Industrial robot automatic grabbing system based on industrial vision

Abstract

The invention discloses an industrial robot automatic grabbing system based on industrial vision, and relates to the technical field of industrial robots, and the technical scheme is characterized by comprising the following steps: collecting real-time grabbing task data of an industrial site; wherein the industrial field real-time grabbing task data corresponds to a production station and a robot operation terminal; obtaining target grabbing task data of the industrial field real-time grabbing task data according to a pre-constructed visual feature dynamic matching library, and obtaining adaptive grabbing task data and a dynamic matching coefficient according to actual dynamic parameters of the target grabbing task data and a dynamic grabbing calibration standard; terminal dynamic operation parameters of the adaptive grabbing task data corresponding to the adaptive operation terminal are obtained, and the dynamic grabbing adaptation degree of the adaptive operation terminal is obtained according to the terminal dynamic operation parameters and the dynamic matching coefficient; and an industrial robot grabbing instruction is generated according to the comprehensive dynamic adaptation degree and the dynamic grabbing adaptation degree. The method has the effect that the safety and reliability of the production process are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of industrial robot technology, and more specifically, to an automated grasping system for industrial robots based on industrial vision. Background Technology

[0002] Early robotic grasping systems mostly employed mechanical positioning or simple visual recognition technology, capable of recognizing only static or low-dynamic targets, and exhibiting extremely low tolerance for changes in ambient lighting and object posture. In mixed-flow production lines for automotive parts or 3C electronics assembly workshops, when the shape, material, or speed of the object to be grasped changes, these systems often fail to adapt quickly, leading to mis-grabbing and positioning errors, resulting in production interruptions or product quality defects. These systems lack assessment of the dynamic adaptability between production workstations and robotic terminals, and cannot adjust operational strategies based on real-time task load and equipment status, often resulting in both idle workstation resources and terminal overload, hindering overall production efficiency.

[0003] With the development of industrial vision technology, some improved systems have introduced visual feature matching mechanisms, but most still use static feature libraries, which cannot cope with the dynamic changes in the contours and textures of objects during movement. At the same time, the adaptability evaluation of existing systems often focuses on a single dimension, only paying attention to the matching degree between tasks and equipment, or only evaluating the operating status of the terminal, lacking a comprehensive consideration of the workstation's carrying capacity and the terminal's collaborative efficiency. This results in the generation of grasping instructions lacking comprehensive data support, making it difficult to meet the dynamic control requirements under complex working conditions. Summary of the Invention

[0004] In view of the shortcomings of existing technologies, the purpose of this invention is to provide an automated grasping system for industrial robots based on industrial vision.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An industrial robot automated grasping system based on industrial vision includes: Data Acquisition Module: Acquires real-time data from industrial sites; wherein the real-time data from industrial sites corresponds to production workstations and robot operation terminals; The first processing module obtains the target grasping task data for real-time grasping tasks in the industrial field based on the pre-built visual feature dynamic matching library, and obtains the adapted grasping task data and dynamic fitting coefficient based on the actual dynamic parameters of the target grasping task data and the dynamic grasping calibration standard. Analysis module: Processes and analyzes the data from the adapted capture task, dynamic fit coefficient, production station and robot operation terminal to obtain the comprehensive dynamic fit degree; The second processing module obtains the dynamic operating parameters of the terminal corresponding to the adapted capture task data of the adapted operation terminal, and obtains the dynamic capture adaptation degree of the adapted operation terminal based on the dynamic operating parameters and dynamic fitting coefficient. Instruction generation module: Generates industrial robot grasping instructions based on comprehensive dynamic adaptability and dynamic grasping adaptability.

[0006] Preferably, the dynamic grasping calibration standard includes a dynamic speed threshold, a dynamic accuracy threshold, and a dynamic load adjustment threshold corresponding to the grasping task condition type; The visual feature dynamic matching library contains dynamic contour features and dynamic texture feature data corresponding to the capture task condition types.

[0007] Preferably, the target grasping task data for real-time grasping tasks in industrial settings is obtained based on a pre-built visual feature dynamic matching library, specifically including the following steps: Dynamic feature matching is performed on the real-time grasping task data in the industrial field based on the dynamic contour feature and dynamic texture feature data of the visual feature dynamic matching library to obtain the target dynamic features. Based on the dynamic characteristics of the target, dynamic feature association data is extracted from the real-time capture task data in the industrial field to obtain the target task data to be captured corresponding to the dynamic characteristics of the target.

[0008] Preferably, the process of obtaining the adapted grasping task data and dynamic fitting coefficient based on the actual dynamic parameters of the target grasping task data and the dynamic grasping calibration standard specifically includes the following steps: The actual dynamic parameters of the target task data are obtained based on the task condition type corresponding to the target task data to be captured. The actual dynamic parameters include the actual dynamic speed value, actual dynamic accuracy value, and actual dynamic load adjustment value corresponding to the target task data to be captured. If all values ​​of the actual dynamic parameters in the real-time grasping task data of the industrial site are within the threshold range corresponding to the dynamic grasping calibration standard, then the real-time grasping task data of the industrial site is determined to be the adapted grasping task data, and the dynamic deviation rate between the actual dynamic parameters and the corresponding threshold is determined. The dynamic fit coefficient of the adaptive grasping task data is obtained by comparing the dynamic deviation rate with the threshold corresponding to the dynamic grasping calibration standard.

[0009] Preferably, the comprehensive dynamic adaptation degree is obtained by processing and analyzing the adapted grasping task data, dynamic fitting coefficient, production station and robot operation terminal, specifically including the following steps: The first workstation adaptation coefficient is obtained based on the task dynamic distribution parameters and dynamic matching coefficient of the adapted capture task data. Mark the robot terminal corresponding to the data to be captured as the adapted terminal. The second workstation adaptation coefficient of the production workstation is obtained based on the dynamic operation parameters of the adapted operation terminal in the production workstation. The overall dynamic adaptation degree is obtained based on the adaptation coefficient of the first workstation and the adaptation coefficient of the second workstation.

[0010] Preferably, the first workstation adaptation coefficient is obtained based on the task dynamic distribution parameters and dynamic matching coefficient of the adapted capture task data, specifically including the following steps: Set the dynamic weight of the working condition of the adapted crawling task data according to the working condition type of the crawling task data, and obtain the dynamic adaptation coefficient of the adapted crawling task data according to the dynamic fitting coefficient and the dynamic weight of the working condition. The task dynamic distribution parameters include the temporal dynamic distribution density and spatial dynamic coverage of the data to be captured in the production workstation. The first dynamic adaptation status value corresponding to the production station is obtained based on the time dynamic distribution density, spatial dynamic coverage range and dynamic adaptation coefficient. The total number of tasks for capturing real-time data in the industrial field at the production workstation is obtained. Based on the total number of tasks and the number of data entries for the adapted capture task at the production workstation, the second dynamic adaptation status value of the production workstation is obtained. The first workstation adaptation coefficient is obtained based on the first dynamic adaptation status value and the second dynamic adaptation status value.

[0011] Preferably, the second workstation adaptation coefficient of the production workstation is obtained based on the dynamic operation parameters of the adapted operation terminal in the production workstation, specifically including the following steps: Determine the collaborative correlation characteristics between the dynamic operation parameters of the adapted operation terminals in the production workstation, and filter out the target parameter items that affect the operation flow of the production workstation based on the collaborative correlation characteristics; Based on the production workstation workflow sequence, determine the work connection nodes of each adapted work terminal, and establish the correspondence between target parameter items and connection nodes; Based on the correspondence, the adaptability and coordination of the dynamic parameters of the terminal target at the connection node are determined, and the degree of cooperative adaptation is obtained. Based on the overall workflow efficiency requirements of each workstation, the coordination and adaptability of each connection node are prioritized to obtain critical connection nodes and normal connection nodes. The coordination deviation of the terminal target dynamic parameters is corrected based on the coordination and adaptation degree of key connection nodes and the adaptation status of normal connection nodes to obtain the terminal connection adaptation value. The adaptability of the workstation operation flow is evaluated based on the terminal connection adaptation value and the terminal operation connection frequency, and the second workstation adaptation coefficient of the production workstation is obtained.

[0012] Preferably, the terminal dynamic operating parameters include the dynamic operating stability, dynamic energy consumption fluctuation value, and dynamic fault warning response speed of the adapted operation terminal corresponding to the adapted capture task data.

[0013] Preferably, the dynamic capture adaptation degree of the adapted operation terminal is obtained based on the terminal's dynamic operating parameters and dynamic matching coefficient, specifically including the following steps: The dynamic stability adaptation coefficient is obtained based on the dynamic operational stability. The dynamic energy consumption adaptation coefficient is obtained by the ratio of the dynamic energy consumption fluctuation value to the preset dynamic energy consumption threshold, and the dynamic early warning adaptation coefficient is obtained by the dynamic fault early warning response speed. The dynamic stability adaptation coefficient, dynamic energy consumption adaptation coefficient, and dynamic early warning adaptation coefficient are weighted and summed to obtain the terminal dynamic operation adaptation coefficient; The dynamic capture adaptation degree of the adapted operation terminal is obtained by using the terminal dynamic operation adaptation coefficient and the dynamic matching coefficient of the adapted capture task data.

[0014] Preferably, the industrial robot grasping instructions are generated based on the comprehensive dynamic adaptability and dynamic grasping adaptability, specifically including the following steps: Prioritize the overall dynamic adaptation and dynamic crawling adaptation to form an adaptation deviation value; The weight ratio of overall dynamic fit and dynamic capture fit is adjusted based on the fit deviation value. The final adaptation evaluation value is obtained based on the comprehensive dynamic adaptation degree, dynamic grasping adaptation degree, and weight ratio. The basic parameters of the grasping action are determined based on the final adaptation evaluation value and the terminal dynamic operation parameters of the adapted operation terminal. The basic parameters include grasping force, motion trajectory, and posture adjustment range. Generate a capture instruction that includes action sequence, execution timing and safety threshold based on the basic parameters and the workstation parameters of the production station.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention links real-time data from production workstations and robotic terminals, combining a visual feature dynamic matching library to quickly identify targets to be grasped under different working conditions. It uses dynamic grasping calibration standards to screen suitable tasks and quantify dynamic fit coefficients, effectively solving the problem of poor adaptability of traditional grasping systems to complex working conditions. By dynamically adjusting the fit standards based on real-time motion parameters, the false grasp rate is significantly reduced, improving the accuracy and efficiency of task matching. A comprehensive dynamic fit is generated by integrating task dynamic distribution parameters and terminal collaborative characteristics, assessing the task load density of the workstation and the operational stability of the robot. When the comprehensive dynamic fit is low, the task allocation strategy is automatically adjusted to avoid workstation overload or terminal idleness, enabling more efficient allocation of production resources. Through priority determination and dynamic weight adjustment, combined with the final fit evaluation value, grasping instructions containing action sequences, execution timing, and safety thresholds are generated. Through quantitative monitoring and dynamic optimization throughout the entire process, the safety and reliability of the production process are effectively improved. Attached Figure Description

[0016] Figure 1 A schematic diagram of a module for an industrial robot automatic grasping system based on industrial vision is provided for embodiments of the present invention; Figure 2 This invention provides a schematic diagram illustrating the steps involved in obtaining a dynamic fit coefficient in an industrial robot automatic grasping system based on industrial vision, as part of an embodiment of the present invention. Detailed Implementation

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0018] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0019] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0020] Reference Figures 1-2 As shown.

[0021] The embodiments further illustrate the automatic grasping system for industrial robots based on industrial vision proposed in this invention.

[0022] An industrial robot automated grasping system based on industrial vision includes: Data Acquisition Module: Acquires real-time data from industrial sites; the real-time data acquisition tasks correspond to production workstations and robot operation terminals. The first processing module obtains the target grasping task data for real-time grasping tasks in the industrial field based on the pre-built visual feature dynamic matching library, and obtains the adapted grasping task data and dynamic fitting coefficient based on the actual dynamic parameters of the target grasping task data and the dynamic grasping calibration standard. Analysis module: Processes and analyzes the data from the adapted capture task, dynamic fit coefficient, production station and robot operation terminal to obtain the comprehensive dynamic fit degree; The second processing module obtains the dynamic operating parameters of the terminal corresponding to the adapted capture task data of the adapted operation terminal, and obtains the dynamic capture adaptation degree of the adapted operation terminal based on the dynamic operating parameters and dynamic fitting coefficient. Instruction generation module: Generates industrial robot grasping instructions based on comprehensive dynamic adaptability and dynamic grasping adaptability.

[0023] The dynamic grasping calibration standard includes dynamic speed threshold, dynamic accuracy threshold, and dynamic load adjustment threshold corresponding to the grasping task working condition type; The visual feature dynamic matching library contains dynamic contour features and dynamic texture feature data corresponding to the working conditions of the grasping task.

[0024] The dynamic grasping calibration standard includes dynamic speed threshold, dynamic accuracy threshold, and dynamic load adjustment threshold. In actual industrial scenarios, different working conditions place significantly different demands on robot operation. For example, in the 3C electronics industry, when grasping precision components, the dynamic speed threshold is typically set between 0.1 and 0.3 meters per second to prevent excessive speed from causing component misalignment or damage. The dynamic accuracy threshold is controlled within ±0.02 millimeters to ensure precise insertion of components into their corresponding mounting holes. The dynamic load adjustment threshold is limited to between 0.5 and 2 kilograms to prevent overload from deforming the robot's end effector. In the case of grasping automotive chassis parts, due to the greater weight of the parts, the dynamic load adjustment threshold is relaxed to 5 to 15 kilograms, and the dynamic speed threshold is adjusted to 0.2 to 0.5 meters per second to balance operational efficiency and equipment safety.

[0025] The visual feature dynamic matching library stores dynamic contour and dynamic texture feature data corresponding to different grasping task conditions. Dynamic contour features represent the changes in the geometry of the object to be grasped over time. For example, a metal gear moving at high speed on a conveyor line exhibits periodic rotational changes in its contour; the visual feature dynamic matching library stores the contour coordinate sequence of this gear at different angles. Dynamic texture features focus on the textural details of the object's surface, such as the rough texture of a rubber seal versus the smooth, reflective texture of a metal bearing, and are converted into grayscale value change sequences stored in the visual feature dynamic matching library.

[0026] The target grasping task data for real-time grasping tasks in industrial settings is obtained based on a pre-built visual feature dynamic matching library, specifically including the following steps: Dynamic feature matching is performed on the real-time grasping task data in the industrial field based on the dynamic contour feature and dynamic texture feature data of the visual feature dynamic matching library to obtain the target dynamic features. Based on the dynamic characteristics of the target, dynamic feature association data is extracted from the real-time capture task data in the industrial field to obtain the target task data to be captured corresponding to the dynamic characteristics of the target.

[0027] Dynamic feature matching is performed on real-time captured task data. Taking the automotive parts assembly scenario as an example, when bolts, nuts, and washers appear mixed on the conveyor line, the dynamic contour features of each object in the real-time image are extracted. The dynamic contour features are then compared frame by frame with the bolts and nuts pre-stored in the visual feature dynamic matching library. The matching process involves calculating the similarity between the real-time contour and the standard contour in the visual feature dynamic matching library. Contour similarity = (number of overlapping pixels between the real-time contour and the standard contour ÷ total number of pixels in the real-time contour) × 100%. If the contour similarity of a certain region reaches 90% or more, the dynamic texture features of that region are extracted. The dynamic texture features are then matched with the texture data of the corresponding object in the visual feature dynamic matching library. Texture similarity = (reciprocal of the sum of the squares of the differences in grayscale values ​​between the real-time texture and the standard texture × 100). When both contour similarity and texture similarity reach a preset threshold, the target dynamic feature corresponding to that region is determined.

[0028] After completing dynamic feature matching, dynamic feature association data is extracted. Based on the target's dynamic features, complete information associated with it is filtered from the real-time grasping task data in the industrial field. For example, after identifying the target dynamic features of a bolt, the bolt's movement speed on the conveyor line, real-time position coordinates, rotation angle, and the status data of the corresponding robot working terminal are extracted, thus forming the target task data to be grasped.

[0029] The process of obtaining adapted capture task data and dynamic fit coefficients based on the actual dynamic parameters of the target capture task data and dynamic capture calibration standards includes the following steps: The actual dynamic parameters of the target task data are obtained based on the task condition type corresponding to the target task data to be captured. The actual dynamic parameters include the actual dynamic speed value, actual dynamic accuracy value, and actual dynamic load adjustment value corresponding to the target task data to be captured; If all values ​​of the actual dynamic parameters in the real-time grasping task data of the industrial site are within the threshold range corresponding to the dynamic grasping calibration standard, then the real-time grasping task data of the industrial site is determined to be the adapted grasping task data, and the dynamic deviation rate between the actual dynamic parameters and the corresponding threshold is determined. The dynamic fit coefficient of the adaptive grasping task data is obtained by comparing the dynamic deviation rate with the threshold corresponding to the dynamic grasping calibration standard.

[0030] The actual dynamic parameters of the target grasping task data are obtained based on the grasping task condition type corresponding to the target task data to be grasped. These actual dynamic parameters include the actual dynamic speed value, the actual dynamic accuracy value, and the actual dynamic load adjustment value. Taking the chip grasping condition in the 3C electronics industry as an example, when the chip to be grasped is identified, the chip's movement speed on the conveyor line is used as the actual dynamic speed value; the deviation between the chip's center coordinates and the visual recognition coordinates is used as the actual dynamic accuracy value; and the chip's weight is used as the actual dynamic load adjustment value.

[0031] These actual dynamic parameters are compared with the corresponding threshold ranges of the dynamic grasping calibration standard. The dynamic grasping calibration standard presets thresholds for different operating conditions. For example, the dynamic speed threshold range for chip grasping is 0.1 m to 0.3 m per second, the dynamic accuracy threshold range is ±0.02 mm, and the dynamic load adjustment threshold range is 0.5 kg to 1 kg. Only when the actual dynamic speed value, actual dynamic accuracy value, and actual dynamic load adjustment value are all within the corresponding threshold range is the real-time grasping task data in the industrial field determined to be suitable grasping task data. If the actual dynamic speed value of a chip grasping task is 0.2 m per second, the actual dynamic accuracy value is ±0.015 mm, and the actual dynamic load adjustment value is 0.8 kg, and if all three indicators are within the threshold range, then it is determined to be suitable grasping task data.

[0032] After determining that the data is suitable for the grasping task, the dynamic deviation rate between the actual dynamic parameters and the corresponding threshold is calculated. Dynamic deviation rate = |Actual value - Mean threshold value| ÷ Mean threshold value × 100%, where the mean threshold value is the average of the upper and lower limits of the corresponding threshold range. Taking the dynamic speed threshold for chip grasping as an example, the mean threshold value is (0.1 + 0.3) ÷ 2 = 0.2 meters per second. If the actual dynamic speed value is 0.22 meters per second, substituting it into the formula, the speed dynamic deviation rate is |0.22 - 0.2| ÷ 0.2 × 100% = 10%. Similarly, the accuracy dynamic deviation rate and load dynamic deviation rate are calculated.

[0033] The dynamic fit coefficient for the adaptive grasping task data is calculated based on the dynamic deviation rate and the threshold corresponding to the dynamic grasping calibration standard. The dynamic fit coefficient = 1 - ((speed deviation rate × 0.3 + accuracy deviation rate × 0.4 + load deviation rate × 0.3) ÷ 3). The weight allocation is set according to the sensitivity of the indicators to the operating conditions. If the speed deviation rate of a certain chip grasping task is 10%, the accuracy deviation rate is 5%, and the load deviation rate is 8%, substituting these values ​​into the formula yields a dynamic fit coefficient of 1 - ((10% × 0.3 + 5% × 0.4 + 8% × 0.3) ÷ 3) ≈ 0.96. The closer this coefficient is to 1, the higher the fit between the actual dynamic parameters and the dynamic grasping calibration standard, and the stronger the reliability of the task execution.

[0034] The comprehensive dynamic fit is obtained by processing and analyzing the data from the capture task, dynamic fit coefficient, production station, and robot operation terminal. This process includes the following steps: The first workstation adaptation coefficient is obtained based on the task dynamic distribution parameters and dynamic matching coefficient of the adapted capture task data. The specific steps include: Set the dynamic weight of the working condition of the adapted crawling task data according to the working condition type of the crawling task data, and obtain the dynamic adaptation coefficient of the adapted crawling task data according to the dynamic fitting coefficient and the dynamic weight of the working condition. Task dynamic distribution parameters include the temporal dynamic distribution density and spatial dynamic coverage of the data to be captured in the production workstation. The first dynamic adaptation status value corresponding to the production station is obtained based on the time dynamic distribution density, spatial dynamic coverage range and dynamic adaptation coefficient. The total number of tasks for capturing real-time data in the industrial field at the production workstation is obtained. Based on the total number of tasks and the number of data entries for the adapted capture task at the production workstation, the second dynamic adaptation status value of the production workstation is obtained. The first workstation adaptation coefficient is obtained based on the first dynamic adaptation status value and the second dynamic adaptation status value. Dynamic weights for different work conditions are set based on the work condition type corresponding to the data being adapted for the crawling task. Different work conditions have varying importance to the production workstation, and the weights are adjusted accordingly. For example, in the automotive engine assembly workstation, the weight for crawling the core component cylinder block is set to 0.7, while the weight for crawling the auxiliary component gasket is set to 0.3. The dynamic adaptation coefficient for a single piece of adapted data is obtained by combining the dynamic fit coefficient of the task: Dynamic Adaptation Coefficient = Dynamic Fit Coefficient × Dynamic Weight of Work Condition. Taking a cylinder block grabbing task as an example, if its dynamic fit coefficient is 0.95 and the working condition dynamic weight is 0.7, substituting into the formula, we get the dynamic fit coefficient as 0.95 × 0.7 = 0.665.

[0035] The task dynamic distribution parameters include the temporal dynamic distribution density and spatial dynamic coverage of the adaptation and capture task data in the production workstation. Temporal dynamic distribution density refers to the number of adaptation tasks that a workstation can handle per unit of time; for example, an engine assembly workstation processes 20 adaptation and capture tasks per hour. Spatial dynamic coverage refers to the breadth of the adaptation tasks' distribution within the workstation's operating space; for example, adaptation tasks covering 80% of the workstation's operating area. The first dynamic adaptation status value of the production workstation is calculated by combining the dynamic adaptation coefficient: First dynamic adaptation status value = (Temporal dynamic distribution density ÷ Maximum temporal distribution density of the workstation × 0.5) + (Spatial dynamic coverage ÷ Total operating space of the workstation × 0.3) + (Dynamic adaptation coefficient × 0.2) Assuming the maximum time distribution density of workstations is 30 per hour and the total working space of workstations is 100 square meters, substituting the example data, the first dynamic adaptation status value is (20÷30×0.5)+(80÷100×0.3)+(0.665×0.2)≈0.333+0.24+0.133=0.706.

[0036] The total number of real-time data capture tasks at each production workstation is calculated. For example, if a workstation receives 30 real-time capture tasks within a certain time period, and 24 of them are adapted capture tasks, the second dynamic adaptation status value is calculated as: (Number of adapted capture tasks ÷ Total number of tasks). Substituting the data, the second dynamic adaptation status value is 24 ÷ 30 = 0.8. The first and second dynamic adaptation status values ​​are then combined to obtain the first workstation adaptation coefficient: First workstation adaptation coefficient = (First dynamic adaptation status value × 0.6) + (Second dynamic adaptation status value × 0.4). Substituting the data, the first workstation adaptation coefficient is (0.706 × 0.6) + (0.8 × 0.4) ≈ 0.424 + 0.32 = 0.744. The closer the first workstation adaptation coefficient is to 1, the stronger the production workstation's ability to handle adapted capture tasks.

[0037] Mark the robot terminal corresponding to the data to be captured as the adapted terminal. The second workstation adaptation coefficient is obtained based on the dynamic operation parameters of the adapted operation terminal in the production workstation. The specific steps include: Determine the collaborative correlation characteristics between the dynamic operation parameters of the adapted operation terminals in the production workstation, and filter out the target parameter items that affect the operation flow of the production workstation based on the collaborative correlation characteristics; Based on the production workstation workflow sequence, determine the work connection nodes of each adapted work terminal, and establish the correspondence between target parameter items and connection nodes; Based on the correspondence, the adaptability and coordination of the dynamic parameters of the terminal target at the connection node are determined, and the degree of cooperative adaptation is obtained. Based on the overall workflow efficiency requirements of each workstation, the coordination and adaptability of each connection node are prioritized to obtain critical connection nodes and normal connection nodes. The coordination deviation of the terminal target dynamic parameters is corrected based on the coordination and adaptation degree of key connection nodes and the adaptation status of normal connection nodes to obtain the terminal connection adaptation value. The adaptability of the workstation operation flow is evaluated based on the terminal connection adaptation value and the terminal operation connection frequency, and the second workstation adaptation coefficient of the production workstation is obtained. In real-world production scenarios, a single production station typically deploys multiple compatible workstations to collaboratively complete tasks. For example, an automotive seat assembly station might include a robot (robot 1) responsible for gripping the seat cushion, a robot (robot 2) responsible for gripping the backrest, and a robot (robot 3) responsible for gripping the frame. First, the collaborative characteristics between the dynamic operational parameters of these compatible workstations are determined, and target parameters affecting the workflow of the production station are identified. Dynamic operational parameters include the terminal's response time, motion trajectory overlap, and load capacity. The interactions between these parameters are assessed; for example, the response time of robot 1 directly affects the start-up timing of robot 2, and the motion trajectory overlap determines whether interference occurs between the terminals. Ultimately, the response time and motion trajectory overlap are selected as the target parameters.

[0038] Based on the workflow sequence of the production station, the work connection nodes of each adapted work terminal are determined, and a correspondence between target parameters and connection nodes is established. Taking the automotive seat assembly station as an example, the workflow is as follows: Robot 1 picks up the seat cushion and places it into the tooling; Robot 2 picks up the backrest and assembles it with the seat cushion; Robot 3 picks up the frame and completes the assembly. Therefore, the connection nodes include the connection point between Robot 1 and Robot 2, and the connection point between Robot 2 and Robot 3. The work response time is mapped to the connection point between Robot 1 and Robot 2, and the motion trajectory overlap is mapped to the connection point between Robot 2 and Robot 3, thus forming a clear mapping relationship.

[0039] Based on the correspondence, the adaptability and coordination of the dynamic parameters of the terminal target at the connection node are determined, thereby obtaining the cooperative adaptation degree. For example, at the connection node between robot 1 and robot 2, the actual operation response time of robot 1 is compared with the preset connection standard time to obtain the time deviation rate. The time deviation rate = |actual response time - standard response time| ÷ standard response time × 100%. If the actual response time is 1.2 seconds and the standard response time is 1 second, then the time deviation rate is 20%, and the corresponding cooperative adaptation degree is 1 - time deviation rate = 0.8. At the connection node between robot 2 and robot 3, the adaptability of the motion trajectory overlap is calculated. If the actual overlap is 15% and the standard overlap is 10%, then the overlap deviation rate is 50%, and the corresponding cooperative adaptation degree is = 1 - overlap deviation rate = 0.5.

[0040] Based on the overall workflow efficiency requirements of each workstation, the coordination and adaptability of each connection node are prioritized to determine critical and normal connection nodes. For example, in the automotive seat assembly workstation, the connection node between robot 2 and robot 3 affects the final product assembly accuracy, and is therefore designated as a critical connection node with a priority weight of 0.7; the connection node between robot 1 and robot 2 is designated as a normal connection node with a priority weight of 0.3.

[0041] Based on priority-based classification, the coordination deviation of the dynamic parameters of the terminal target is corrected to obtain the terminal connection adaptation value. Terminal connection adaptation value = (coordination adaptation degree of key connection nodes × 0.7 + coordination adaptation degree of normal connection nodes × 0.3) × (1 - coordination deviation rate). Substituting the data, the coordination adaptation degree of key connection nodes is 0.5, the coordination adaptation degree of normal connection nodes is 0.8, and the coordination deviation rate is (20% + 50%) ÷ 2 = 35%. Therefore, the terminal connection adaptation value is approximately (0.5 × 0.7 + 0.8 × 0.3) × (1 - 0.35) ≈ 0.384.

[0042] The adaptability of workstation workflow is assessed by combining the terminal operation connection frequency, thus obtaining the second workstation adaptation coefficient. Second workstation adaptation coefficient = terminal connection adaptation value × (connection frequency ÷ average workstation connection frequency). If the connection frequency of this workstation is 15 times per hour and the average workstation connection frequency is 12 times per hour, substituting the data, we get the second workstation adaptation coefficient = 0.384 × (15 ÷ 12) = 0.48. The closer the second workstation adaptation coefficient is to 1, the stronger the collaborative operation capability of multiple terminals within the production workstation.

[0043] The overall dynamic adaptation degree is obtained based on the adaptation coefficient of the first workstation and the adaptation coefficient of the second workstation.

[0044] The first workstation adaptation coefficient measures the static carrying capacity of the production workstation for adapting grasping tasks, including the temporal distribution density of tasks, spatial coverage, and the adaptability of the tasks themselves; the second workstation adaptation coefficient focuses on the dynamic collaborative capability of multiple adapting operation terminals within the workstation, including the connection efficiency between terminals and parameter coordination.

[0045] Taking the automotive engine assembly station as an example, assume the adaptation coefficient of the first station is 0.744 and the adaptation coefficient of the second station is 0.48. In actual production, different working conditions have different weights for static load-bearing capacity and dynamic coordination requirements. For example, in stable batch production, the weight of static load-bearing capacity is slightly higher, while in flexible production with multiple varieties and small batches, the weight of dynamic coordination capacity is higher. Weights are usually preset according to the type of working condition. For example, in stable batch production, the weight of the adaptation coefficient of the first station is set to 0.6, and the weight of the adaptation coefficient of the second station is set to 0.4. The overall dynamic adaptation degree = adaptation coefficient of the first station × first weight + adaptation coefficient of the second station × second weight. Substituting the data, we get the overall dynamic adaptation degree = 0.744 × 0.6 + 0.48 × 0.4 ≈ 0.638. The closer the overall dynamic adaptation degree is to 1, the stronger the overall adaptation capability of the production station to the current grasping task.

[0046] The terminal's dynamic operating parameters include the dynamic operating stability, dynamic energy consumption fluctuation value, and dynamic fault warning response speed of the adapted task data capture terminal.

[0047] The dynamic capture adaptation degree of the adapted operation terminal is obtained based on the terminal's dynamic operating parameters and dynamic matching coefficient, specifically including the following steps: The dynamic stability adaptation coefficient is obtained based on the dynamic operational stability. The dynamic energy consumption adaptation coefficient is obtained by the ratio of the dynamic energy consumption fluctuation value to the preset dynamic energy consumption threshold, and the dynamic early warning adaptation coefficient is obtained by the dynamic fault early warning response speed. The dynamic stability adaptation coefficient, dynamic energy consumption adaptation coefficient, and dynamic early warning adaptation coefficient are weighted and summed to obtain the terminal dynamic operation adaptation coefficient; The dynamic capture adaptation degree of the adapted operation terminal is obtained by using the terminal dynamic operation adaptation coefficient and the dynamic matching coefficient of the adapted capture task data.

[0048] Focusing on the dynamic operating parameters of the terminal for adapted operations, these parameters include dynamic operating stability, dynamic energy consumption fluctuation, and dynamic fault warning response speed. Dynamic operating stability reflects the fluctuation range of the terminal's motion parameters during continuous operation, such as the positioning accuracy fluctuation of the robot's end effector; dynamic energy consumption fluctuation reflects the deviation range between the terminal's actual energy consumption and the preset energy consumption; and dynamic fault warning response speed reflects the time it takes for the terminal to issue a warning after detecting an anomaly.

[0049] Taking the No. 3 robot terminal in the automotive parts assembly station as an example, the dynamic stability adaptation coefficient is first calculated based on the dynamic operation stability. Assuming that the fluctuation range of the end-positioning accuracy of the terminal in continuous operation is 0.01 mm, and the maximum allowable fluctuation range is 0.05 mm, the dynamic stability adaptation coefficient = 1 - (parameter fluctuation range per unit time ÷ maximum allowable fluctuation range). Substituting the example data, we get the dynamic stability adaptation coefficient = 1 - (0.01 ÷ 0.05) = 0.8.

[0050] The dynamic energy consumption adaptation coefficient is calculated based on the dynamic energy consumption fluctuation value. If the dynamic energy consumption fluctuation value of the terminal is 10 watts and the preset dynamic energy consumption threshold is 50 watts, the dynamic energy consumption adaptation coefficient = 1 - (dynamic energy consumption fluctuation value ÷ preset dynamic energy consumption threshold). Substituting the data, the dynamic energy consumption adaptation coefficient = 1 - (10 ÷ 50) = 0.8. The dynamic warning adaptation coefficient is calculated based on the dynamic fault warning response speed. If the fault warning response speed of the terminal is 0.2 seconds and the preset optimal response speed is 0.1 seconds, the dynamic warning adaptation coefficient = 1 - (actual response speed ÷ preset optimal response speed - 1). Substituting the data, the dynamic warning adaptation coefficient = 1 - (0.2 ÷ 0.1 - 1) = 0.

[0051] Terminal dynamic operation adaptation coefficient = dynamic stability adaptation coefficient × 0.4 + dynamic energy consumption adaptation coefficient × 0.3 + dynamic early warning adaptation coefficient × 0.3. Substituting the data, we get the terminal dynamic operation adaptation coefficient = 0.8 × 0.4 + 0.8 × 0.3 + 0 × 0.3 = 0.56.

[0052] The dynamic capture adaptation degree of the adapted terminal is calculated by combining the dynamic matching coefficient of the capture task data. Assuming the dynamic matching coefficient of this task is 0.96, the dynamic capture adaptation degree = terminal dynamic running adaptation coefficient × dynamic matching coefficient. Substituting the data, we get the dynamic capture adaptation degree = 0.56 × 0.96 ≈ 0.538. The closer this value is to 1, the stronger the adapted terminal's execution capability for the current capture task.

[0053] The industrial robot's grasping instructions are generated based on the comprehensive dynamic adaptability and dynamic grasping adaptability, specifically including the following steps: Prioritize the overall dynamic adaptation and dynamic crawling adaptation to form an adaptation deviation value; The weight ratio of overall dynamic fit and dynamic capture fit is adjusted based on the fit deviation value. The final adaptation evaluation value is obtained based on the comprehensive dynamic adaptation degree, dynamic grasping adaptation degree, and weight ratio. The basic parameters of the grasping action are determined based on the final adaptation evaluation value and the terminal dynamic operation parameters of the adapted operation terminal. Among them, the basic parameters include grasping force, motion trajectory, and posture adjustment range. Generate a capture instruction that includes action sequence, execution timing and safety threshold based on the basic parameters and the workstation parameters of the production station.

[0054] First, prioritize the overall dynamic adaptability and dynamic crawling adaptability to determine the adaptability deviation value. For example, in a mass production scenario at an automotive engine assembly station, if the overall dynamic adaptability is 0.638, the dynamic crawling adaptability is 0.538, and the preset baseline value is 0.7, calculate the difference between each and the baseline value: Overall dynamic adaptability deviation = 0.7 - 0.638 = 0.062, Dynamic crawling adaptability deviation = 0.7 - 0.538 = 0.162. The adaptability deviation value = |Overall Dynamic Adaptability - Dynamic Crawling Adaptability|, substituting the data, we get the adaptability deviation value = |0.638 - 0.538| = 0.1. Adjust the weight ratio of overall dynamic adaptability and dynamic crawling adaptability based on the adaptability deviation value. If the deviation value is small, it indicates that the adaptability levels of the two are similar, and the weight can be set to 0.5 for each; if the deviation value is large, it indicates that one of the adaptability capabilities has a weakness, and the weight of the weaker item should be increased to optimize the overall adaptability. For example, when the adaptation deviation value is 0.1, the dynamic crawling adaptation is considered relatively weak. Therefore, the weight of the overall dynamic adaptation is set to 0.4, and the weight of the dynamic crawling adaptation is set to 0.6 to increase the proportion of terminal execution capability in the final evaluation. The final adaptation evaluation value = overall dynamic adaptation × overall weight + dynamic crawling adaptation × dynamic weight. Substituting the data, we get the final adaptation evaluation value = 0.638 × 0.4 + 0.538 × 0.6 ≈ 0.578.

[0055] Based on the final adaptation evaluation value and the dynamic operating parameters of the adapted terminal, the basic parameters of the grasping action are determined, including grasping force, motion trajectory, and attitude adjustment range. For example, when the final adaptation evaluation value is 0.578, combined with the terminal's dynamic operating stability of 0.8, the grasping force is set to 80% of the rated value to improve stability, the motion trajectory is adjusted to a smoother curve to reduce energy consumption fluctuations, and the attitude adjustment range is controlled within ±5 degrees to reduce the risk of failure.

[0056] The system generates complete grasping instructions, including action sequences, execution timing, and safety thresholds, based on the workstation parameters. For example, in an automotive engine assembly station, the instructions clearly state that the robot first grasps the cylinder block and moves it to the tooling position, then grasps the bolts for assembly. The execution timing is set with an interval of 1.5 seconds between each action, and the safety threshold limits the robot's movement speed to no more than 0.3 meters per second and the load to no more than 10 kilograms, to ensure both operational safety and efficiency.

[0057] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0058] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An industrial robot automatic grasping system based on industrial vision, characterized in that, include: Data Acquisition Module: Acquires real-time data from industrial sites; wherein the real-time data from industrial sites corresponds to production workstations and robot operation terminals; The first processing module obtains the target grasping task data for real-time grasping tasks in the industrial field based on the pre-built visual feature dynamic matching library, and obtains the adapted grasping task data and dynamic fitting coefficient based on the actual dynamic parameters of the target grasping task data and the dynamic grasping calibration standard. Analysis module: Processes and analyzes the data from the adapted capture task, dynamic fit coefficient, production station and robot operation terminal to obtain the comprehensive dynamic fit degree; The second processing module obtains the dynamic operating parameters of the terminal corresponding to the adapted capture task data of the adapted operation terminal, and obtains the dynamic capture adaptation degree of the adapted operation terminal based on the dynamic operating parameters and dynamic fitting coefficient. Instruction generation module: Generates industrial robot grasping instructions based on comprehensive dynamic adaptability and dynamic grasping adaptability.

2. The industrial robot automatic grasping system based on industrial vision according to claim 1, characterized in that, The dynamic grasping calibration standard includes dynamic speed threshold, dynamic accuracy threshold, and dynamic load adjustment threshold corresponding to the grasping task working condition type; The visual feature dynamic matching library contains dynamic contour features and dynamic texture feature data corresponding to the capture task condition types.

3. The industrial robot automatic grasping system based on industrial vision according to claim 2, characterized in that, Based on a pre-built visual feature dynamic matching library, target grasping task data for real-time grasping tasks in industrial settings is obtained. Includes the following steps: Dynamic feature matching is performed on the real-time grasping task data in the industrial field based on the dynamic contour feature and dynamic texture feature data of the visual feature dynamic matching library to obtain the target dynamic features. Based on the dynamic characteristics of the target, dynamic feature association data is extracted from the real-time capture task data in the industrial field to obtain the target task data to be captured corresponding to the dynamic characteristics of the target.

4. The industrial robot automatic grasping system based on industrial vision according to claim 1, characterized in that, The process of obtaining adapted capture task data and dynamic fit coefficients based on the actual dynamic parameters of the target capture task data and dynamic capture calibration standards includes the following steps: The actual dynamic parameters of the target task data are obtained based on the task condition type corresponding to the target task data to be captured. The actual dynamic parameters include the actual dynamic speed value, actual dynamic accuracy value, and actual dynamic load adjustment value corresponding to the target task data to be captured. If all values ​​of the actual dynamic parameters in the real-time grasping task data of the industrial site are within the threshold range corresponding to the dynamic grasping calibration standard, then the real-time grasping task data of the industrial site is determined to be the adapted grasping task data, and the dynamic deviation rate between the actual dynamic parameters and the corresponding threshold is determined. The dynamic fit coefficient of the adaptive grasping task data is obtained by comparing the dynamic deviation rate with the threshold corresponding to the dynamic grasping calibration standard.

5. The industrial robot automatic grasping system based on industrial vision according to claim 1, characterized in that, The comprehensive dynamic fit is obtained by processing and analyzing the data from the capture task, dynamic fit coefficient, production station, and robot operation terminal. This process includes the following steps: The first workstation adaptation coefficient is obtained based on the task dynamic distribution parameters and dynamic matching coefficient of the adapted capture task data. Mark the robot terminal corresponding to the data to be captured as the adapted terminal. The second workstation adaptation coefficient of the production workstation is obtained based on the dynamic operation parameters of the adapted operation terminal in the production workstation. The overall dynamic adaptation degree is obtained based on the adaptation coefficient of the first workstation and the adaptation coefficient of the second workstation.

6. The industrial robot automatic grasping system based on industrial vision according to claim 5, characterized in that, The first workstation adaptation coefficient is obtained based on the task dynamic distribution parameters and dynamic fit coefficient of the adapted capture task data. Includes the following steps: Set the dynamic weight of the working condition of the adapted crawling task data according to the working condition type of the crawling task data, and obtain the dynamic adaptation coefficient of the adapted crawling task data according to the dynamic fitting coefficient and the dynamic weight of the working condition. The task dynamic distribution parameters include the temporal dynamic distribution density and spatial dynamic coverage of the data to be captured in the production workstation. The first dynamic adaptation status value corresponding to the production station is obtained based on the time dynamic distribution density, spatial dynamic coverage range and dynamic adaptation coefficient. The total number of tasks for capturing real-time data in the industrial field at the production workstation is obtained. Based on the total number of tasks and the number of data entries for the adapted capture task at the production workstation, the second dynamic adaptation status value of the production workstation is obtained. The first workstation adaptation coefficient is obtained based on the first dynamic adaptation status value and the second dynamic adaptation status value.

7. The industrial robot automatic grasping system based on industrial vision according to claim 5, characterized in that, The second workstation adaptation coefficient is obtained based on the dynamic operation parameters of the adapted operation terminal in the production workstation. The specific steps include: Determine the collaborative correlation characteristics between the dynamic operation parameters of the adapted operation terminals in the production workstation, and filter out the target parameter items that affect the operation flow of the production workstation based on the collaborative correlation characteristics; Based on the production workstation workflow sequence, determine the work connection nodes of each adapted work terminal, and establish the correspondence between target parameter items and connection nodes; Based on the correspondence, the adaptability and coordination of the dynamic parameters of the terminal target at the connection node are determined, and the degree of cooperative adaptation is obtained. Based on the overall workflow efficiency requirements of each workstation, the coordination and adaptability of each connection node are prioritized to obtain critical connection nodes and normal connection nodes. The coordination deviation of the terminal target dynamic parameters is corrected based on the coordination and adaptation degree of key connection nodes and the adaptation status of normal connection nodes to obtain the terminal connection adaptation value. The adaptability of the workstation operation flow is evaluated based on the terminal connection adaptation value and the terminal operation connection frequency, and the second workstation adaptation coefficient of the production workstation is obtained.

8. The industrial robot automatic grasping system based on industrial vision according to claim 1, characterized in that, The terminal's dynamic operating parameters include the dynamic operating stability, dynamic energy consumption fluctuation value, and dynamic fault warning response speed of the adapted terminal corresponding to the data captured in the task.

9. The industrial robot automatic grasping system based on industrial vision according to claim 8, characterized in that, The dynamic capture adaptation degree of the adapted operation terminal is obtained based on the terminal's dynamic operating parameters and dynamic matching coefficient, specifically including the following steps: The dynamic stability adaptation coefficient is obtained based on the dynamic operational stability. The dynamic energy consumption adaptation coefficient is obtained by the ratio of the dynamic energy consumption fluctuation value to the preset dynamic energy consumption threshold, and the dynamic early warning adaptation coefficient is obtained by the dynamic fault early warning response speed. The dynamic stability adaptation coefficient, dynamic energy consumption adaptation coefficient, and dynamic early warning adaptation coefficient are weighted and summed to obtain the terminal dynamic operation adaptation coefficient; The dynamic capture adaptation degree of the adapted operation terminal is obtained by using the terminal dynamic operation adaptation coefficient and the dynamic matching coefficient of the adapted capture task data.

10. The industrial robot automatic grasping system based on industrial vision according to claim 9, characterized in that, The industrial robot's grasping instructions are generated based on the comprehensive dynamic adaptability and dynamic grasping adaptability, specifically including the following steps: Prioritize the overall dynamic adaptation and dynamic crawling adaptation to form an adaptation deviation value; The weight ratio of overall dynamic fit and dynamic capture fit is adjusted based on the fit deviation value. The final adaptation evaluation value is obtained based on the comprehensive dynamic adaptation degree, dynamic grasping adaptation degree, and weight ratio. The basic parameters of the grasping action are determined based on the final adaptation evaluation value and the terminal dynamic operation parameters of the adapted operation terminal. The basic parameters include grasping force, motion trajectory, and posture adjustment range. Generate a capture instruction that includes action sequence, execution timing and safety threshold based on the basic parameters and the workstation parameters of the production station.

Images