Method, system and computer device for designing software to cooperatively control a knitwear equipment

Abstract

This invention relates to the field of CNC technology, specifically to a method, system, and computer device for collaborative control of knitting equipment using design software. The steps include: acquiring knitting design data output by the design software; performing semantic analysis on the design data to obtain structured instructions containing knitting rules, pattern features, and process parameters; and establishing a semantic mapping between the design data and the knitting equipment instructions. This invention constructs a semantic collaborative mechanism between design software and knitting equipment, which not only improves the quality stability of knitted products, such as the consistency of pattern accuracy and loop density, but also achieves efficient collaboration between design and manufacturing. For example, design modifications can be quickly updated with new instructions without requiring equipment downtime for re-analysis. This provides technical support for the integration of design and manufacturing in the knitting industry and promotes the upgrade of knitting manufacturing from mechanical execution to intelligent collaboration.

Description

Technical Field

[0001] This invention relates to the field of numerical control technology, specifically to a method, system, and computer device for designing software to collaboratively control knitting equipment. Background Technology

[0002] Design software, as the carrier of design intent, needs to transform pattern ideas, process rules, and parameter requirements into executable manufacturing instructions; while knitting equipment, such as computerized flat knitting machines and circular knitting machines, acts as the execution end of manufacturing intent, needing to accurately interpret design instructions and complete the knitting task. In existing technologies, the collaboration between the two largely relies on a linear model of design file transmission and direct equipment parsing: design software outputs standardized DXF files containing graphics, dimensions, and other information, and knitting equipment extracts geometric features such as lines and coordinates from the files to generate mechanical motion instructions. However, this collaborative method does not transcend the data-level transmission, lacking semantic association of design data and dynamic linkage of equipment status.

[0003] Specifically, the core design intent in design software, such as the texture direction of jacquard patterns, weaving rules such as the gradient change of loop density, and process parameters such as the dynamic threshold of yarn tension, often exists in unstructured graphic or text form. This cannot be directly mapped into structured instructions such as needle bed movement trajectory and yarn feeding rhythm that knitting equipment can understand. Even if some systems attempt to convert design data into equipment instructions, they are mostly simple one-to-one mappings, such as directly mapping jacquard positions to needle numbers on the needle bed, without considering the semantic relationship between design features and equipment execution logic. For example, the continuity of jacquard patterns should correspond to the coherence of needle bed oscillation.

[0004] Furthermore, real-time status data generated by knitting equipment during operation, such as machine speed fluctuations and yarn tension deviations, are not correlated with corresponding features in the design data. Deviation adjustments largely rely on operator judgment or simple threshold triggers, such as stopping the machine when yarn tension exceeds the upper limit. Precise optimization based on design priorities, such as prioritizing loop density consistency over knitting efficiency, is impossible. This semantic fragmentation and inaccurate adjustment coordination mode makes it difficult to accurately convey design intent to the manufacturing process. This not only causes significant fluctuations in key quality indicators of knitted sweaters, such as pattern accuracy and loop density (e.g., jacquard position deviations and uneven loop tension), but also reduces the efficiency of design-manufacturing collaboration. For example, design modifications require regenerating all instructions, necessitating equipment shutdown for adjustments. This has become a technical bottleneck restricting the integration of design and manufacturing in the knitted sweater industry. Summary of the Invention

[0005] This disclosure presents a method, system, and computer device for designing software collaborative control of knitting sweater equipment, with the aim of overcoming at least one of the deficiencies in the prior art.

[0006] To achieve the above objectives, the technical solution disclosed in this invention is as follows:

[0007] According to one aspect of this disclosure, a method for designing software-coordinated control of a knitting sweater equipment is provided, comprising the steps of:

[0008] Obtain the knitted sweater design data output by the design software, perform semantic parsing on the design data to obtain structured instructions containing knitting rules, pattern features and process parameters. Semantic parsing is used to establish a semantic mapping between design data and knitting equipment instructions.

[0009] The structured instructions are sent to the knitting equipment, and the real-time operating status data returned by the knitting equipment is received.

[0010] Based on the deviation between real-time operating status data and structured instructions, adjustment instructions are generated through a collaborative control model. The collaborative control model is used to optimize the adjustment strategy based on the deviation characteristics.

[0011] The adjustment command is sent to the knitting equipment to control the knitting equipment to adjust its operating parameters to perform the knitting task.

[0012] Furthermore, semantic parsing includes: extracting multimodal features from the design data, including pattern and texture features of the design image, process description features of the design text, and numerical features of the design parameters; based on a pre-built knitting process knowledge graph, mapping the multimodal features to corresponding knitting rules, pattern features, and process parameters, with the knowledge graph used to store the semantic association between design features and knitting equipment instructions.

[0013] Furthermore, the steps for receiving real-time operating status data returned by the knitting equipment include: receiving sensor data from the knitting equipment, yarn status image data captured by the camera, and historical operating data of the knitting equipment. The sensor data includes machine speed sensor data, yarn tension sensor data, and needle bed position sensor data. The sensor data, image data, and historical operating data are fused to obtain real-time operating status data in order to eliminate data noise and supplement missing data.

[0014] Furthermore, the steps for generating adjustment instructions include: dividing the features in the structured instructions into key features and secondary features. Key features include coil density, jacquard pattern position, and yarn tension threshold. Secondary features include machine speed and needle bed oscillation frequency. Calculate the first deviation between the key features in the real-time operating status data and the corresponding key features in the structured instructions, and the second deviation between the secondary features and the corresponding secondary features in the structured instructions. Based on the weighting coefficients of the first and second deviations, obtain the comprehensive deviation, and generate adjustment instructions through a collaborative control model.

[0015] Furthermore, after generating adjustment instructions through the collaborative control model, the process also includes: inputting the adjustment instructions into the digital twin model of the knitting equipment to simulate the operating state and knitting actions of the knitting equipment after executing the adjustment instructions; and sending the adjustment instructions to the knitting equipment when the simulated operating state meets the preset equipment safety threshold and the simulated knitting actions meet the requirements of the design data.

[0016] Furthermore, sending structured instructions to the knitting equipment includes: when multiple knitting equipment are performing knitting tasks simultaneously, obtaining the current task progress and resource usage of each knitting equipment, coordinating the sending order of structured instructions and the process parameters of each knitting equipment, avoiding resource conflicts between equipment, and the resource usage includes the remaining amount of yarn and the idle status of the needle bed.

[0017] Furthermore, after obtaining the knitted sweater design data output by the design software, the process also includes: monitoring whether the design software has modified the design data; when modification is detected, re-semantically parsing the modified design data to obtain updated structured instructions; and adjusting the previously generated adjustment instructions based on the updated structured instructions to ensure consistency between the adjustment instructions and the updated structured instructions.

[0018] Furthermore, after sending the adjustment command to the knitting equipment, the process also includes: collecting the operating status data of the knitting equipment after executing the adjustment command and the quality inspection data of the knitted products, including the coil density test data and the pattern accuracy test data; feeding back the operating status data and the quality inspection data to the design software, and adjusting the design parameters according to the feedback results to optimize the subsequent knitting design data.

[0019] According to another aspect of this disclosure, a system for designing software to collaboratively control a knitting sweater machine is provided, and a method for implementing the above-mentioned design software collaborative control of the knitting sweater machine includes:

[0020] The design data parsing module is used to acquire the knitted sweater design data output by the design software, perform semantic parsing on the design data, and obtain structured instructions containing knitting rules, pattern features and process parameters. The design data parsing module includes a multimodal semantic mapping unit, which is used to extract multimodal features from the design data and map them to the corresponding knitting rules, pattern features and process parameters based on a pre-built knitting process knowledge graph.

[0021] The status interaction module is used to send structured instructions to the knitting sweater equipment and receive real-time operating status data returned by the knitting sweater equipment. The status interaction module includes a multi-source data fusion unit, which is used to fuse sensor data of the knitting sweater equipment, yarn status image data collected by the camera, and historical operating data to obtain real-time operating status data.

[0022] The deviation adjustment module is used to generate adjustment instructions through a collaborative control model based on the deviation between real-time running status data and structured instructions. The deviation adjustment module includes a hierarchical deviation evaluation unit and a deep reinforcement learning control unit. The hierarchical deviation evaluation unit is used to divide the features in the structured instructions into key features and secondary features and calculate the corresponding deviations. The deep reinforcement learning control unit is used to optimize the adjustment strategy based on the deviation features through a deep reinforcement learning model.

[0023] The equipment control module is used to send adjustment instructions to the knitting equipment and control the knitting equipment to adjust its operating parameters to perform the knitting task. The equipment control module includes a digital twin verification unit, which is used to input the adjustment instructions into the digital twin model of the knitting equipment to simulate the execution effect. When the simulation result meets the requirements, the adjustment instruction is sent.

[0024] According to another aspect of this disclosure, a computer device is provided, including a processor, a memory, and a computer program stored in the memory, wherein the processor executes the computer program to implement the above-described method for collaboratively controlling a knitting sweater device with design software.

[0025] The beneficial effects of this invention are:

[0026] This invention effectively solves the technical problems of semantic separation between design intent and device instruction semantics and inaccurate deviation adjustment in the prior art through a collaborative mechanism of semantic parsing, real-time status interaction and collaborative deviation adjustment.

[0027] Specifically, this invention performs semantic parsing on design data, extracts weaving rules, pattern features, and process parameters, and establishes a semantic mapping with knitting equipment instructions. This transforms unstructured design intentions in the design software, such as the texture direction of jacquard patterns, into structured instructions that the equipment can understand, such as the trajectory and frequency of needle bed oscillation. This ensures the accurate association between design features and equipment execution logic, avoiding the intention deviation caused by direct transmission of design data in the prior art.

[0028] Furthermore, this invention receives real-time operating status data of the knitting equipment and compares it with structured instructions. Combined with a collaborative control model, it optimizes the weighting of deviation features, such as key and secondary features, to generate adjustment instructions prioritizing design intent. For example, it prioritizes adjusting loop density deviations affecting product quality, followed by machine speed deviations affecting efficiency. This achieves dynamic linkage between design data, equipment status, and adjustment strategies, solving the problem of prioritizing deviation adjustments based on experience or simple thresholds in existing technologies. By sending adjustment instructions to the equipment to control its operating parameters, a control logic of design, execution, feedback, and optimization is formed, ensuring the accurate execution of design intent in the manufacturing process.

[0029] Furthermore, this invention constructs a semantic collaborative mechanism between design software and knitting equipment, which not only improves the quality stability of knitted sweater products, such as the consistency of pattern accuracy and loop density, but also realizes efficient collaboration between design and manufacturing. For example, design modifications can be quickly updated with instructions, and the equipment does not need to be stopped for re-analysis. This provides technical support for the knitting industry to achieve integrated design and manufacturing, and promotes the upgrade of knitting manufacturing from mechanical execution to intelligent collaboration.

[0030] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0031] Figure 1 This is a flowchart illustrating a method for designing software to collaboratively control a knitting sweater equipment according to an embodiment of the present invention;

[0032] Figure 2 This is a semantic mapping topology diagram of design features and device instructions in one embodiment of the present invention;

[0033] Figure 3a This is a heatmap showing the weight distribution of different design features in multimodal extraction according to an embodiment of the present invention.

[0034] Figure 3b This is a mapping result diagram of knowledge graph to multimodal features in one embodiment of the present invention;

[0035] Figure 4a This is a schematic diagram of stratification deviation analysis in one embodiment of the present invention;

[0036] Figure 4b This is a schematic diagram of control decision-making in one embodiment of the present invention;

[0037] Figure 5 This is a schematic diagram of the verification and optimization process of a digital twin system in one embodiment of the present invention;

[0038] Figure 6a This is a diagram of the original sensor data from a multi-source data fusion process in one embodiment of the present invention.

[0039] Figure 6b This is a feature extraction map of a yarn state image processed by multi-source data fusion in one embodiment of the present invention;

[0040] Figure 6c This is a correlation analysis diagram of historical operational data in a multi-source data fusion processing embodiment of the present invention;

[0041] Figure 6d This is a state diagram after multi-source data fusion processing in one embodiment of the present invention;

[0042] Figure 7 This is a schematic diagram illustrating the optimization of the deviation adjustment strategy in one embodiment of the present invention;

[0043] Figure 8a This is a schematic diagram illustrating the verification effect of the digital twin model before adjustment in one embodiment of the present invention;

[0044] Figure 8b This is a schematic diagram illustrating the verification effect of the adjusted digital twin model in one embodiment of the present invention. Detailed Implementation

[0045] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0046] In embodiments of the present invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0047] The present invention provides the following preferred embodiments:

[0048] Example 1: To address the problems of semantic disconnect between design intent and equipment instructions, and inaccurate deviation adjustment in existing technologies, this example provides a method for collaborative control of knitting equipment by design software. It establishes a semantic association between design data and equipment instructions through semantic parsing, and combines real-time status interaction and collaborative deviation adjustment to achieve accurate transmission and dynamic optimization of design intent to the manufacturing process. For example... Figure 1 As shown, the steps of the method are as follows:

[0049] S100: Acquires knitted sweater design data output by the design software, performs semantic parsing on the design data, and obtains structured instructions containing knitting rules, pattern features and process parameters. Semantic parsing is used to establish a semantic mapping between design data and knitting equipment instructions.

[0050] S200: Sends structured instructions to the knitting equipment and receives real-time operating status data returned by the knitting equipment.

[0051] S300: Based on the deviation between real-time operating status data and structured instructions, adjustment instructions are generated through a collaborative control model. The collaborative control model is used to optimize the adjustment strategy based on the deviation characteristics.

[0052] S400: Sends adjustment instructions to the knitting equipment to control the knitting equipment to adjust operating parameters to perform the knitting task.

[0053] Specifically, the knitwear design data output by design software typically includes graphical patterns, textual process rules, and numerical process parameters. Examples include CAD files for jacquard patterns, textual descriptions such as "loop density increases gradually from the neckline to the hem," and parameter settings such as "yarn tension threshold 15cN." Figure 2 As shown, this data exists in unstructured or semi-structured form and cannot be directly understood by knitting equipment. It needs to be converted into structured instructions that conform to the equipment's execution logic through semantic parsing. The key to semantic parsing lies in extracting the intent from the design data and establishing a semantic mapping with the equipment instructions, rather than simple format conversion. First, considering the multimodal characteristics of the design data, image features (such as the texture direction of jacquard patterns and the geometric shape of pattern outlines), text features (such as the process description of "changes in coil density gradient"), and parameter features (such as the dynamic threshold of yarn tension and the specific values ​​of dimensional parameters) are extracted. For example... Figure 3a and Figure 3b As shown, multimodal feature parsing transforms unstructured design information into quantifiable feature vectors through image recognition, natural language processing, and numerical extraction techniques. Subsequently, based on a pre-constructed knitting process knowledge graph, these feature vectors are mapped to corresponding knitting rules, pattern features, and process parameters. The knowledge graph stores the semantic relationships between design features and equipment execution logic. For example, "continuity of jacquard texture" corresponds to "coherence of needle bed oscillation" to ensure the integrity of the jacquard pattern, and "stability of yarn feeding rhythm" to avoid pattern interruptions caused by yarn breakage. "Gradient density" corresponds to "gradient adjustment of machine speed control" to ensure that the loop density changes uniformly with the knitting progress. Figure 2As shown, the semantic mapping topology clearly demonstrates the many-to-many semantic relationships between design features such as jacquard texture, gradient density, and yarn tension, and equipment instructions such as needle bed trajectory, yarn feeding rhythm, and machine speed control. This relationship is not a simple one-to-one mapping, but rather considers the semantic consistency between design features and equipment execution logic. For example, jacquard texture corresponds not only to the needle bed trajectory but also to the oscillation path, because the continuity of the jacquard pattern requires the coherence of the needle bed oscillation. Through semantic parsing, design data is transformed into structured instructions containing weaving rules such as "loop density gradient change rules," pattern features such as "jacquard pattern texture direction," and process parameters such as "yarn tension threshold 15cN." These instructions directly correspond to the equipment execution logic, such as needle bed movement trajectory, yarn feeding rhythm, and machine speed control, to ensure the accurate transmission of design intent.

[0054] Furthermore, after the structured instructions are sent to the knitting equipment, the equipment begins to execute the knitting task. During execution, the equipment returns real-time operating status data through sensors (machine speed sensor, yarn tension sensor, needle bed position sensor), cameras, and storage modules. Sensor data reflects the equipment's real-time operating parameters, such as machine speed fluctuations, yarn tension deviations, and needle bed position errors. Camera data reflects the yarn's condition, such as yarn breakage, excessive fuzz, and yarn feeding uniformity. Historical operating data provides information on the equipment's past performance, such as the yarn tension variation pattern at a certain machine speed and the jacquard position error corresponding to a certain needle bed trajectory. These data come from different sources and have different formats, requiring multi-source data fusion processing to eliminate noise, supplement missing information, and integrate data. Figure 6a , Figure 6b , Figure 6c as well as Figure 6d As shown, multi-source data fusion utilizes Kalman filtering (to eliminate sensor noise), interpolation (to supplement missing data), and feature fusion (such as associating visual features of yarn condition images with tension sensor data) to obtain accurate and complete real-time operating status data. For example, when abnormal fluctuations occur in yarn tension sensor data, the sensor data can be corrected using the "machine speed-tension" relationship in historical operating data and yarn condition images from the camera (such as whether the yarn is slack), ensuring the reliability of real-time status data.

[0055] Furthermore, based on the deviation between real-time operating status data and structured instructions, adjustment instructions are generated through a collaborative control model. Deviation calculations need to be performed separately for key features (such as coil density, jacquard position, and yarn tension) and secondary features (such as machine speed and needle bed oscillation frequency) in the structured instructions. Figure 7As shown. First, compare the differences between real-time status and structured instructions: for example, the real-time coil density is 20 coils / inch (structured instructions are 22 coils / inch), with a deviation of -2 coils / inch; the real-time machine speed is 80 rpm (structured instructions are 75 rpm), with a deviation of +5 rpm. The core of the collaborative control model lies in optimizing and adjusting strategies based on deviation characteristics, considering the priority of design intent, i.e., key features directly affect product quality, while secondary features affect production efficiency, and the accuracy of key features must be prioritized. For example... Figure 4a and Figure 4b As shown, the layered deviation analysis divides features into a critical layer (loop density, jacquard position, yarn tension) and a secondary layer (machine speed, oscillation frequency), assigning higher weights to critical features and calculating the overall deviation. Subsequently, the collaborative control model generates adjustment instructions based on the overall deviation and deviation characteristics: for example, prioritizing the adjustment of loop density deviation by increasing the needle bed pressure to the set value to increase loop density, and then adjusting the machine speed deviation, such as reducing the machine speed to 75 rpm to maintain loop density stability. The content of the adjustment instructions must specifically correspond to the equipment's execution parameters, such as "increase needle bed pressure by 5%", "decrease machine speed by 5 rpm", or "increase yarn feeding rhythm by 10%", to ensure accurate execution by the equipment.

[0056] Furthermore, after the adjustment command is sent to the knitting equipment, the equipment adjusts its operating parameters according to the command and executes the knitting task. To ensure the accuracy of the adjustment command, it can be input into the digital twin model of the knitting equipment, such as... Figure 5 As shown, the simulation model depicts the operating status and knitting motion of the equipment after executing an adjustment command. Based on the equipment's physical characteristics (such as the mechanical limitations of the needle bed movement and the mechanical properties of the yarn) and historical data (such as changes in operating status corresponding to a specific adjustment command), the digital twin model accurately simulates changes in machine speed, yarn tension, and needle bed movement trajectory after adjustment. For example, after simulating a 5% increase in needle bed pressure, it checks whether the loop density increases to 22 loops / inch, and whether the yarn tension stabilizes below 15 cN after a 5 rpm decrease in machine speed. When the simulated operating status meets preset equipment safety thresholds (such as machine speed not exceeding 100 rpm and yarn tension not exceeding 20 cN) and the simulated knitting motion meets the design data requirements (such as jacquard position error less than 0.5 mm and loop density uniformity greater than 95%), an adjustment command is sent to the equipment to avoid equipment malfunctions or product defects caused by incorrect adjustment commands.

[0057] This embodiment establishes a semantic association between design intent and equipment instructions through semantic parsing, ensuring an accurate correspondence between design features and equipment execution logic. It acquires equipment operation data through real-time status interaction, providing a reliable basis for deviation analysis. Furthermore, it generates adjustment instructions that prioritize design intent through a collaborative control model, achieving dynamic linkage between design data, equipment status, and adjustment strategies. This collaborative mechanism solves the problems of semantic separation between design intent and equipment instructions, as well as inaccurate deviation adjustments, ensuring the accurate execution of design intent in the manufacturing process and improving the quality stability of knitted sweater products and the collaborative efficiency of design and manufacturing.

[0058] Example 2: To address the issue of unclear semantic relationships between design data and knitting equipment instructions, this example further refines the multimodal feature extraction and knowledge graph mapping steps in the semantic parsing process. Semantic parsing transforms design data into semantic information that the equipment can understand, and multimodal feature extraction is the foundation for achieving this transformation. Pattern texture features in the design image, such as the direction of jacquard textures and the edge features of pattern outlines, reflect the visual design intent of the sweater. Process description features in the design text, such as "using plain knit" and "cuff decrease density of 12 stitches / inch," convey specific process requirements. Numerical features in the design parameters, such as the set value of yarn tension and the specific values ​​of size parameters, quantify the technical indicators of the design. These multimodal features cover the visual, textual, and numerical dimensions of the design, comprehensively capturing different expressions of the design intent.

[0059] Based on a pre-constructed knitting process knowledge graph, multimodal features are mapped to corresponding knitting rules, pattern features, and process parameters. The knitting process knowledge graph stores the semantic relationships between design features and equipment instructions, such as the association between jacquard texture features and needle bed trajectory instructions, the association between process description features and needle selection logic instructions, and the association between numerical features and yarn feeding rhythm instructions. For example... Figure 3a and Figure 3b As shown, the multimodal feature parsing process includes two key sub-steps: Figure 3a The heatmap shows the weight distribution of different design features in multimodal extraction. For example, the extraction weight of "jacquard texture" in image features is as high as 0.85, and the weight of "yarn tension" in parametric features is 0.70, which clearly reflects the contribution of each design feature to multimodal semantics. Figure 3bThe polar coordinate graph visually presents the mapping results of the knowledge graph to multimodal features. Centered on target features such as "jacquard position" and "loop density," the polar radius length represents the mapping weight of image, text, and parameter features. For example, in the mapping weight of "jacquard position," image features account for 75%, text features account for 15%, and parameter features account for 10%, realizing the transformation of multimodal features into device-understandable semantics. For instance, the jacquard texture feature in the design image is mapped to the needle bed trajectory parameter in the knitting rules through the knowledge graph; the "plain knitting" process description feature in the design text is mapped to the loop structure parameter in the pattern feature; and the yarn tension value feature in the design parameters is mapped to the yarn feeding rhythm parameter in the process parameters.

[0060] The multimodal feature extraction in this embodiment ensures the comprehensiveness of design information and avoids information omissions that may occur due to single-modal features. Knowledge graph mapping establishes a semantic bridge between design features and device instructions, transforming abstract design intentions into specific parameters that the device can execute. Through this step, the semantic information in the design data is accurately parsed, laying the foundation for the subsequent generation of structured instructions.

[0061] Example 3: To accurately acquire the real-time operating status of the knitting equipment to support subsequent collaborative control, this example further optimizes the reception and processing of real-time operating status data. The accuracy of real-time operating status data directly affects the effectiveness of collaborative control; therefore, it is necessary to cover multiple dimensions of equipment operation and perform fusion processing. The received data types include sensor data, yarn status image data captured by cameras, and historical operating data: sensor data comes from machine speed sensors, yarn tension sensors, and needle bed position sensors, reflecting the equipment's operating speed, yarn tension, and the real-time position of the needle bed, respectively; the yarn status image data captured by cameras reflects the appearance of the yarn through visual information, such as whether there are problems like breakage or excessive fuzz; historical operating data includes data such as machine speed, yarn tension, and needle bed position during past operations, serving as a contextual reference for the current status.

[0062] like Figure 6a , Figure 6b , Figure 6c as well as Figure 6dAs shown, the multi-source data fusion processing flow consists of three core stages: data acquisition, preprocessing, fusion, and authenticity verification and weighted integration, resulting in unified real-time operating status data. Specifically, firstly, sensor data is preprocessed. A low-pass filtering algorithm is used to eliminate high-frequency noise in machine speed data, avoiding the impact of instantaneous fluctuations on status judgment. A moving average method is used to smooth yarn tension data, reducing numerical fluctuations caused by sensor jitter. The needle bed position sensor data is calibrated to ensure the accuracy of position information. Secondly, feature extraction is performed on the yarn status image data acquired by the camera. Image recognition algorithms are used to identify features such as yarn breakage points and hairiness, and these visual features are converted into numerical status parameters, such as "yarn breakage probability" and "hairiness index." Finally, sensor data, image data, and historical operating data are fused together. When a sensor data point is missing due to a fault, it is supplemented using similar operating condition data from historical data. When image data identifies yarn breakage, the yarn tension mutation information in the sensor data is combined to confirm the authenticity of the yarn condition. The same state parameter from multiple sources, such as the yarn tension sensor data and the hairiness index in the image data, is integrated using a weighted average method to obtain a more accurate real-time state value.

[0063] The reception of multi-source data covers three dimensions of equipment operation: numerical, visual, and historical data, ensuring the comprehensiveness of status information. Fusion processing, through preprocessing and multi-source integration, improves the accuracy and completeness of real-time operating status data; for example, the fused value of yarn tension has a 35% lower error than data from a single sensor. Through this embodiment, the real-time operating status of the equipment is accurately captured, providing a reliable foundation for subsequent deviation analysis and adjustment command generation.

[0064] Example 4: To generate precise adjustment instructions that match the equipment's operating status and achieve design-equipment synergy, this example further refines the adjustment instruction generation logic, introducing feature layering and deviation weighting processing steps. The generation of adjustment instructions needs to consider both the design intent and the current equipment state; therefore, the features in the structured instructions need to be categorized. The features in the structured instructions are divided into key features and secondary features: key features include coil density, jacquard pattern position, and yarn tension threshold; secondary features include machine speed and needle bed oscillation frequency.

[0065] Specifically, deviations are calculated to generate adjustment instructions. First, the first deviation between key features and structured instructions in the real-time operating status data is calculated, such as the difference between real-time coil density and design value, and the offset of jacquard position. Second, the second deviation between secondary features and structured instructions is calculated, such as the difference between real-time machine speed and design value, and the deviation of oscillation frequency. To balance the influence of both, a higher weight is assigned to the first deviation according to the knitting process requirements, such as a weight of 0.7 for key features and 0.3 for secondary features. The comprehensive deviation is obtained by "first deviation × key weight + second deviation × secondary weight", which reflects the overall degree of deviation between the equipment status and the design intent.

[0066] like Figure 7 As shown, the generation logic of adjustment instructions is based on a 3D surface model: peak areas of the surface, i.e., large deviations in key features, correspond to high adjustment intensity. For example, when the jacquard position offset exceeds 2mm, the adjustment strategy intensity reaches 0.9. Flat areas, i.e., large deviations in secondary features, correspond to low adjustment intensity. For example, when the machine speed deviation is 10%, the adjustment intensity is 0.4. The collaborative control model generates adjustment instructions based on the magnitude and direction of the comprehensive deviation. When the first deviation is large, instructions to adjust yarn tension or jacquard position are generated first (quickly correcting quality problems); when the second deviation is large, instructions to adjust machine speed or oscillation frequency are generated (balancing efficiency and equipment wear). For example, when image data identifies yarn breakage, the yarn tension threshold deviation in the first deviation reaches +30%, and sensor data shows a sudden change in yarn tension, the first deviation is confirmed, and the collaborative control model will prioritize generating an adjustment instruction to "reduce the yarn feeding rhythm by 20%"; when the machine speed is too high, and the secondary feature deviation reaches +15% but does not affect quality, an instruction to "reduce the machine speed by 10%" is generated to avoid excessive equipment wear.

[0067] Feature layering makes the generation of adjustment instructions more targeted, and prioritizing key features ensures the stability of finished product quality. Deviation weighting and comprehensive deviation calculation balance the influence of different features, avoiding over- or under-adjustment caused by a single feature deviation. The collaborative control model generates adjustment instructions based on comprehensive deviation, achieving dynamic matching between design intent and equipment operating status. Through this embodiment, the adjustment instructions accurately reflect design requirements and adapt to the real-time status of the equipment, providing precise instruction support for the collaborative control of design software and knitting equipment.

[0068] Example 5: To avoid operational risks or product quality issues that may arise from directly sending adjustment instructions to the knitting equipment, this example further optimizes the verification process for adjustment instructions. After the collaborative control model generates the adjustment instructions, a digital twin model simulation verification step is added.

[0069] The collaborative control model, after generating adjustment commands based on design data and equipment status, does not directly send the commands to the equipment. Instead, it first inputs them into the digital twin model of the knitting equipment. This model is constructed based on the equipment's physical structural parameters (such as needle bed dimensions and yarn feed mechanism mechanical characteristics), knitting process mechanical models (such as tension transmission laws during coil formation), and geometric constraints of the design data (such as the pattern accuracy requirements of jacquard textures). It can simulate the equipment's operating state and knitting actions after executing adjustment commands with high fidelity. Figure 5 As shown in the diagram, the verification and optimization process of the digital twin system clearly illustrates this process: After the adjustment command input, the digital twin model outputs real-time operating status data such as changes in machine speed, yarn tension fluctuations, and needle bed trajectory execution. At the same time, it simulates specific actions such as loop formation and jacquard texture presentation through a virtual knitting engine. For example, when simulating jacquard texture knitting, the model calculates the matching degree of the needle bed position, yarn feed amount, and loop density of each stitch to verify whether it meets the requirements of the design pattern.

[0070] During the simulation, the collaborative control model compares these data with preset equipment safety thresholds and design data requirements. Equipment safety thresholds include the upper limit of yarn tension (to prevent yarn breakage), the allowable machine speed range (to prevent equipment overload), and the limit of needle bed acceleration (to protect the mechanical structure), while design data requirements cover pattern errors in jacquard weave and the execution accuracy of dimensional parameters. When the simulated operating state does not exceed the safety thresholds and the knitting action meets the design data requirements, it indicates that the adjustment commands can achieve the design intent while ensuring equipment safety. Furthermore, if... Figure 8a and Figure 8b As shown, the digital twin model also generates heat maps of process stability before and after adjustment. By comparing the distribution of process stability on the two-dimensional plane of needle bed position and yarn tension, such as the increase in the proportion of high stability area after adjustment, the effect of adjustment command on process optimization is further verified.

[0071] In this embodiment, the collaborative control model will only send adjustment instructions to the knitting equipment when all the above simulation verifications pass. This step, through virtual verification using a digital twin model, identifies potential equipment malfunctions or product quality issues that may result from the adjustment instructions in advance, ensuring the safety and reliability of instruction execution.

[0072] Example 6: In order to solve the resource conflict problem that may occur when multiple knitting sweater devices perform knitting tasks at the same time, this example further refines the structured instruction sending coordination mechanism, and adds the device status and resource perception and coordination steps before sending instructions.

[0073] When multiple knitting machines are simultaneously performing knitting tasks, the collaborative control model first obtains the current task progress and resource usage of each machine in real time through the machine communication interface. Task progress includes data such as the number of rows completed and the remaining pattern knitting portions, such as unfinished areas of jacquard patterns; this data is reported in real time by the machine control system. Resource usage includes the remaining yarn amount and the idle status of the needle bed, collected by weight sensors on the yarn spool and transmitted to the collaborative control model, which then feeds back needle position information, such as the proportion of idle needle positions to total needle positions. Figure 6a , Figure 6b , Figure 6c as well as Figure 6d As shown in the diagram, the multi-source data fusion processing diagram illustrates the data fusion process. The collaborative control model integrates task progress and resource data from different devices to form a unified device status view.

[0074] Based on this view, the collaborative control model analyzes the matching degree between the resource requirements of each device and the currently available resources. For example, when the remaining yarn of a device is less than 10% of the current task, it is determined that its yarn resource is under pressure; when the number of idle needle positions on a device's needle bed is less than 20% of the total number of needle positions, it is determined that its needle bed resource status is under pressure. Subsequently, the collaborative control model coordinates the sending order of structured instructions: instructions are prioritized to devices with lagging task progress and ample resource usage, such as devices with sufficient remaining yarn and idle needle beds, to ensure that they can quickly advance the task; for devices with under pressure, the collaborative control model adjusts their process parameters to alleviate resource pressure. For example, for devices with insufficient remaining yarn, the yarn feeding rhythm is appropriately reduced, such as adjusting the yarn feeding speed from 100mm / s to 80mm / s, reducing yarn consumption per unit time and avoiding downtime due to yarn exhaustion; for devices with under pressure in needle bed idle status, the needle selection logic is adjusted, such as optimizing the needle position allocation order, improving the utilization efficiency of needle bed positions, and reducing needle bed conflicts.

[0075] Through the coordination mechanism in this embodiment, the collaborative control model can effectively avoid resource shortages or equipment waiting caused by multiple devices competing for the same resource (such as the same yarn roll or the same type of needle position), ensuring the smooth progress of multi-device knitting tasks. At the same time, the adjustment of process parameters also ensures that the equipment can maintain reasonable knitting efficiency under resource constraints, avoiding delays in the overall production schedule due to resource conflicts.

[0076] Example 7: In order to address the issue of inconsistencies between the adjustment instructions and design requirements after the design software modifies the design data, this example further optimizes the dynamic adjustment mechanism after the design data changes by adding design modification monitoring and instruction update steps after obtaining the design data.

[0077] Specifically, the collaborative control model maintains a connection with the design software through a real-time data interface, continuously monitoring whether the design software modifies the design data. When a designer modifies the jacquard pattern or changes dimensional parameters in the design software, the interface sends a change notification and the modified design data to the collaborative control model. At this point, the collaborative control model needs to re-parse the modified design data semantically to generate updated structured instructions.

[0078] like Figure 2 As shown, the semantic mapping topology diagram of design features and equipment instructions illustrates the correspondence between design data and equipment instructions. Jacquard textures correspond to needle bed trajectories and oscillation paths, dimensional parameters correspond to needle removal strategies, and process rules correspond to needle selection logic and yarn feeding rhythm. Modified design data needs to be remapped to the corresponding equipment instructions: for example, modified jacquard textures require denser needle bed trajectories, such as increasing the number of needle bed movement steps from 10 to 15 per centimeter, and adjusted oscillation paths, such as adjusting the oscillation amplitude from 20mm to 25mm; modified dimensional parameters require adjustments to the needle removal strategy, such as increasing the number of needle removal rows from 10 to 15. Figure 3a and Figure 3b As shown in the diagram, the multimodal feature parsing schematic illustrates the re-parsing process. The collaborative control model extracts multimodal features from the modified design data, and then, based on... Figure 2 The semantic mapping topology is used to remap these features to the corresponding device instructions, generating updated structured instructions.

[0079] Furthermore, the collaborative control model compares the updated structured instructions with the previously generated adjustment instructions, identifying inconsistencies. For example, the previous needle bed trajectory adjustment instructions were based on the old diamond pattern, while the new circular pattern requires a different needle bed trajectory. Subsequently, the collaborative control model adjusts the previous adjustment instructions to align with the updated structured instructions. For instance, it updates the needle bed trajectory adjustment instruction from "10 steps per centimeter" to "15 steps per centimeter," and the needle withdrawal strategy adjustment instruction from "10 rows of withdrawal" to "15 rows of withdrawal."

[0080] Through the dynamic adjustment mechanism in this embodiment, the collaborative control model ensures that adjustment commands are always consistent with design data, avoiding deviations between equipment execution actions and design requirements caused by design modifications, such as jacquard textures not conforming to the new pattern or dimensions not meeting new requirements. Simultaneously, the real-time monitoring and update process ensures that design changes can be quickly transmitted to the equipment control stage, maintaining the consistency between design and production.

[0081] Example 8: In order to verify the effect of the adjustment command and continuously optimize the design parameters, this example further sets up a data collection and feedback optimization step after sending the adjustment command to the knitting sweater equipment.

[0082] Specifically, after the knitting equipment completes the knitting task by executing adjustment instructions, the system automatically collects two types of data: one is the equipment's operating status data, including the actual execution path of the needle bed trajectory, the real-time changes in the yarn feeding rhythm, and the fluctuations in machine speed control. This data is collected and transmitted in real time through the equipment's built-in sensors. The other type is the quality inspection data of the knitted product. Among them, the loop density detection data is obtained by scanning the loop arrangement density of the knitted fabric line by line through a high-precision density sensor, while the pattern accuracy detection data is obtained by capturing the pattern outline of the knitted product through an industrial camera, and then comparing it with the original pattern features in the design software using image recognition algorithms to obtain the deviation value of the pattern outline.

[0083] Furthermore, the aforementioned operational status data and quality inspection data are fed back to the design software via the status interaction module. Upon receiving the feedback data, the design software compares it with the original design parameters, such as the design density of the jacquard texture and the distribution rules of the gradient density, against the expected effect of the adjustment command, analyzing whether the execution effect of the adjustment command meets the design requirements. For example, if the coil density detection data shows a deviation between the actual density and the design density, the design software will adjust the corresponding design parameters, such as the gradient density setting, and use a digital twin model, such as... Figure 5 The digital twin system verification and optimization process shown simulates the impact of adjusted design parameters on subsequent knitting tasks, ensuring that the optimized design parameters more accurately match the operating characteristics of the knitting equipment. Through this closed-loop feedback mechanism, the design software can continuously optimize subsequent knitting design data, gradually reducing the deviation between design expectations and actual knitting results.

[0084] Example 9: In order to achieve collaborative control between design software and knitting equipment, this example provides a system for collaborative control of knitting equipment by design software. The system realizes the full-process collaboration of design data parsing, status interaction, deviation adjustment and equipment control through modular design.

[0085] Specifically, the design data parsing module is the system's input interface. Its function is to acquire the knitted sweater design data output by the design software, including patterns, process rules, and size parameters, and to perform semantic parsing on this data, converting unstructured design data into structured instructions containing knitting rules, pattern features, and process parameters. The multimodal semantic mapping unit is a key component of this module. It can extract multimodal features from the design data, such as image features from patterns, text features from process rules, and numerical features from size parameters. Based on a pre-built knitting process knowledge graph, it includes the semantic mapping relationship between design features and equipment instructions, such as... Figure 2The semantic mapping topology between the design features and equipment instructions shown maps these multimodal features to corresponding weaving rules, pattern features and process parameters, ensuring that the design data can be understood and executed by the equipment.

[0086] Furthermore, the status interaction module serves as the system's communication hub. On one hand, it sends structured instructions generated by the design data parsing module to the knitting equipment, guiding it to execute knitting tasks; on the other hand, it receives real-time operational status data returned by the equipment. To ensure the integrity and accuracy of the real-time operational status data, this module incorporates a multi-source data fusion unit to integrate three types of data: first, real-time data collected by the equipment's sensors, such as tension values ​​from the yarn tension sensor and rotational speed values ​​from the machine speed sensor; second, yarn status image data captured by industrial cameras, such as yarn fuzziness and potential breakage; and third, historical operational data from the equipment, such as machine speed variation curves and yarn feeding rhythm adjustment records from previous knitting tasks. Through multi-source data fusion algorithms, such as... Figure 6a , Figure 6b , Figure 6c as well as Figure 6d As shown, these data are integrated into unified real-time operational status data, providing a foundation for subsequent deviation analysis.

[0087] Furthermore, the deviation adjustment module generates adjustment instructions based on the deviation between real-time running status data and structured instructions. This module comprises two key units: a hierarchical deviation evaluation unit and a deep reinforcement learning control unit. The hierarchical deviation evaluation unit first categorizes the features in the structured instructions into two classes: key features and secondary features, and then calculates the deviation values ​​for each class of features, such as... Figure 4a and Figure 4b As shown; the deep reinforcement learning control unit, based on these bias characteristics, uses a deep reinforcement learning model, such as Figure 7 The diagram shown illustrates the optimization of the deviation adjustment strategy. This optimization strategy generates more precise adjustment instructions. For example, when the deviation of a key feature is large, the control parameters of the needle bed trajectory are adjusted first; when the deviation of a secondary feature is large, the parameters of the machine speed control are adjusted.

[0088] Furthermore, the equipment control module is the system's execution terminal. Its function is to send the adjustment commands generated by the deviation adjustment module to the knitting equipment, controlling the equipment to adjust its operating parameters. To ensure the validity of the adjustment commands, this module is equipped with a digital twin verification unit, used to input the adjustment commands into the digital twin model of the knitting equipment, such as... Figure 5 As shown, this simulates the effect of the device executing adjustment commands. The digital twin verification unit only sends the adjustment command to the actual device when the simulation results show that the coil density, pattern accuracy, and other indicators of the braided product meet the design requirements; otherwise, it returns to the deviation adjustment module to re-optimize the adjustment strategy.

[0089] Example 10: In order to realize the above-mentioned method of collaborative control of knitting equipment by design software, this example provides a computer device. This device realizes full-process automation of design data parsing, status interaction, deviation adjustment and equipment control through the combination of hardware and software.

[0090] The computer device includes a processor, a memory, and a computer program stored in the memory. The memory stores the computer program that implements the above method, including program code for designing a data parsing module, a status interaction module, a deviation adjustment module, and a device control module. The processor executes the computer program, and the specific execution process is as follows:

[0091] First, the processor calls the program code of the design data parsing module to obtain the knitted sweater design data output by the design software. Next, the program of the multimodal semantic mapping unit is launched to extract multimodal features from the design data, such as image features from the pattern file, text features from the process rule text, and numerical features from the size parameter table. Based on the pre-built knitting process knowledge graph, these features are mapped to the corresponding knitting rules, pattern features, and process parameters to generate structured instructions.

[0092] Furthermore, the processor calls the program code of the status interaction module to send structured instructions to the knitting sweater device and starts the program of the multi-source data fusion unit to receive sensor data, camera image data and historical operation data returned by the device, and integrates these data into real-time operation status data through the multi-source data fusion algorithm.

[0093] Furthermore, the processor calls the program code of the deviation adjustment module, starts the program of the hierarchical deviation evaluation unit, divides the features in the structured instructions into key features and secondary features, and calculates the deviation values ​​of the two types of features; then it starts the program of the deep reinforcement learning control unit, and based on these deviation values, optimizes the adjustment strategy through the deep reinforcement learning model to generate adjustment instructions.

[0094] Furthermore, the processor calls the program code of the device control module to start the program of the digital twin verification unit, inputting the adjustment command into the digital twin model of the knitting sweater equipment to simulate the effect of the equipment executing the adjustment command. When the simulation results show that the coil density, pattern accuracy, and other indicators of the knitted product meet the design requirements, the adjustment command is sent to the actual equipment to control the equipment to adjust the operating parameters; if the simulation results do not meet the requirements, the process returns to the deviation adjustment module to re-optimize the adjustment strategy.

[0095] Furthermore, the processor also executes a data collection and feedback program, collecting operational status data and quality inspection data after the equipment executes adjustment instructions, and feeding this data back to the design software to adjust design parameters and optimize subsequent knitted sweater design data. Through this process, the computer device achieves collaborative control between the design software and the knitted sweater equipment, ensuring the quality and efficiency of the knitted products.

[0096] Although the present invention has been specifically described above with reference to preferred embodiments, it should be understood that the present invention is not limited to the embodiments described above. Various modifications and variations can be made by those skilled in the art without departing from the spirit of the present invention, and such modifications and variations should fall within the scope defined by the appended claims and their equivalents.

Claims

1. A method for designing software to collaboratively control a knitting sweater equipment, characterized by the following steps: include: Obtain the knitted sweater design data output by the design software, perform semantic parsing on the design data to obtain structured instructions containing knitting rules, pattern features and process parameters. The semantic parsing is used to establish a semantic mapping between the design data and the knitting equipment instructions. The structured instructions are sent to the knitting equipment, and the real-time operating status data returned by the knitting equipment is received. Based on the deviation between the real-time operating status data and the structured instructions, adjustment instructions are generated through a collaborative control model, which is used to optimize the adjustment strategy based on the deviation characteristics. The adjustment command is sent to the knitting equipment to control the knitting equipment to adjust its operating parameters to perform the knitting task; The steps for generating the adjustment instruction include: dividing the features in the structured instruction into key features and secondary features, wherein the key features include coil density, jacquard pattern position, and yarn tension threshold, and the secondary features include machine speed and needle bed oscillation frequency; calculating the first deviation between the key features in the real-time operating status data and the corresponding key features in the structured instruction, and the second deviation between the secondary features and the corresponding secondary features in the structured instruction; obtaining the comprehensive deviation based on the weighting coefficients of the first deviation and the second deviation, and generating the adjustment instruction through a collaborative control model; After generating the adjustment command through the collaborative control model, the method further includes: inputting the adjustment command into the digital twin model of the knitting sweater equipment to simulate the operating state and knitting action of the knitting sweater equipment after executing the adjustment command; when the simulated operating state meets the preset equipment safety threshold and the simulated knitting action meets the requirements of the design data, the adjustment command is sent to the knitting sweater equipment. After sending the adjustment command to the knitting equipment, the method further includes: collecting the operating status data and quality inspection data of the knitting equipment after executing the adjustment command, the quality inspection data including loop density inspection data and pattern accuracy inspection data; feeding back the operating status data and quality inspection data to the design software, the design software adjusting the design parameters according to the feedback results to optimize the subsequent knitting design data; The semantic parsing includes: extracting multimodal features from the design data, including pattern texture features of the design image, process description features of the design text, and numerical features of the design parameters; and mapping the multimodal features to corresponding knitting rules, pattern features, and process parameters based on a pre-constructed knitting process knowledge graph, wherein the knowledge graph is used to store the semantic association between design features and knitting equipment instructions.

2. The method for collaborative control of knitting equipment by design software as described in claim 1, characterized in that, The step of receiving the real-time operating status data returned by the knitting equipment includes: receiving sensor data from the knitting equipment, yarn status image data captured by the camera, and historical operating data of the knitting equipment. The sensor data includes machine speed sensor data, yarn tension sensor data, and needle bed position sensor data. The sensor data, image data, and historical operating data are fused to obtain the real-time operating status data, so as to eliminate data noise and supplement missing data.

3. The method for collaborative control of knitting equipment by design software as described in claim 1, characterized in that, Sending the structured instructions to the knitting equipment includes: when multiple knitting equipments are performing knitting tasks simultaneously, obtaining the current task progress and resource usage of each knitting equipment, coordinating the sending order of the structured instructions and the process parameters of each knitting equipment, and avoiding resource conflicts between equipment. The resource usage includes the remaining amount of yarn and the idle status of the needle bed.

4. The method for collaborative control of knitting equipment by design software as described in claim 1, characterized in that, After obtaining the knitted sweater design data output by the design software, the method further includes: monitoring whether the design software has modified the design data; when modification is detected, re-semantically parsing the modified design data to obtain updated structured instructions; and adjusting the previously generated adjustment instructions according to the updated structured instructions to keep the adjustment instructions consistent with the updated structured instructions.

5. A system for collaboratively controlling a knitting sweater equipment using design software, for implementing the method for collaboratively controlling a knitting sweater equipment using design software as described in any one of claims 1-4, characterized in that, include: The design data parsing module is used to acquire knitted sweater design data output by the design software, perform semantic parsing on the design data, and obtain structured instructions containing knitting rules, pattern features and process parameters. The design data parsing module includes a multimodal semantic mapping unit, which is used to extract multimodal features from the design data and map them to corresponding knitting rules, pattern features and process parameters based on a pre-built knitting process knowledge graph. The status interaction module is used to send the structured instructions to the knitting sweater equipment and receive the real-time operating status data returned by the knitting sweater equipment. The status interaction module includes a multi-source data fusion unit, which is used to fuse the sensor data of the knitting sweater equipment, the yarn status image data collected by the camera, and the historical operating data to obtain the real-time operating status data. The deviation adjustment module is used to generate adjustment instructions through a collaborative control model based on the deviation between the real-time running status data and the structured instructions. The deviation adjustment module includes a hierarchical deviation evaluation unit and a deep reinforcement learning control unit. The hierarchical deviation evaluation unit is used to divide the features in the structured instructions into key features and secondary features and calculate the corresponding deviations. The deep reinforcement learning control unit is used to optimize the adjustment strategy based on the deviation features through a deep reinforcement learning model. The equipment control module is used to send the adjustment command to the knitting sweater equipment and control the knitting sweater equipment to adjust the operating parameters to perform the knitting task. The equipment control module includes a digital twin verification unit, which is used to input the adjustment command to the digital twin model of the knitting sweater equipment to simulate the execution effect. When the simulation result meets the requirements, the adjustment command is sent.

6. A computer device, comprising a processor, a memory, and a computer program stored in the memory, characterized in that, When the processor executes the computer program, it implements the method for collaborative control of a knitting sweater equipment by design software as described in any one of claims 1-4.

Images