A Data Processing Method for Optimizing Energy Consumption in Flash Trimming of Injection Molded Products

Abstract

The present invention relates to the technical field of post-processing of injection molded products, and particularly to a data processing method for optimizing energy consumption in trimming the flash of injection molded products. The present invention specifically includes the following steps: synchronously collecting the operation data of each stage, including energy consumption, images, time, and tool models, and extracting the initial flash features according to the image data; calculating the quality scores of each stage respectively according to the operation data; constructing an energy consumption-quality coupling matrix based on the energy consumption and quality scores to locate the stages to be optimized; calculating the theoretical energy consumption and theoretical quality scores according to the initial flash features; comparing and analyzing the theoretical energy consumption and theoretical quality scores with the actual ones, and formulating different optimization strategies for different stages; verifying after implementing the optimization strategies, and storing the verified effective optimization parameters into the strategy knowledge base; The present invention can accurately locate the most serious link of energy consumption waste in the trimming process, point out the direction for energy consumption optimization, and help reduce ineffective energy consumption.

Description

Technical Field

[0001] This invention relates to the field of post-processing technology for injection molded products, and in particular to a data processing method for optimizing energy consumption during the flash trimming process of injection molded products. Background Technology

[0002] Traditional finishing techniques rely heavily on the manual experience of technicians in actual operation. Quality evaluation is largely based on experiential indicators derived from technicians' on-site observation and touch, while energy management focuses on overall energy consumption statistics without separate measurement for each stage. Process adjustments are primarily based on experience accumulated through long-term practice, lacking a systematic mechanism linking theoretical models with actual data. The operational experience of skilled technicians is mostly passed down orally from master to apprentice, with the overall goal of ensuring basic product quality, resulting in relatively fixed production operation procedures. This leads to:

[0003] The overall management is too extensive, making it difficult to balance the relationship between product quality, energy consumption, and production cycle; it is difficult to accurately locate the links with serious energy waste, resulting in a lack of direction for energy consumption optimization and difficulty in reducing ineffective energy consumption; it is impossible to achieve unified and objective quality evaluation, which is prone to imbalance in quality control due to subjective bias, affecting the stability of product quality; it is impossible to identify abnormal working conditions and inefficient operations in production in a timely manner, making it difficult to correct problems in a targeted manner, and process optimization lacks data support, which can only maintain the original level and cannot achieve continuous improvement; the experience of excellent technicians is only passed down orally, which is difficult to solidify and promote throughout the factory, and production is highly dependent on personal experience. Personnel changes are prone to process fluctuations, making it impossible to achieve data-based adaptive optimization and resulting in poor production stability; it is difficult to balance quality, energy consumption, and production cycle, often resulting in excessive energy consumption to ensure quality or sacrificing quality to reduce energy consumption, leading to high overall energy consumption, long production cycle, increased production costs, and difficulty in effectively improving the company's economic benefits.

[0004] Therefore, a data processing method for optimizing energy consumption in injection molding product flash trimming is needed to solve the above problems. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a data processing method for optimizing energy consumption in the trimming of overflow of injection molded products, so as to solve the problems existing in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a data processing method for optimizing energy consumption in edge trimming of injection molded products, characterized by comprising the following steps:

[0007] S1. Divide the trimming process into three stages: rough trimming, fine trimming, and detail trimming. Collect the operation data of each stage simultaneously, including energy consumption, image, time, and tool model, and extract the initial overflow features based on the image data.

[0008] S2. Calculate the quality score for each stage based on the operational data;

[0009] S3. Construct an energy consumption and quality coupling matrix based on energy consumption and quality scores, and analyze the matrix to identify the stages that need optimization.

[0010] S4. For the stages that need optimization, calculate the theoretical energy consumption and theoretical quality score based on the initial overflow characteristics.

[0011] S5. Compare and analyze the theoretical energy consumption and theoretical quality score with the actual energy consumption and actual quality score, and formulate different optimization strategies for different stages.

[0012] S6. After implementing the optimization strategy, verify it and store the verified and effective optimization parameters in the strategy knowledge base.

[0013] The energy consumption quality coupling matrix is ​​constructed as follows: the horizontal dimension of the matrix includes coarse trimming, fine trimming, and detailed trimming; the vertical dimension includes current energy consumption, cumulative energy consumption, current quality score, cumulative quality score, and unit energy consumption quality efficiency; where unit energy consumption quality efficiency is obtained by comparing current quality score with current energy consumption; the dimensions are cross-matched to form the energy consumption quality coupling matrix.

[0014] The theoretical energy consumption and theoretical quality score are calculated as follows: For theoretical energy consumption, based on the initial overflow characteristics and product material properties, a feature mapping relationship is established through a machine learning model using a historical database to predict the theoretical trimming time required for each trimming stage; the theoretical time is converted into theoretical energy consumption using the rated power of the trimming tool as the conversion coefficient; for theoretical quality score, it is preset according to product standards and quality requirements.

[0015] The method of formulating different optimization strategies for different stages is as follows: compare theoretical energy consumption and theoretical quality score with actual energy consumption and actual quality score, analyze whether the energy consumption is lower than the theoretical value and whether the quality score meets the standard, and analyze the reasons for the difference; based on clarifying the root cause of the difference, implement specific optimization strategies for different stages: for the stage where the unit energy consumption quality efficiency in the energy consumption quality coupling matrix is ​​lower than the preset threshold, dynamically adjust the process parameters of the stage; for the detail finishing stage, set a threshold for the final quality, and automatically terminate the finishing operation once the threshold is reached.

[0016] The operational data for each stage includes: rough retouching energy consumption, rough retouching image, rough retouching time, and rough retouching tool model; fine retouching energy consumption, fine retouching image, fine retouching time, and fine retouching tool model; and detail retouching energy consumption, detail retouching image, detail retouching time, and detail retouching tool model.

[0017] The quality score for each stage is calculated as follows: based on the image data collected before and after each stage of trimming, the percentage of overflow area removed in each stage, the surface roughness after each stage of trimming, and the image similarity between each stage and the standard sample are calculated using machine vision algorithms; these are then weighted and summed to obtain the quality score for each stage, with the weights allocated according to the product type.

[0018] The method for identifying the stages requiring optimization is as follows: Analyze the matrix, and the stage with the lowest unit energy consumption quality efficiency is the stage with the lowest energy consumption efficiency and the one that needs the most optimization. Under normal circumstances, the coarse trimming stage has the highest energy consumption efficiency and the fine trimming stage has the lowest energy consumption efficiency. If the energy consumption efficiency of the fine trimming stage is lower than a preset threshold, then the fine trimming stage needs optimization. If the energy consumption efficiency of the coarse trimming stage is higher than the preset threshold, then the coarse trimming stage needs optimization. If the energy consumption efficiency of the fine trimming stage is higher than that of the coarse trimming stage or lower than that of the fine trimming stage, determine whether it exceeds the preset threshold range. If it does, then optimization is required.

[0019] The method for verifying the implementation of the optimization strategy is as follows: compare the optimized detection data with the historical benchmark data before optimization, and evaluate whether the final quality score reaches or exceeds the target threshold, whether the total energy consumption decreases to the set threshold, and whether the unit energy consumption efficiency increases to the preset threshold; when the data proves that the optimization strategy has reached the required threshold in terms of reducing energy consumption, improving efficiency, or improving quality, and has not introduced new quality defects, the strategy is determined to be effective.

[0020] The technical effects and advantages of this invention are as follows:

[0021] 1. This invention achieves energy consumption visualization through phased energy consumption measurement and efficiency matrix, which can accurately locate the most serious energy waste in the maintenance process, point out the direction for energy consumption optimization, and help reduce ineffective energy consumption.

[0022] 2. This invention transforms the ambiguous trimming effect into a quantifiable quality score, unifies quality standards, makes quality evaluation more objective, avoids subjective bias, and provides a clear basis for quality control.

[0023] 3. By comparing theoretical models with actual data, this invention can promptly identify production anomalies and inefficiencies, correct problems, continuously drive the optimization of the finishing process, and ensure efficient production operation.

[0024] 4. The optimization strategy library formed by this invention can solidify the experience of excellent technicians and promote it throughout the factory, realize data-driven adaptive optimization, reduce dependence on human experience, and improve production stability.

[0025] 5. This invention significantly reduces the overall energy consumption of the finishing process, shortens the production cycle, reduces costs, and effectively improves the economic benefits of enterprises while ensuring stable product quality, thus achieving a win-win situation for both quality and efficiency. Attached Figure Description

[0026] Figure 1 This is a flowchart of the method of the present invention.

[0027] Figure 2 This is a schematic diagram of the energy consumption quality coupling matrix of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Reference Figure 1 This invention provides a data processing method for optimizing energy consumption in the trimming of overflow in injection molded products, comprising six steps.

[0030] In a more specific application of the present invention, the trimming process is divided into three stages: rough trimming, fine trimming and detail trimming. The running data of each stage is collected simultaneously and the initial overflow features are extracted.

[0031] The method for collecting operational data at each stage is as follows: For energy consumption data, a smart meter or power sensor is installed on the power supply line of each finishing tool to collect and record the tool's instantaneous power data in real time at a high frequency. The power data is integrated over time using an industrial control computer or data acquisition card to calculate the actual energy consumed in processing a single workpiece at that stage, thus obtaining the energy consumption for rough finishing, fine finishing, and detailed finishing.

[0032] For image data, an industrial camera with a constant light source is installed at a fixed position at each finishing station. The camera's angle of view, focal length, and lighting conditions are ensured to remain consistent throughout the production process. After the robot or fixture positions the workpiece at the finishing station but before the finishing tool begins its movement, the camera is triggered to acquire a "before finishing" image. When the finishing action is completed at this stage but the workpiece has not yet moved out of the station, the camera is triggered again to acquire a "after finishing" image. High-quality image pairs are obtained simultaneously at each stage, providing a foundation for subsequent quality scoring and feature extraction.

[0033] For time data, the start and stop signals of each dressing tool are directly obtained from the PLC of the production line; the complete time span from the start to the stop of the dressing tool is recorded, which is the actual dressing time for that stage; this time data should be synchronized with the energy consumption and image data in terms of timestamps.

[0034] For tool model data, a digital work order is created for the processing path of each workpiece in the production management system. When the system assigns a trimming task to a specific workpiece, the specific tool model used will be automatically recorded and associated with the production data package of that workpiece, thereby realizing the accurate collection and traceability of tool model information.

[0035] The initial overflow feature is extracted by analyzing the initial image data using machine vision algorithms. The acquired raw images are preprocessed, including using Gaussian or median filtering algorithms to remove noise and performing illumination correction to standardize image brightness and contrast, laying the foundation for subsequent accurate analysis.

[0036] By using a trained deep learning segmentation model or an advanced threshold segmentation and edge detection algorithm, the overflow region is accurately separated from the product background, generating a clear binary "overflow mask" image, where white pixels represent the overflow and black pixels represent the product background.

[0037] By using pre-calibrated camera parameters, the pixel area of ​​the overflow region in the mask is converted into the actual area; then, combined with the depth information obtained by the structured light or laser displacement sensor, the average height of the overflow is estimated; the total volume V is calculated by integrating the area of ​​each pixel region with its corresponding height information; the distribution density M is quantified by calculating the ratio of the total number of overflow pixels to the total number of pixels in the entire region of interest.

[0038] The three key feature parameters extracted—total overflow volume V, average thickness D, and distribution density M—will be structured and stored together with the product material code, providing accurate input data for subsequent theoretical model prediction and optimization strategy generation.

[0039] The quality score for each stage is calculated based on the operational data. The calculation method is as follows: based on the image data collected before and after each stage, the percentage of overflow area removed, the surface roughness after each stage, and the image similarity between each stage and the standard sample are calculated using machine vision algorithms. Weights are assigned to the percentage of overflow area removed, the surface roughness after each stage, and the image similarity between each stage and the standard sample according to the product type. The weighted sum is then used to obtain the quality score for each stage.

[0040] An energy consumption and quality coupling matrix is ​​constructed based on energy consumption and quality scores, and the matrix is ​​analyzed to identify stages requiring optimization. (See also...) Figure 2As shown, the energy consumption quality coupling matrix is ​​constructed as follows: the horizontal dimension of the matrix includes coarse trimming, fine trimming and detailed trimming; the vertical dimension includes current energy consumption, cumulative energy consumption, current quality score, cumulative quality score and unit energy consumption quality efficiency; where unit energy consumption quality efficiency is obtained by comparing current quality score with current energy consumption; the dimensions are cross-matched to form the energy consumption quality coupling matrix.

[0041] The method for identifying the stages that need optimization is as follows: Analyze the matrix, and the stage with the lowest energy consumption efficiency is the stage with the lowest energy consumption efficiency and the one that needs the most optimization. The roughing stage has the highest cost-effectiveness and energy consumption efficiency because it removes a large number of easily handled overflow edges. The fine-tuning stage is next. When it comes to the detail finishing stage, its main task is to deal with minor flaws and perform surface polishing. In order to pursue the last few points of quality improvement, it often requires a disproportionately large amount of energy. Therefore, its energy consumption efficiency is the lowest under normal circumstances.

[0042] Based on historical production data, technical feasibility, and economic cost-benefit analysis, minimum efficiency boundaries, maximum efficiency boundaries, and fine-tuning efficiency boundaries are comprehensively set. If the energy consumption efficiency in the fine-tuning stage is lower than the preset minimum efficiency boundary, the fine-tuning stage needs to be optimized; if the energy consumption efficiency in the rough-tuning stage is higher than the maximum efficiency boundary, the rough-tuning stage needs to be optimized; if the energy consumption efficiency in the fine-tuning stage is higher than that in the rough-tuning stage or lower than that in the fine-tuning stage, it is determined whether it exceeds the preset fine-tuning efficiency boundary range, and if it does, it needs to be optimized.

[0043] For the stages requiring optimization, theoretical energy consumption and theoretical quality scores are calculated based on the initial overflow characteristics. The calculation method for theoretical energy consumption and theoretical quality scores is as follows: For theoretical energy consumption, the system combines the extracted total overflow volume, average thickness, distribution density, and material attribute code of the product into a feature vector. This vector is then input into a pre-trained machine learning model to establish a complex nonlinear mapping relationship from "overflow characteristics" to "ideal trimming time." By learning and imitating the process parameters of numerous successful cases in the historical database, the theoretical trimming time required for each trimming stage under optimal process parameters is predicted under the current overflow condition and material. After obtaining the theoretical time, the rated power of the tools used in that trimming stage is multiplied by the predicted theoretical trimming time to obtain the theoretical energy value that should be consumed to complete that stage.

[0044] The theoretical quality score is predefined and stored in the system's standard process library before production begins. It represents the final quality level that the workpiece should achieve under ideal conditions after a complete standard finishing process. The theoretical quality score is a given target benchmark that actual production needs to strive to achieve.

[0045] By comparing and analyzing theoretical energy consumption and theoretical quality scores with actual energy consumption and actual quality scores, different optimization strategies are formulated for different stages. The actual total energy consumption and final quality score generated during the processing of each product are compared one by one with the theoretical energy consumption and target quality score predicted based on its initial overflow characteristics. If the actual energy consumption is higher than the theoretical value but the quality score fails to meet the standard, it may indicate poor equipment condition, unreasonable process parameter settings, or severe tool wear. If the actual energy consumption is close to the theoretical value but the quality score is lower than the target, it may indicate that the process parameters on which the prediction model is based are not optimal for the product's material or overflow characteristics, resulting in over-processing or under-processing. Through this model analysis, the root cause of the discrepancy can be accurately identified—whether it is a problem with the equipment, tools, parameters, or the model itself.

[0046] Based on a clear understanding of the root causes of the differences, the system can implement customized optimization strategies for the characteristics of different products. For stages exhibiting significantly low unit energy efficiency in the energy consumption-quality coupling matrix, the core strategy is to dynamically adjust their process parameters. The system will automatically generate parameter optimization schemes based on the results of the difference analysis. For example, if the unit energy efficiency in the finishing stage is lower than a preset threshold, it is recommended to appropriately reduce the tool speed or shorten the action time in this stage to reduce energy consumption while maintaining quality; conversely, if the efficiency in the roughing stage is insufficient, it will recommend increasing the feed speed or replacing it with a more aggressive grinding head to improve the material removal efficiency in this stage, thereby reducing the burden on subsequent stages and optimizing energy consumption allocation overall.

[0047] For the final finishing stage of the finishing process, which is usually the least energy-efficient and prone to energy waste due to the pursuit of ultimate quality, the system sets a scientific quality threshold for different types of products. This threshold is pre-written into the product's processing instructions. During production, the system monitors the calculation results of the quality score in real time. Once the preset threshold is reached or exceeded, the system immediately sends an instruction to the PLC to automatically terminate the finishing operation, thereby effectively avoiding unnecessary energy consumption and achieving the optimal balance between quality and energy consumption.

[0048] All proven and effective optimization strategies and their corresponding product characteristics will be summarized and stored in the strategy knowledge base, enabling the system to continuously learn and automatically recommend the best production plan that has been tested in practice when faced with new products.

[0049] Through the above description of the embodiments, those skilled in the art can clearly understand that the various embodiments of this application can be implemented by means of software or software combined with necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware functions; based on this understanding, the technical solution of this application, in essence or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to cause a computer device, such as including but not limited to a personal computer, server, or network device, to execute all or part of the steps of the method described in any embodiment of this application.

[0050] The foregoing describes exemplary embodiments of this application. It should be understood that the above exemplary embodiments are not restrictive but illustrative, and the scope of protection of this application is not limited thereto. It should be understood that those skilled in the art can make modifications and variations to the embodiments of this application without departing from the spirit and scope of this application, and such modifications and variations should be within the scope of protection of this application.

Claims

1. A method for processing energy consumption optimization data of an injection molding product flash trimming, characterized in that, Specifically comprising the following steps: S1, dividing the trimming process into three stages of rough trimming, fine trimming and detail trimming, synchronously collecting operation data of each stage, including energy consumption, image, time and tool type, and extracting initial overflow edge features according to image data; S2, calculating quality scores of each stage according to operation data; S3, constructing an energy consumption-quality coupling matrix according to energy consumption and quality scores, and analyzing the matrix to locate the stage needing optimization; S4, calculating theoretical energy consumption and theoretical quality score according to initial overflow edge features for the stage needing optimization; S5, comparing and analyzing theoretical energy consumption and theoretical quality score with actual energy consumption and actual quality score, and formulating different optimization strategies for different stages; S6, verifying the optimization strategies after implementation, and storing the optimization parameters verified as effective into a strategy knowledge base; The construction method of the energy consumption-quality coupling matrix is that: the horizontal dimension of the matrix includes rough trimming, fine trimming and detail trimming; and the vertical dimension includes current period energy consumption, cumulative energy consumption, current period quality score, cumulative quality score and unit energy consumption-quality efficiency; The unit energy consumption-quality efficiency is obtained by dividing the current period quality score by the current period energy consumption; and each dimension is cross-matched to form the energy consumption-quality coupling matrix; The calculation method of the theoretical energy consumption and the theoretical quality score is that: for the theoretical energy consumption, the feature mapping relationship is established through a machine learning model in a historical database according to initial overflow edge features combined with product material properties, to predict the theoretical trimming time required in each trimming stage; the rated power of the trimming tool is taken as a conversion coefficient to convert the theoretical time into theoretical energy consumption; and for the theoretical quality score, it is pre-set according to product standards and quality requirements; The method of formulating different optimization strategies for different stages is that: the theoretical energy consumption and the theoretical quality score are compared with the actual energy consumption and the actual quality score, whether the energy consumption is lower than the theoretical value and whether the quality score meets the standard are analyzed, and the reasons for the differences are analyzed; on the basis of identifying the root causes of the differences, specific optimization strategies are implemented for different stages: for the stage with unit energy consumption-quality efficiency lower than the pre-set threshold in the energy consumption-quality coupling matrix, the process parameters of the stage are dynamically adjusted; for the detail trimming stage, a threshold is set for the final quality, and once the threshold is reached, the trimming operation is automatically terminated.

2. The energy consumption optimization data processing method for trimming overflow edge of injection molding product according to claim 1, characterized in that: The operation data of each stage includes: rough trimming energy consumption, rough trimming image, rough trimming time and rough trimming tool type; fine trimming energy consumption, fine trimming image, fine trimming time and fine trimming tool type; and detail trimming energy consumption, detail trimming image, detail trimming time and detail trimming tool type.

3. The energy consumption optimization data processing method for trimming the overflow edge of an injection molding product according to claim 1, characterized in that: The calculation method of the quality scores of each stage is that: according to the collected image data before and after trimming in each stage, the removed overflow edge area proportion in each stage, the roughness of the surface after trimming in each stage and the image similarity of each stage with a standard sample are calculated based on a machine vision algorithm; and the quality scores of each stage are obtained by weighted summation, and the weights are allocated according to different product types.

4. The energy consumption optimization data processing method for trimming the overflow edge of an injection molding product according to claim 1, characterized in that: The positioning of the stage to be optimized is performed by analyzing the matrix, wherein the stage with the lowest unit energy consumption quality efficiency is the stage with the lowest energy consumption efficiency and the most in need of optimization; under normal circumstances, the energy consumption efficiency of the rough adjustment stage is the highest and the energy consumption efficiency of the detail adjustment stage is the lowest, if the energy consumption efficiency of the detail adjustment stage is lower than a preset threshold, the detail adjustment stage needs to be optimized; if the energy consumption efficiency of the rough adjustment stage is higher than a preset threshold, the rough adjustment stage needs to be optimized; if the energy consumption efficiency of the fine adjustment stage is higher than that of the rough adjustment stage or lower than that of the detail adjustment stage, it is judged whether it exceeds a preset threshold range, if it exceeds, it needs to be optimized.

5. The energy consumption optimization data processing method for trimming the overflow edge of an injection molding product according to claim 1, characterized in that: The verification of the optimization strategy is performed by comparing the optimized detection data with the historical baseline data before optimization, evaluating whether the final quality score reaches or exceeds the target threshold, whether the total energy consumption decreases to the set threshold and whether the unit energy consumption efficiency improves to the preset threshold; when the data proves that the optimization strategy has reached the required threshold in terms of reducing energy consumption, improving efficiency or improving quality, and has not introduced new quality defects, the strategy is determined to be effective.

Images