile sensor are similar, and the entire contact surface exhibits a uniform force. Specifically, this is manifested in the output pressure values ​​of all sensing units being at the same level, with no obvious high-pressure or low-pressure areas, and the pressure distribution map showing a smooth transition. This distribution indicates that the suction head adheres well to the object, without collapse, wrinkles, or foreign objects. The combination of these two factors indicates complete contact, allowing the process to proceed to the handling stage.

[0040] For example, when gripping a smooth, rigid cardboard box, the vacuum pressure stabilizes within a preset range of -48 kPa to -52 kPa after the vacuum head makes contact, with fluctuations less than ±2 kPa. The pressure values ​​in each area of ​​the tactile sensor are between 0.5 N and 0.8 N, and the pressure distribution map shows a uniform distribution with a slightly higher center and smooth transitions at the edges. Therefore, it is determined that the contact is complete, and an instruction to maintain the current adsorption parameters and perform the handling can be generated.

[0041] This implementation establishes a two-dimensional judgment rule based on vacuum pressure and tactile distribution, categorizing complex working conditions during the grasping process into three typical states. This enables the control system to quickly and accurately determine the current grasping state. When grasping flexible objects, due to their uncertain shape, susceptibility to deformation, and potential surface wrinkles, the tactile sensor perceives the contact force and pressure distribution in real time. This data is then fused and analyzed with vacuum pressure data to effectively identify hidden anomalies such as false adsorption and false contact. This achieves accurate identification and stable grasping of flexible objects, significantly improving the reliability and safety of the grasping process.

[0042] Furthermore, by introducing a preset negative pressure range instead of a single threshold judgment, the assessment of the adsorption state is made more consistent with actual working conditions, avoiding misjudgments caused by normal system fluctuations. By identifying the pressure distribution characteristics of local zero-value areas, precise location of collapses and wrinkles is achieved. The uniform force distribution characteristics across the entire contact surface ensure reliable identification of the contact integrity state. These judgment rules enable this embodiment to achieve higher recognition accuracy and positioning precision for various abnormal working conditions during the grasping process of flexible objects.

[0043] As another optional implementation, based on the determination that there is local collapse or surface wrinkles, the degree of anomaly is further quantitatively analyzed to achieve a refined assessment of the degree of object deformation. Specifically, this includes the following steps: Step S310: When it is determined to be a local collapse or surface wrinkling, a real-time pressure distribution map is constructed based on the pressure values ​​collected by multiple sensing units. In this embodiment, based on the identification of localized collapse or surface wrinkles, the abnormal areas are visualized. Tactile sensors typically comprise multiple independent sensing units arranged in an array, each outputting the pressure value at its location. Real-time data from all sensing units is acquired and mapped onto a two-dimensional plane to form a real-time pressure distribution map reflecting the pressure distribution characteristics of the contact interface. This distribution map can intuitively display the location and extent of high-pressure and low-pressure areas.

[0044] For example, after determining that wrinkles exist on the surface in step S302, the pressure values ​​of the 32 sensing units of the tactile sensor are retrieved to generate a pressure distribution map. The map shows two low-pressure areas where the pressure values ​​of adjacent units are close to zero, corresponding to the suspended areas formed by the thin film wrinkles.

[0045] Step S320: Calculate the offset of the pressure center or the variance of the pressure values ​​in the pressure distribution map; In this embodiment, the pressure center refers to the equivalent point of application of the contact pressure field. Its coordinates are calculated by weighted averaging of the pressure values ​​of each sensing unit, reflecting the spatial distribution center of the contact pressure. The pressure center offset is the distance between this equivalent point of application and the geometric center of the vacuum head. The magnitude of the offset indicates the degree of symmetry in the force distribution of the vacuum head. A larger offset indicates a more uneven force distribution on the head, indirectly reflecting the severity of local collapse of the object or tilting of the head. The variance of the pressure values ​​represents the degree of dispersion of the pressure values ​​of each sensing unit relative to the average pressure. A larger variance indicates a more significant spatial difference in the pressure distribution, i.e., a poorer uniformity of contact interface adhesion, which may be caused by wrinkles, foreign objects, or local suspension on the object surface.

[0046] Step S330: Based on the offset or the variance, quantify the degree of collapse or the uniformity of adhesion of the object, and use the degree of collapse or the uniformity of adhesion as the deformation degree parameter of the current grasping state.

[0047] In this embodiment, the calculated offset or variance is converted into interpretable deformation parameters. Offset thresholds and variance thresholds can be preset to classify the degree of collapse into levels such as slight, moderate, and severe, or to measure the uniformity of fit as a percentage value from 0 to 100%. These deformation parameters are added to the current grasping state to provide a basis for subsequent control strategies.

[0048] This implementation method, based on the identification of local collapse or surface anomalies, transforms abstract qualitative phenomena such as collapse and wrinkles into quantifiable deformation parameters by constructing a pressure distribution map and calculating the pressure center offset or pressure value variance. Even if the object being grasped is flexible, its easily deformable and uneven surface makes it difficult to distinguish the severity of anomalies based solely on qualitative judgment. However, this implementation method, through quantitative analysis, can accurately assess the severity of collapse and the uniformity of adhesion, enabling the control system to dynamically adjust its control strategy according to the degree of anomaly. For example, it can reduce the suction growth rate only in the case of minor collapse, and pause adsorption and adjust the posture in the case of severe collapse. This achieves refined and differentiated grasping control of flexible objects, further improving the success rate and safety of grasping.

[0049] Step S4: Based on the current grasping state, generate control commands, which are used to adjust the adsorption parameters of the vacuum head and / or the motion parameters of the vacuum head.

[0050] In this embodiment, based on the current grasping state determined in step S3, corresponding control commands are generated to adjust the adsorption parameters or motion parameters in real time to optimize the grasping process or cope with abnormal working conditions. By adjusting the adsorption and motion parameters in real time according to the grasping state, the vacuum suction head has dynamic adaptability and can autonomously adjust its behavior strategy according to different working conditions.

[0051] As an optional implementation, based on the current grasping state determined in step S3, differentiated control commands are generated to achieve precise responses to different abnormal operating conditions. Specifically, this includes the following steps: Step S410: When the current gripping state is a partial collapse or the surface has wrinkles, reduce the growth rate of vacuum suction, pause adsorption, or adjust the suction head posture. In this embodiment, when a partial collapse is determined, it indicates that the object has been excessively sucked into the suction port. If the suction force is increased further, the collapse will be aggravated or even damaged. Therefore, the suction force growth rate is reduced to slow down the adsorption process, or the adsorption is paused to allow the object to bounce back and reset. When a wrinkle is determined to be present on the surface, it indicates that the suction head does not fit well with the object. In this case, adjusting the suction head posture can change the contact angle and allow the suction head to better fit the object surface.

[0052] Step S420: When the current gripping state is poor sealing or leakage, control the vacuum suction head to re-perform the contact action or adjust the contact position; In this embodiment, when a poor seal is determined, it indicates that the vacuum nozzle and the object have failed to form an effective seal. At this time, re-performing the contact action can allow the nozzle to make contact in a new posture, or adjust the contact position to move to other areas on the object surface to avoid the leakage point.

[0053] Step S430: When the current grasping state is complete contact and successful grasping, maintain the current adsorption parameters and carry out the transfer, and continuously monitor the changes in tactile data during the transfer process.

[0054] In this embodiment, when complete contact is determined, it indicates a good grasping state. The control system maintains the current adsorption parameters, such as suction strength and growth rate, and performs the transport action. Simultaneously, tactile data is continuously monitored during transport to promptly detect and address any anomalies such as slippage that may occur.

[0055] Optionally, changes in tactile data are continuously monitored during the handling process, further including real-time prediction and compensation for slippage. Specifically, this includes the following steps: Step S431: When a continuous moving trajectory is detected at the center of contact force, predict the sliding direction and sliding speed of the object; In this embodiment, tactile data is analyzed in real time during the handling process. The change in the position of the contact force center reflects the relative motion trend between the suction head and the object. When the contact force center shows a continuous unidirectional movement over time, it indicates that the object is slipping relative to the vacuum suction head. By analyzing the direction and rate of the movement trajectory, the direction and speed of slippage can be predicted, providing a basis for subsequent compensation.

[0056] Step S432: Based on the sliding direction and sliding speed, generate motion compensation commands to adjust the motion speed and / or motion direction of the vacuum suction head.

[0057] In this embodiment, an active compensation command is generated based on the predicted slippage trend. The control system controls the robotic arm or actuator to superimpose a compensating motion with the same direction and speed as the slippage on the original motion, so that the vacuum suction head follows the object, thereby counteracting the relative slippage and keeping the object relatively stable.

[0058] By monitoring the movement trajectory of the center of contact force in real time during the handling process, predicting slippage trends, and actively compensating, the problem of the inability to correct object slippage in real time during handling is solved. Even if the object being grasped is a flexible object, it is more prone to slippage during handling due to its smooth surface and easy deformation. This implementation method uses tactile data to perceive slippage trends in real time and actively adjusts the suction head movement, which can suppress slippage in its early stages, preventing the object from falling off midway and achieving dynamic stability control throughout the grasping process.

[0059] This implementation generates differentiated control commands for different grasping states. Even if the grasped object is a flexible object, due to its characteristics such as easy collapse, easy leakage, and possible surface wrinkles, a uniform response strategy is often inefficient or ineffective. This implementation takes measures such as slowing down, pausing, and adjusting to prevent damage to the object in case of collapse / wrinkles, and adopts strategies such as retrying and relocation to better seal leaks. In normal state, parameters are maintained and handling monitoring is performed, enabling the control system to autonomously select the optimal response method according to the actual working conditions, thereby improving the intelligence and adaptability of the grasping process.

[0060] As another optional implementation, step S4, adjusting the adsorption parameters and / or motion parameters of the vacuum nozzle, further includes correcting the contact position and closed-loop control of the contact force. Specifically, it includes the following steps: Step S401: In the initial stage of the contact action, based on the moment when the contact force signal is first detected, correct the actual contact position coordinates of the vacuum nozzle; In this embodiment, at the initial stage of the contact action, the moment the tactile data first exceeds the contact threshold, the actual spatial coordinates of the current vacuum head are recorded and used as the true contact point. These coordinates are then used to update the preset contact position in the robot system. Subsequently, the robot uses these corrected coordinates as a reference for subsequent motion control.

[0061] Step S402: During the adsorption process, based on the trend of contact force change fed back by tactile data, the output power of the vacuum generator in the vacuum suction head is adaptively adjusted to keep the contact force within the preset safety threshold range.

[0062] In this embodiment, during the adsorption process, the magnitude of the contact force in the tactile data is monitored in real time. When the contact force tends to exceed the safety threshold, the output power of the vacuum generator is dynamically adjusted to change the suction force, ensuring that the contact force is always maintained between preset upper and lower limits. The upper limit is used to prevent excessive suction force from damaging the object, while the lower limit is used to ensure sufficient adsorption force to prevent it from falling off.

[0063] This implementation corrects the position coordinates through the initial contact signal, solving the problem of inconsistency between the robot's positioning error and the actual position of the object. By adjusting the suction force according to the trend of contact force changes, it achieves compliant grasping of flexible objects. Even if the grasped object is flexible, due to its characteristics such as easy fluctuation in position and easy deformation under force, this implementation simultaneously achieves position correction and force control closed loop through tactile feedback, keeping the contact force within a safe threshold and effectively avoiding object damage caused by excessive suction force or grasping failure caused by insufficient suction force.

[0064] Optionally, the method further includes a self-learning optimization step, which records historical crawling data and optimizes and updates the crawling parameters to enable the system to continuously evolve. Specifically, it includes the following steps: Step S5: Record the vacuum pressure data and tactile data during each grasping process, as well as the corresponding grasping results; In this embodiment, after each grasping operation is completed, the vacuum pressure data sequence and tactile data sequence recorded during the grasping process, including changes in contact force and pressure distribution, as well as the final grasping result, are associated and stored in a local database or cloud server. This data constitutes the basic sample set for subsequent parameter optimization.

[0065] Step S6: Based on the grasping results, optimize and update the grasping parameters for objects of the same type. The grasping parameters include contact speed, suction growth rate, and / or lifting speed.

[0066] In this embodiment, historical data is analyzed periodically or in real time. For the same type of object, the parameter differences between successful and unsuccessful grasps are compared to identify the parameter value range positively correlated with the success rate. For example, if the analysis finds that for a certain type of sponge object, the success rate is highest when the suction growth rate is between 30% and 50% / s, while the risk of collapse increases significantly when it exceeds 80% / s, then the default suction growth rate for this type of object is automatically updated to 40% / s.

[0067] This implementation method records sensor data and grasping results during each grasping process and optimizes and updates grasping parameters for similar objects, enabling the system to have self-learning capabilities. Even when grasping flexible objects, due to their diverse types and characteristics, fixed parameters are difficult to adapt to all working conditions. However, this implementation method continuously accumulates grasping experience and automatically optimizes parameters, allowing the system to continuously evolve with the increase in the number of grasping attempts. The grasping parameters become more accurate with each use, and the grasping success rate continues to improve.

[0068] In this embodiment, by deeply integrating tactile sensing with vacuum pressure, the system achieves state perception, anomaly recognition, differentiated control, and self-learning optimization throughout the entire process of grasping flexible objects, significantly improving the success rate, safety, and intelligence level of grasping.

[0069] Based on the same inventive concept, the present invention also provides an apparatus corresponding to the method in Embodiment 1, as shown in Embodiment 2.

[0070] Example 2 Based on Embodiment 1, another embodiment of the present invention is proposed, with reference to... Figure 3 A vacuum suction head gripping control device, the device comprising: a vacuum suction head 1, a tactile sensor 2, a vacuum pressure sensor (not shown in the figure) and a controller (not shown in the figure).

[0071] Vacuum suction head 1, located at the front end of the end effector, is used to apply negative pressure to the object to generate suction force; The vacuum suction head 1 has an adsorption cavity inside and is connected to a vacuum generating device through a vacuum tube 3.

[0072] Optionally, the vacuum nozzle 1 is the component that directly contacts the object and performs adsorption. The vacuum nozzle 1 has a deformable structure, including but not limited to a silicone suction cup, a flexible diaphragm nozzle, or a multi-chamber nozzle. The deformable structure can adapt to the undulations of the object's surface within a certain range, improving the sealing effect. Silicone suction cups are suitable for objects with flat or slightly curved surfaces, flexible diaphragm nozzles are suitable for thin-walled flexible objects, and multi-chamber nozzles are suitable for porous or breathable objects.

[0073] Optionally, refer to Figure 4The vacuum suction head 1 has one or more rows of vacuum suction holes 5 on its adsorption working surface. These vacuum suction holes 5 are connected to the vacuum tube 3 and are used to evenly distribute negative pressure during adsorption, preventing excessive local suction that could cause excessive deformation of the flexible object. The pore size, number, and distribution density of the vacuum suction holes 5 can be designed according to the characteristics of the target object. For example, for thin-walled flexible objects, using densely packed small-diameter suction holes can reduce local pressure and prevent suction-induced damage.

[0074] The tactile sensor 2 is integrated on the adsorption working surface of the vacuum head or on the peripheral structure adjacent to the adsorption working surface, and is used to collect tactile signals at the interface between the object and the vacuum head in real time. Optionally, the tactile sensor 2 is the core sensing component of this device. The tactile sensor 2 is one of a flexible pressure sensing array, a capacitive tactile sensing membrane, or a distributed pressure sensor. The flexible pressure sensing array 4 consists of multiple independent sensing units arranged in an array, which can provide pressure distribution information with high spatial resolution; the capacitive tactile sensing membrane is thin and flexible, and is suitable for being attached to curved surfaces; the distributed pressure sensor is composed of multiple discrete sensors, which can be flexibly arranged according to the shape of the suction head.

[0075] Optionally, the sensing area of ​​the tactile sensor 2 covers both the central and edge areas of the working surface of the vacuum suction head 1. The sensing of the central area is used to determine whether the contact is sufficient and whether there is any collapse; the sensing of the edge area is used to determine whether the seal is good and whether there is any leakage.

[0076] Optionally, the multiple sensing units 4 of the tactile sensor 2 are distributed around or in the gaps of the vacuum suction holes 5, so that the contact state near each vacuum suction hole 5 can be independently sensed. When the vacuum suction holes 5 in a certain area are blocked due to the local collapse of the object, the tactile sensor 2 in that area can detect abnormal pressure changes, thereby providing the controller with accurate positioning information.

[0077] Optionally, when the vacuum suction head 1 is a deformable structure, the tactile sensor 2 is attached to the inner or outer surface of the deformable structure and deforms synchronously with the deformation of the vacuum suction head 1. For example, when the vacuum suction head 1 is a silicone suction cup, the tactile sensor 2 can be a flexible pressure sensor array attached to the inner surface of the suction cup. When the suction cup deforms due to adsorption, the sensor bends accordingly, always maintaining consistency with the surface of the suction cup, thereby accurately reflecting the contact state between the suction cup and the object.

[0078] It should be noted that the specific implementation of tactile sensors is not limited to the types listed above. Any other type of tactile sensor that can sense the magnitude of contact force and / or pressure distribution, such as piezoresistive, piezoelectric, and optical tactile sensors, can be selected and integrated according to actual application requirements, and all fall within the protection scope of this invention.

[0079] A vacuum pressure sensor is connected to the air path of the vacuum nozzle 1 and is used to monitor the vacuum pressure data in the air path in real time. Optionally, a vacuum pressure sensor is installed on the gas path between the vacuum tip 1 and the vacuum generator, or integrated into the gas path interface of the vacuum tip 1 body, to measure the negative pressure value inside the tip in real time. The vacuum pressure data reflects the degree of sealing between the tip and the object and the magnitude of the adsorption force.

[0080] The controller is electrically connected to both the tactile sensor and the vacuum pressure sensor. Alternatively, the controller is typically implemented using an embedded processor, a programmable logic controller, or an industrial control computer. The controller integrates a data acquisition module, a signal processing module, a fusion analysis module, and a control output module, enabling it to simultaneously acquire data from tactile sensors and vacuum pressure sensors, and perform real-time processing and analysis.

[0081] The controller is configured to perform the following operations: Step 1: Receive the tactile signal and the vacuum pressure data; The controller synchronously receives tactile signals from the tactile sensor and vacuum pressure data from the vacuum pressure sensor at a fixed sampling frequency, and performs preprocessing such as filtering and noise reduction on the raw data to ensure the accuracy and real-time performance of the data.

[0082] Step 2: Determine the current grasping state based on the fusion result of the tactile signal and the vacuum pressure data; The controller fuses and analyzes the pre-processed tactile data with the vacuum pressure data, and determines the current grasping state according to preset judgment rules. The grasping state includes at least the following categories: poor sealing or leakage, partial collapse or surface wrinkles, and complete contact, as well as deformation parameters such as the degree of collapse and the uniformity of adhesion. For specific fusion analysis methods, please refer to step S3 of Embodiment 1 and its various optional implementations.

[0083] Step 3: Generate control commands based on the current grasping state to adjust the adsorption parameters and / or motion parameters of the vacuum head.

[0084] The controller generates corresponding control commands based on the determined current grasping state. These control commands can be sent to the vacuum generator to adjust adsorption parameters, such as suction strength, suction increase rate, and pumping rate, or to the end effector or robotic arm to adjust motion parameters, such as contact speed, lifting speed, direction of movement, and suction head posture. For specific control strategies, refer to step S4 of Embodiment 1 and its various optional implementations.

[0085] As an optional implementation, the device further includes a vacuum generator, and the controller is connected to the vacuum generator.

[0086] A vacuum generator is an actuator that generates negative pressure, and can take the form of a venturi tube, vacuum pump, etc. The input end of the vacuum generator is connected to a compressed air source or power supply, and the output end is connected to the vacuum suction head through an air circuit. Its operating state is regulated by control commands output by the controller.

[0087] Optionally, the controller is further configured to: When the current gripping state is detected to include partial collapse or surface wrinkles, an output command is sent to control the vacuum generator to reduce the pumping rate or stop pumping. When the current gripping state is detected to include poor sealing or leakage, an instruction is output to control the vacuum head to re-perform the contact action or adjust the contact position.

[0088] In this embodiment, the controller achieves direct control of the vacuum generator and motion control of the end effector. For collapse or wrinkle conditions, the controller slows down the adsorption process by reducing the pumping rate or stops pumping to allow the object to spring back and reset, preventing damage. For leakage conditions, the controller controls the end effector to make the vacuum head detach and re-engage, or to move it on the object surface to find a better contact position.

[0089] In this embodiment, by integrating a tactile sensor into the vacuum suction head and forming a closed-loop control system together with a vacuum pressure sensor and a controller, a hardware implementation foundation is provided for the method described in Embodiment 1. Even if the object being grasped is a flexible object, due to its uncertain shape, easy deformation, and possible surface wrinkles, related vacuum suction head devices struggle to perceive the actual contact state. However, this embodiment, through the deep integration of the tactile sensor and the vacuum suction head, enables the device to perceive contact force, pressure distribution, and degree of deformation. It can acquire multi-dimensional state information during the grasping process in real time, and through the controller's fusion analysis and differentiated control, achieve precise, stable, and non-destructive grasping of flexible objects.

[0090] Furthermore, by setting one or more rows of vacuum suction holes on the suction working surface of the vacuum nozzle, a uniform distribution of negative pressure is achieved, effectively avoiding damage to objects caused by excessive local suction. Simultaneously, the sensing units of the tactile sensor are distributed around the vacuum suction holes, allowing the contact state near each hole to be independently sensed, providing hardware support for the precise location of anomalies such as local collapse and blockage. The optimized design, including the tactile sensor covering the center and edge areas of the nozzle and conforming to deformable structural surfaces, further enhances the comprehensiveness and realism of the sensing. The vacuum generator is incorporated into closed-loop control, enabling suction adjustment to quickly respond to tactile feedback, ensuring the effective implementation of the control method from a hardware perspective.

[0091] Since the apparatus described in Embodiment 2 of the present invention is an apparatus used to implement the method of Embodiment 1 of the present invention, those skilled in the art can understand the specific structure and variations of the apparatus based on the method described in Embodiment 1 of the present invention, and therefore will not be described again here. All apparatuses used in the method of Embodiment 1 of the present invention fall within the scope of protection of the present invention.

[0092] Example 3 In this embodiment of the invention, a computer storage medium is provided, which stores a computer program. When the computer program is executed by a processor, it implements the vacuum suction head gripping control method described in Embodiment 1 and its various optional embodiments.

[0093] Reference Figure 5 , Figure 5 This is a schematic diagram of the terminal structure of the hardware operating environment involved in an embodiment of the present invention.

[0094] like Figure 5 As shown, the control terminal may include: a processor 1001, such as a CPU, a network interface 1003, a memory 1004, and a communication bus 1002. The communication bus 1002 is used to enable communication between these components. The network interface 1003 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1004 may be high-speed RAM or stable non-volatile memory, such as disk storage. Alternatively, the memory 1004 may be a storage device independent of the aforementioned processor 1001.

[0095] Those skilled in the art will understand that Figure 5 The terminal structure shown does not constitute a limitation on the terminal and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0096] like Figure 5As shown, the memory 1004, which serves as a computer storage medium, may include an operating system, a network communication module, and a vacuum suction head gripping control program.

[0097] exist Figure 5 In the hardware structure of the vacuum suction head gripping control device shown, the processor 1001 can call the vacuum suction head gripping control program stored in the memory 1004 and perform the following operations: Real-time acquisition of vacuum pressure data from the vacuum suction head, and simultaneous acquisition of tactile data; When the change in the tactile data exceeds a preset contact threshold, it is determined that the vacuum nozzle is in contact with the object; In response to the contact, the vacuum pressure data and the tactile data are fused and analyzed to determine the current gripping state of the vacuum suction head; Based on the current grasping state, a control command is generated, which is used to adjust the adsorption parameters of the vacuum head and / or the motion parameters of the vacuum head.

[0098] Optionally, the processor 1001 may call the vacuum suction head gripping control program stored in the memory 1004 and also perform the following operations: When the vacuum pressure data does not reach the preset negative pressure threshold and the tactile data shows that contact force has been generated, the current grasping state is determined to be poor sealing or leakage. When the vacuum pressure data reaches the preset negative pressure threshold, and the tactile data shows that the pressure distribution is uneven or there is a sudden change in local pressure, the current grasping state is determined to be a local collapse or a wrinkled surface. When the vacuum pressure data is stable within a preset range and the tactile data shows a uniform pressure distribution, the current grasping state is determined to be complete contact.

[0099] Optionally, the processor 1001 may call the vacuum suction head gripping control program stored in the memory 1004 and also perform the following operations: When a local collapse or surface wrinkling is determined, a real-time pressure distribution map is constructed based on the pressure values ​​collected by multiple sensing units. Calculate the offset of the pressure center or the variance of the pressure values ​​in the pressure distribution map; Based on the offset or the variance, the degree of collapse or the uniformity of fit of the object is quantified, and the degree of collapse or the uniformity of fit is used as the deformation parameter of the current gripping state.

[0100] Optionally, the processor 1001 may call the vacuum suction head gripping control program stored in the memory 1004 and also perform the following operations: When the current gripping state is a partial collapse or the surface has wrinkles, reduce the rate of increase of vacuum suction, pause adsorption, or adjust the suction head posture; When the current gripping state is poor sealing or leakage, control the vacuum suction head to re-perform the contact action or adjust the contact position; Once the current grasping state is complete and the grasping is successful, the current adsorption parameters are maintained and the object is transported, while the changes in tactile data are continuously monitored during the transport process.

[0101] Optionally, the processor 1001 may call the vacuum suction head gripping control program stored in the memory 1004 and also perform the following operations: When a continuous moving trajectory is detected at the center of contact force, the sliding direction and sliding speed of the object are predicted. Based on the sliding direction and sliding speed, a motion compensation command is generated to adjust the movement speed and / or movement direction of the vacuum suction head.

[0102] Optionally, the processor 1001 may call the vacuum suction head gripping control program stored in the memory 1004 and also perform the following operations: In the initial stage of the contact action, the actual contact position coordinates of the vacuum nozzle are corrected based on the moment when the contact force signal is first detected. During the adsorption process, the output power of the vacuum generator in the vacuum head is adaptively adjusted based on the trend of contact force changes fed back by tactile data, so that the contact force is maintained within the preset safety threshold range.

[0103] Optionally, the processor 1001 may call the vacuum suction head gripping control program stored in the memory 1004 and also perform the following operations: Record the vacuum pressure data and tactile data during each grasping process, as well as the corresponding grasping results; Based on the grasping results, the grasping parameters for similar objects are optimized and updated. The grasping parameters include contact speed, suction growth rate, and / or lifting speed.

[0104] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0105] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0106] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0107] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0108] It should be noted that any reference signs placed between parentheses in the claims should not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, third, etc., does not indicate any order. These words can be interpreted as names.

[0109] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0110] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A vacuum suction head gripping control method, characterized in that, The method includes: Real-time acquisition of vacuum pressure data from the vacuum suction head, and simultaneous acquisition of tactile data; When the change in the tactile data exceeds a preset contact threshold, it is determined that the vacuum nozzle is in contact with the object; In response to the contact, the vacuum pressure data and the tactile data are fused and analyzed to determine the current gripping state of the vacuum suction head; Based on the current grasping state, a control command is generated, which is used to adjust the adsorption parameters of the vacuum head and / or the motion parameters of the vacuum head.

2. The method as described in claim 1, characterized in that, The step of fusing and analyzing the vacuum pressure data and the tactile data to determine the current gripping state of the vacuum suction head includes: When the vacuum pressure data does not reach the preset negative pressure threshold and the tactile data shows that contact force has been generated, the current grasping state is determined to be poor sealing or leakage. When the vacuum pressure data reaches the preset negative pressure threshold, and the tactile data shows that the pressure distribution is uneven or there is a sudden change in local pressure, the current grasping state is determined to be a local collapse or a wrinkled surface. When the vacuum pressure data is stable within a preset range and the tactile data shows a uniform pressure distribution, the current grasping state is determined to be complete contact.

3. The method as described in claim 2, characterized in that, The step of fusing and analyzing the vacuum pressure data and the tactile data to determine the current gripping state of the vacuum suction head further includes: When a local collapse or surface wrinkling is determined, a real-time pressure distribution map is constructed based on the pressure values ​​collected by multiple sensing units. Calculate the offset of the pressure center or the variance of the pressure values ​​in the pressure distribution map; Based on the offset or the variance, the degree of collapse or the uniformity of fit of the object is quantified, and the degree of collapse or the uniformity of fit is used as the deformation parameter of the current gripping state.

4. The method as described in claim 1, characterized in that, The step of generating control commands based on the current grasping state includes: When the current gripping state is a partial collapse or the surface has wrinkles, reduce the rate of increase of vacuum suction, pause adsorption, or adjust the suction head posture; When the current gripping state is poor sealing or leakage, control the vacuum suction head to re-perform the contact action or adjust the contact position; Once the current grasping state is complete and the grasping is successful, the current adsorption parameters are maintained and the object is transported, while the changes in tactile data are continuously monitored during the transport process.

5. The method as described in claim 4, characterized in that, The step of continuously monitoring changes in tactile data during the handling process includes: When a continuous moving trajectory is detected at the center of contact force, the sliding direction and sliding speed of the object are predicted. Based on the sliding direction and sliding speed, a motion compensation command is generated to adjust the movement speed and / or movement direction of the vacuum suction head.

6. The method as described in claim 4, characterized in that, The steps of adjusting the adsorption parameters and / or motion parameters of the vacuum nozzle include: In the initial stage of the contact action, the actual contact position coordinates of the vacuum nozzle are corrected based on the moment when the contact force signal is first detected. During the adsorption process, the output power of the vacuum generator in the vacuum head is adaptively adjusted based on the trend of contact force changes fed back by tactile data, so that the contact force is maintained within the preset safety threshold range.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: Record the vacuum pressure data and tactile data during each grasping process, as well as the corresponding grasping results; Based on the grasping results, the grasping parameters for similar objects are optimized and updated. The grasping parameters include contact speed, suction growth rate, and / or lifting speed.

8. A vacuum suction head gripping device, characterized in that, The device includes: A vacuum suction head, located at the front end of the end effector, is used to apply negative pressure to an object to generate suction force; A tactile sensor is integrated into the adsorption working surface of the vacuum head or the peripheral structure adjacent to the adsorption working surface, and is used to collect tactile signals at the interface between the object and the vacuum head in real time. A vacuum pressure sensor is connected to the air path of the vacuum nozzle and is used to monitor the vacuum pressure data in the air path in real time. The controller is electrically connected to both the tactile sensor and the vacuum pressure sensor. The controller is configured to perform the following operations: Receive the tactile signal and the vacuum pressure data; Based on the fusion result of the tactile signal and the vacuum pressure data, the current grasping state is determined; Based on the current grasping state, control commands are generated to adjust the adsorption parameters and / or motion parameters of the vacuum head.

9. The apparatus as claimed in claim 8, characterized in that, The tactile sensing module is one of a flexible pressure sensing array, a capacitive tactile sensing membrane, or a distributed pressure sensor. The sensing area of ​​the tactile sensor covers the central and edge areas of the working surface of the vacuum suction head; The vacuum suction head has a deformable structure and includes a silicone suction cup, a flexible diaphragm suction head, or a multi-chamber suction head. The tactile sensor is attached to the inner or outer surface of the deformable structure and deforms synchronously with the deformation of the vacuum suction head.

10. The apparatus as claimed in claim 8, characterized in that, The device also includes a vacuum generator, and the controller is connected to the vacuum generator; The controller is also configured to: When the current gripping state is detected to include partial collapse or surface wrinkles, an output command is sent to control the vacuum generator to reduce the pumping rate or stop pumping. When the current gripping state is detected to include poor sealing or leakage, an instruction is output to control the vacuum head to re-perform the contact action or adjust the contact position.

Images