Search

CN-121979162-A - Digital twinning-fused part production control method and system

CN121979162ACN 121979162 ACN121979162 ACN 121979162ACN-121979162-A

Abstract

The invention relates to the technical field of component production control, in particular to a component production control method and system integrating digital twinning. According to the system, the regional image is subjected to gridding segmentation, so that accurate positioning distribution of parts is realized, the problem of repeated display across grids is solved, in the subsequent staged evaluation process, the staged evaluation route of the parts is constructed, the digital twin technology is utilized to simulate the evaluation process, the marking parameters of the corresponding parts in different stages are obtained, the marking parameters are utilized to regulate the evaluation route of the next stage, the evaluation result of each stage is obtained, the problem of repeated parameter acquisition in the evaluation process is avoided, the closed loop of the production control of the parts is ensured, and the abnormal response efficiency is improved. And finally, determining the production state of the current part according to the evaluation results of each stage, and carrying out adaptive adjustment on production equipment by combining the production state.

Inventors

  • RONG WEIBIN
  • LI TIANYI
  • YAN JIHONG

Assignees

  • 哈尔滨工业大学

Dates

Publication Date
20260505
Application Date
20260407

Claims (10)

  1. 1. The production control method of the fused digital twin component is characterized by comprising the following steps of: s1, configuring monitoring equipment and a monitoring route, and acquiring an image of a part placement area to obtain an area image; S2, carrying out gridding segmentation on the regional image, and carrying out clear positioning distribution on each part; s3, constructing a part staged evaluation route, and utilizing a digital twin technology to simulate an evaluation process to obtain the marking parameters of the corresponding parts in different stages; s4, extracting the sign parameters of each stage, regulating and controlling an evaluation route of the next stage by using the sign parameters, regulating and controlling monitoring equipment according to the evaluation route, and obtaining an evaluation result of each stage; s5, positioning the current production state of the parts according to the evaluation results of each stage, and carrying out adaptive adjustment on production equipment by combining the production state.
  2. 2. The method for controlling production of a component with digital twinning integration according to claim 1, wherein the monitoring device configured in S1 is a 3D structured light camera.
  3. 3. The method for controlling the production of a component with digital twinning integration according to claim 2, wherein the method for gridding and dividing the area image in S2 comprises the following steps: S2.1, determining the size of a unit grid according to the size of the part placement area; S2.2, mapping the region image after grid division into a two-dimensional coordinate system, and carrying out coordinate positioning on each unit grid through the two-dimensional coordinate system; S2.3, obtaining masks of each part by using the example segmentation model, and extracting accurate outer contours of the parts at different positions by matching with binary pixel matrixes in the example segmentation model; s2.4, determining the distribution positions of the parts in a two-dimensional coordinate system according to the outer contour of the parts.
  4. 4. The method for controlling production of digital twinned integrated components according to claim 3, wherein the method for performing the clear positioning distribution on each component in S2 comprises the following steps: S2.10, traversing all unit grids, and calculating the intersection area of the outer contour of the current part and the current unit grid; s2.11, carrying out single-area positioning on the part based on the intersecting area, and extracting all unit grids intersecting with the outer contour of the current part; S2.12, selecting a unit grid with the largest intersection area, distributing the part to the unit grid, marking the unit grid as a main weight grid, and marking the rest unit grids intersected with the outer contour of the current part as auxiliary weight grids; S2.13, traversing all unit grids, and for each unit grid: drawing only those parts assigned to the current unit mesh; Parts overlapping the unit grid but having the main right belonging to other unit grids are hidden.
  5. 5. The method for controlling production of digital twinned integrated components according to claim 4, wherein the step of constructing the component staged evaluation route in step S3 includes component placement distribution, component number of unit grids, and component quality monitoring.
  6. 6. The method for controlling production of integrated digital twin parts according to claim 5, wherein the marking parameter of the part placement distribution is a coordinate mark and a positional relationship of each part stack, and the marking parameter of the number of parts in the unit grid is a difference rate.
  7. 7. The method for controlling production of a component integrated with digital twinning according to claim 6, wherein the method for evaluating the component placement distribution comprises the steps of: s3.1, marking the instance IDs of all parts of different unit grids; s3.2, extracting 3D point clouds belonging to corresponding parts in each unit grid from the 3D point clouds corresponding to the original part mask according to the instance ID, marking the point clouds as point sets, wherein each point set represents one part in the unit grid; s3.3, merging point sets of all parts in the current unit grid to obtain a total point cloud set representing all parts in the unit grid; S3.4, constructing a space rectangular coordinate system, and dividing a bearing plane from the total point cloud set by adopting a random sampling consistency plane fitting algorithm; s3.5, calculating the vertical distance from the point with the largest Z coordinate value in the total point cloud set to the fitted bearing plane, and marking the vertical distance as the stacking height of the part stack; S3.6, processing all 3D point clouds in the current unit grid by using a clustering algorithm based on Euclidean distance, classifying spatially adjacent points into the same cluster, and classifying distant points into different clusters; S3.7, determining the number of clusters obtained by a clustering algorithm as the number of piled parts in the current unit grid; And S3.8, counting the number of piled parts in each unit grid and the piling height of the parts, wherein the parts are summarized and marked as parts placement distribution of the corresponding unit grid.
  8. 8. The method for controlling production of components by fusion digital twinning according to claim 6, wherein the method for evaluating the number of components of the unit grid comprises the steps of: s3.10, counting the single-placed parts in the unit grid according to the part mask; s3.11, combining the diversity of the part masks, and counting single parts under different placing postures; s3.12, obtaining the number of single parts with different placing postures in the unit grids, and marking the number as a difference rate; S3.13, when the parts are judged to be stacked according to the difference rate, combining the coordinate marks and the position relation of each part stack, constructing an image two-wheel acquisition route of the 3D structure light camera, and carrying out targeted image acquisition on the part stacks of different unit grids; S3.14, identifying and counting according to the part masks in the stacked state, and counting the number of parts in each unit grid.
  9. 9. The method for controlling production of a component with digital twinning integration according to claim 6, wherein the method for evaluating quality monitoring of the component comprises the steps of: s3.20, obtaining the difference rate of each unit grid ; S3.21, making a difference rate threshold And the difference rate is carried out on each unit grid And (3) comparison: Will be the difference rate Below the difference rate threshold Removing the unit grids of the (3); The difference rate is not lower than the difference rate threshold value The unit grids of (2) are marked as evaluation unit grids; S3.22, according to the difference rate The size performs order sequencing on each evaluation unit grid; s3.23, sequentially collecting sample images of each evaluation unit grid according to the sequence numbers of the order sequences to obtain a corresponding number of sample parts; S3.24, comparing part images in the standard database, and evaluating the abnormality rate of the parts in the current placement area.
  10. 10. A system for implementing the fused digital twin component production control method of claim 1, comprising a device side, a communication side, and a processing side; The equipment end is provided with a 3D structure light camera for collecting images of the component placement areas, the 3D structure light camera is matched with a control equipment for position adjustment to perform directional image collection work, and the equipment end is also provided with a production equipment for adapting to the current component stacking state by regulating and controlling the parameters of the production equipment; The communication end is used for constructing a network channel between the equipment end and the processing end, is used for uploading the area image acquired by the equipment end to the processing end, is used for processing the area image by utilizing an edge computing module carried by the processing end, is used for constructing a part staged evaluation route, is used for digital twin simulation evaluation, is used for extracting mark parameters of each stage, is used for regulating and controlling an evaluation route of the next stage, is used for acquiring an evaluation result of each stage, is matched with the network channel to be fed back to the equipment end, and the control equipment of the equipment end performs position route planning adjustment on the 3D structured light camera according to the evaluation result and the evaluation route, and is used for executing the evaluation route and the targeted acquisition work.

Description

Digital twinning-fused part production control method and system Technical Field The invention relates to the technical field of component production control, in particular to a component production control method and system integrating digital twinning. Background In the field of intelligent manufacturing, quality control and process optimization of component production are highly dependent on real-time evaluation and feedback of production results. The existing mainstream production control method generally follows a feedback paradigm of sensing, evaluating and regulating, namely, state data (such as stacking gesture, appearance flaw, dimensional tolerance and the like) of parts after production are collected through visual sensors, laser measurement and other devices, the state data are fed back to a central control system, and then process parameters (such as mechanical arm track, feeding quantity of a processing machine tool, spraying pressure and the like) of production line devices are adaptively adjusted according to the state data, so that quality is improved in subsequent production. However, as the product customization degree increases and the production tact increases, the inherent limitations of this traditional paradigm are increasingly highlighted, and the core problem is serialization of the evaluation flow and islanding of the data application, which is embodied as follows: The first, multi-step serial evaluation results in an efficiency bottleneck in that the existing evaluation process is typically broken down into multiple single, sequentially executed sub-tasks. For example, the 2D vision is used for preliminary positioning and counting, then the 3D sensor is triggered to perform fine scanning on the specific area to obtain the gesture, and finally the data is transferred to another independent quality analysis module for defect judgment. The steps are linearly carried out according to a fixed flow, each step needs to wait for finishing the last step and transmitting data, a large amount of computing resources are in an idle state during the last step, and delay of any link is accumulated, so that the whole response period from state acquisition to generation of a regulation instruction is overlong, and the real-time requirement of high-beat dynamic production cannot be met. Secondly, data island prevents information fusion and depth utilization, namely data (such as texture information in a 2D image, space geometric information in a 3D point cloud and semantic information in a quality inspection result) generated in different evaluation steps are often processed by independent subsystems, and a unified data model is not used for association and expression. For example, three-dimensional posture data of a part is separated from scratch detection results of the surface of the part at a data level, and a control system has difficulty in directly knowing deep association knowledge such as what kind of defects are more likely to occur at a specific position under what posture. This results in massive high-value process data being used only to complete the current single judgment, the process-quality association rules hidden behind it being unexcavated, and the data utilization being low. Therefore, a method and a system for controlling the production of parts by integrating digital twinning, which can break the barrier of the evaluation step and realize the full-flow data penetration, are needed. Disclosure of Invention The invention aims to provide a method and a system for controlling production of components fused with digital twinning, which are used for solving the problems in the background technology. In order to achieve the above object, one of the objects of the present invention is to provide a method for controlling production of a component with digital twinning integration, comprising the steps of: s1, configuring monitoring equipment and a monitoring route, and acquiring an image of a part placement area to obtain an area image; S2, carrying out gridding segmentation on the regional image, and carrying out clear positioning distribution on each part; s3, constructing a part staged evaluation route, and utilizing a digital twin technology to simulate an evaluation process to obtain the marking parameters of the corresponding parts in different stages; s4, extracting the sign parameters of each stage, regulating and controlling an evaluation route of the next stage by using the sign parameters, regulating and controlling monitoring equipment according to the evaluation route, and obtaining an evaluation result of each stage; s5, positioning the current production state of the parts according to the evaluation results of each stage, and carrying out adaptive adjustment on production equipment by combining the production state. Preferably, the monitoring device configured in S1 is a 3D structured light camera. Preferably, the method for gridding and segmenting the area image in S