Production process parameter dynamic adjustment method and device based on AI visual detection

Abstract

The application discloses a production process parameter dynamic adjustment method and device based on AI vision detection. The server method comprises the following steps: acquiring a production video stream collected by a 5G explosion-proof camera pre-deployed in a production environment; inputting the production video stream into a pre-trained AI vision detection model to output a plurality of appearance characteristic parameters corresponding to the production video stream and used for reflecting the appearance of a chemical product; analyzing production process parameters in a production device of the production environment which do not meet preset product production conditions according to the plurality of appearance characteristic parameters, wherein the parameters are obtained based on a preset process knowledge guide map, and the preset process knowledge guide map is used for recording a correlation relationship between appearance process marks and a production process parameter set of the production device in the production environment; and adjusting the parameter value of the production process parameters which do not meet the preset product production conditions. Therefore, the application can meet the high requirement for quality control and reduce the product error probability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and in particular to a production process parameter dynamic adjustment method and device based on AI visual detection. BACKGROUND

[0002] In the production process of chemical products, especially in the field of fine chemicals, the appearance characteristics (such as color, shape, surface texture, etc.) of the products are important indicators for measuring their quality and performance. In actual production, these appearance characteristics need to be monitored in real time on the production line in order to timely adjust the production process parameters and ensure the consistency and stability of product quality.

[0003] In related technologies, appearance detection in the production of chemical products mainly relies on traditional detection equipment. Traditional detection equipment can usually only detect simple physical parameters of products, and has low detection accuracy for deep appearance characteristics such as shape and texture, which is difficult to meet the high requirements of quality control. At the same time, traditional detection equipment cannot adjust the parameters of production equipment in real time according to the detection results, which increases the probability of product errors. SUMMARY

[0004] The embodiments of the present application provide a production process parameter dynamic adjustment method and device based on AI visual detection. In order to have a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This part is not a general review, nor is it intended to determine the key / important components or delineate the protection scope of these embodiments. Its only purpose is to present some concepts in a simple form as a prelude to the detailed description that follows.

[0005] In a first aspect, the embodiments of the present application provide a production process parameter dynamic adjustment method based on AI visual detection, applied to a server, the method comprising:

[0006] acquiring a production video stream collected by a 5G explosion-proof camera pre-deployed in a production environment according to a preset period, the production video stream being used to represent internal information of a reaction kettle, material feeding information, and appearance information of a chemical product;

[0007] inputting the production video stream into a pre-trained AI visual detection model, and outputting a plurality of appearance characteristic parameters corresponding to the production video stream and reflecting the appearance of the chemical product;

[0008] analyzing production process parameters in the production equipment of the production environment that do not meet preset product production conditions based on the plurality of appearance characteristic parameters, the production process parameters that do not meet the preset product production conditions being obtained based on a preset process knowledge oriented graph, and the preset process knowledge oriented graph being used to record the correlation between appearance process marks and a production process parameter set of the production equipment in the production environment;

[0009] Adjust the parameter value of the production process parameter that does not meet the preset product production condition according to a preset step size.

[0010] In a second aspect, the embodiments of the present application provide a production process parameter dynamic adjustment device based on AI visual detection, which comprises:

[0011] A production video stream acquisition module is configured to acquire a production video stream collected by a 5G explosion-proof camera pre-deployed in a production environment within a preset time period, and the production video stream is used to represent internal information of a reaction kettle, material feeding information, and appearance information of a chemical product.

[0012] An appearance feature parameter output module is configured to input the production video stream into a pre-trained AI visual detection model and output a plurality of appearance feature parameters corresponding to the production video stream and used to reflect the appearance of the chemical product.

[0013] A production process parameter analysis module is configured to analyze production process parameters in a production device in a production environment that do not meet a preset product production condition based on the plurality of appearance feature parameters, wherein the production process parameters that do not meet the preset product production condition are obtained based on a preset process knowledge guide map, and the preset process knowledge guide map is used to record a correlation relationship between appearance process marks and a production process parameter set of the production device in the production environment.

[0014] A parameter value adjustment module is configured to adjust the parameter value of the production process parameter that does not meet the preset product production condition according to a preset step size.

[0015] The technical scheme provided by the embodiments of the present application can include the following beneficial effects:

[0016] In the embodiments of the present application, on the one hand, the pre-trained AI visual detection model can extract a plurality of appearance feature parameters including physical appearance parameters (such as color and shape) and non-physical appearance parameters (such as texture and luster) from the production video stream collected by the 5G explosion-proof camera. These parameters can comprehensively reflect the appearance information of the chemical product and provide more accurate and detailed data support for quality control. Through analysis of these appearance feature parameters, any slight changes in the appearance of the product can be found in time, and high requirements for quality control can be met. On the other hand, by acquiring a plurality of appearance feature parameters reflecting the appearance of the chemical product in real time, these parameters are associated with the production process parameter set recorded in the preset process knowledge guide map to trace the production process parameters that do not meet the preset product production condition. The system can automatically adjust the values of these parameters to adapt to the current production conditions. The dynamic online adjustment mechanism can timely correct the deviation in the production process and reduce the product quality problems caused by improper parameter setting, thereby significantly reducing the product error probability.

[0017] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application, as claimed. BRIEF DESCRIPTION OF DRAWINGS

[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and serve to explain the principles of the application, in which, like reference numerals designate corresponding parts throughout the several views.

[0019] Figure 1 is a method flow diagram of a production process parameter dynamic adjustment method based on AI visual inspection provided by an embodiment of the application;

[0020] Figure 2 is an image diagram of an image frame collected in an actual scene of a 5G explosion-proof camera provided by an embodiment of the application;

[0021] Figure 3 is a scene diagram of an application scene provided by an embodiment of the application;

[0022] Figure 4 is a model architecture diagram of a pre-trained AI visual inspection model provided by an embodiment of the application;

[0023] Figure 5A is a parameter representation diagram of a physical appearance parameter provided by an embodiment of the application;

[0024] Figure 5B is a parameter representation diagram of a non-physical appearance parameter provided by an embodiment of the application;

[0025] Figure 6 is a knowledge-oriented diagram of a root node and multiple child nodes provided by an embodiment of the application;

[0026] Figure 7 is a diagram of part of the content in a preset process knowledge-oriented diagram provided by an embodiment of the application;

[0027] Figure 8 is a flow diagram of a model training method of an action parameter analysis model provided by an embodiment of the application;

[0028] Figure 9 is a structural diagram of a production process parameter dynamic adjustment device based on AI visual inspection provided by an embodiment of the application;

[0029] Figure 10 is a structural diagram of an electronic device provided by an embodiment of the application. DETAILED DESCRIPTION

[0030] The following description and drawings are illustrative of specific embodiments of the application and are not intended to limit the scope of the application, as claimed.

[0031] It should be noted that the described embodiments are merely some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0032] The following description refers to the accompanying drawings. Unless otherwise indicated, same or similar elements in different drawings are denoted by same or similar reference numerals. The implementations described in the following exemplary embodiments are not meant to represent all implementations consistent with the present application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the present application as detailed in the appended claims.

[0033] In the description of the present application, it should be understood that the terms "first", "second" and the like are used only for descriptive purposes and are not to be construed as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances. In addition, in the description of the present application, unless otherwise specified, "multiple" means two or more. The association between the associated objects is described by "and / or", which means that there can be three relationships, for example, A and / or B can represent three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after it.

[0034] At present, appearance detection in chemical product production mainly relies on traditional detection equipment.

[0035] The inventors realized that traditional detection equipment can usually only detect simple physical parameters of products, and the detection accuracy of morphological and textural features is low, which is difficult to meet the high requirements of quality control. At the same time, the traditional detection equipment cannot adjust the parameters of the production equipment in real time according to the detection results, which increases the probability of product errors.

[0036] To solve the above problems, the application provides a production process parameter dynamic adjustment method and device based on AI visual detection to solve the problems existing in the above related technical problems. In the embodiments of the application, on the one hand, the pre-trained AI visual detection model can extract a plurality of appearance characteristic parameters including physical appearance parameters (such as color, shape) and non-physical appearance parameters (such as texture, luster) from the production video stream collected by the 5G explosion-proof camera. These parameters can comprehensively reflect the appearance information of the chemical product, and provide more accurate and detailed data support for quality control. Through the analysis of these appearance characteristic parameters, any slight change in the appearance of the product can be found in time, and the high requirements for quality control can be met. On the other hand, by obtaining a plurality of appearance characteristic parameters reflecting the appearance of the chemical product in real time, the parameters are associated with a production process parameter set recorded in a preset process knowledge guide chart, the non-compliant production process parameters are traced, and the system can automatically adjust the values of the parameters to adapt to the current production conditions. The dynamic online adjustment mechanism can correct the deviation in the production process in time, reduce the product quality problems caused by improper parameter setting, thereby significantly reducing the product error probability. The following will be described in detail by exemplary embodiments.

[0037] The following will be described in detail by exemplary embodiments. Figure 1 - The following will be described in detail by exemplary embodiments. Figure 8 The production process parameter dynamic adjustment method based on AI visual detection provided by the embodiments of the application will be described in detail. The method can be realized by relying on a computer program and can be run on an AI visual detection-based production process parameter dynamic adjustment device based on the von Neumann system. The computer program can be integrated in an application or can be run as an independent tool application.

[0038] Please refer to Figure 1 A flowchart of a production process parameter dynamic adjustment method based on AI visual detection provided by the embodiments of the application is provided, which is applied to a server. As shown in the figure, Figure 1 The method provided by the embodiments of the application includes the following steps:

[0039] S101, a production video stream collected by a 5G explosion-proof camera pre-deployed in a production environment is obtained according to a preset period. The production video stream is used to represent internal information of a reaction kettle, material feeding information, and appearance information of a chemical product;

[0040] The preset period is a predetermined time interval for periodically performing production process parameter dynamic adjustment. In this application, it is the time interval for periodically obtaining video streams from the camera. The production video stream is real-time captured and continuously transmitted video data from the camera, which is used to monitor and analyze various information in the production process. The reaction kettle is a commonly used container in chemical production, which is used for chemical reactions. The material feeding information is the type, quantity and time of the raw materials fed into the reaction kettle during the production process. The appearance information of the chemical product refers to the physical appearance characteristics of the chemical product, such as color, shape, etc. The 5G explosion-proof camera is an image acquisition device that can obtain clear video streams in harsh production environments. In the face of complex environments such as high temperature and high pressure in the reaction kettle, there are certain difficulties in online real-time quality visualization detection and analysis of products. By arranging 5G explosion-proof cameras in the reaction kettle, the camera has functions such as custom cooling back blowing, circulating cooling water, supporting variable focal mode, explosion-proof, etc. The image frames collected by the camera in the actual production environment are shown in, for example Figure 2

[0041] In some embodiments of the present application, 5G explosion-proof cameras are deployed at key positions in the production environment, such as near the reaction kettle, the material feeding port, etc. The camera automatically starts and begins to collect production video streams. The server obtains the production video streams collected by the 5G explosion-proof cameras pre-deployed in the production environment according to the preset period. The video stream is used to represent the internal information of the reaction kettle, the material feeding information, and the appearance information of the chemical product.

[0042] For example Figure 3 As shown in, the 5G explosion-proof camera can collect production video streams that can represent the internal information of the reaction kettle, the material feeding information, and the appearance information of the chemical product according to the preset period, and send the collected production video streams to the server through the 5G network.

[0043] S102, input the production video stream into the pre-trained AI visual detection model, and output a plurality of appearance characteristic parameters corresponding to the production video stream for reflecting the appearance of the chemical product;

[0044] The pre-trained AI visual detection model is a mathematical model that uses a large amount of labeled data to train the model so that it can recognize and understand the features in the video. The plurality of appearance characteristic parameters includes physical appearance parameters and non-physical appearance parameters. The physical appearance parameters are directly observable and measurable, such as color, shape, etc. The non-physical appearance parameters are not directly observable and measurable, such as complex texture, luster, transparency, etc.

[0045] For example Figure 4 ​As shown, the pre-trained AI visual inspection model includes a basic appearance feature extraction module and a deep appearance feature extraction module; the basic appearance feature extraction module is used to extract physical appearance parameters of the chemical product in the production video stream, and the deep appearance feature extraction module is used to extract non-physical appearance parameters of the chemical product in the production video stream.

[0046] In some embodiments of the present application, the production video stream is input into the pre-trained AI visual inspection model, and the specific process of outputting the production video stream corresponding to the multiple appearance feature parameters reflecting the appearance of the chemical product includes: using the basic appearance feature extraction module to extract the physical appearance parameters of the chemical product in the production video stream to obtain the original appearance feature parameters; using the deep appearance feature extraction module to extract the non-physical appearance parameters of the chemical product in the production video stream to obtain the deep appearance feature parameters; and taking the original appearance feature parameters and the deep appearance feature parameters as the multiple appearance feature parameters corresponding to the production video stream and reflecting the appearance of the chemical product.

[0047] Among them, the basic appearance feature extraction module is used to extract the physical appearance attributes of the chemical product that can be directly observed from the production video stream, which is directly observable and measurable, such as color, shape, etc. The deep appearance feature extraction module is used to extract the physical appearance attributes of the chemical product that cannot be directly observed from the production video stream, which cannot be directly observed and measured, such as the complex texture, luster, transparency, etc. of the product.

[0048] In some embodiments of the present application, the basic appearance feature extraction module of the AI model analyzes the video stream to extract the physical appearance parameters of the product, such as color, shape, etc., to obtain the original appearance feature parameters. The deep appearance feature extraction module of the AI model further analyzes the video stream to extract the non-physical appearance parameters of the product, such as texture, luster, etc., to obtain the deep appearance feature parameters. The original appearance feature parameters and the deep appearance feature parameters are integrated to form a set of multiple appearance feature parameters that comprehensively reflect the appearance of the chemical product.

[0049] Among them, the data table of the physical appearance parameters is, for example Figure 5A As shown, the data table of the non-physical appearance parameters is, for example Figure 5B As shown.

[0050] In some embodiments of the present application, the specific process of generating the pre-trained AI visual inspection model includes: obtaining sample production video streams of the production environment within a preset period; from the sample production video streams, obtaining each sample key frame reflecting the key visual information of the reaction mid-stage, the material feeding moment, and the appearance of the chemical product of the reaction kettle; for each sample key frame, labeling the physical appearance parameters about color and shape and the non-physical appearance parameters about texture and luster to obtain model training samples; using an appearance parameter recognition algorithm to construct a basic appearance feature extraction module; using a neural network algorithm to construct a deep appearance feature extraction module; according to the basic appearance feature extraction module and the deep appearance feature extraction module, building an AI visual inspection model; according to the model training samples, machine learning the AI visual inspection model to obtain the pre-trained AI visual inspection model.

[0051] Specifically, the specific process of machine learning the AI visual inspection model according to the model training samples to obtain the pre-trained AI visual inspection model includes: inputting the model training samples into the AI visual inspection model to output a model loss value; in the case that the model loss value reaches a minimum, generating the pre-trained AI visual inspection model; or in the case that the model loss value does not reach the minimum, updating the model parameters of the AI visual inspection model and continuing to perform the step of inputting the model training samples into the AI visual inspection model until the model loss value reaches the minimum.

[0052] S103, according to the plurality of appearance feature parameters, analyzing the production process parameters in the production equipment of the production environment that do not meet the preset product production conditions; the production process parameters that do not meet the preset product production conditions are obtained based on the preset process knowledge oriented graph, and the preset process knowledge oriented graph is used to record the correlation relationship between the appearance process marks and the production process parameter set of the production equipment in the production environment;

[0053] Wherein, the appearance feature parameter refers to a parameter extracted from the production video stream for reflecting the appearance characteristics of the chemical product, including physical and non-physical appearance parameters. The production environment is the manufacturing place of the chemical product, including all equipment, tools and conditions. The production equipment is the machine and device used for manufacturing the chemical product in the production environment. The preset product production condition is the production condition and standard defined in advance to ensure product quality. The preset process knowledge oriented graph is a chart used to record and analyze the correlation relationship between the production process parameters and the product appearance characteristics. The appearance process mark refers to the label or identification of the product appearance characteristics related to the production process parameters.

[0054] Wherein, the plurality of appearance feature parameters includes the appearance process mark and the appearance feature quantization value. The appearance feature quantization value is the specific numerical value of the appearance feature parameter, such as the brightness of color.

[0055] In some embodiments of the present application, the specific process of analyzing production process parameters that do not meet the preset product production conditions in the production equipment of the production environment according to the plurality of appearance feature parameters includes: obtaining the corresponding appearance feature quantization threshold range from the mapping relationship between the appearance process marks and the appearance feature quantization threshold range pre-established; comparing the appearance feature quantization value with the appearance feature quantization threshold range to determine the appearance process marks corresponding to the appearance feature quantization values not in the appearance feature quantization threshold range as a plurality of abnormal appearance process marks; performing production process parameter tracing on each abnormal appearance process mark from the preset process knowledge guide map to obtain a plurality of candidate production process parameters; and comparing the current value of each candidate production process parameter with its preset standard range to determine the production process parameters that do not meet the preset product generation conditions in the production equipment of the production environment.

[0056] Wherein, the appearance feature quantization value range refers to the acceptable range of appearance feature quantization values for judging product quality. The abnormal appearance process mark is an appearance process mark that exceeds the quantization value range and may affect product quality. Production process parameter tracing is the process of finding production process parameters related to abnormal appearance process marks from the preset process knowledge guide map. The candidate production process parameter is a production process parameter that needs to be adjusted to correct problems in the production process. The current value is the actual measured value of the production process parameter in the production process. The preset standard range is the ideal range that the production process parameter should maintain to ensure product quality.

[0057] In some embodiments of the present application, the specific process of obtaining a plurality of candidate production process parameters by tracing production process parameters from the preset process knowledge guide map for each abnormal appearance process mark includes: taking each abnormal appearance process mark as a tracing condition, traversing the preset process knowledge guide map to find candidate root nodes that meet the tracing condition; obtaining all child nodes on the candidate root nodes; taking the union set of all child nodes to obtain a plurality of candidate production process parameters.

[0058] In a possible implementation, the production video stream from the 5G explosion-proof camera is processed using an AI visual detection model, and the extracted appearance feature parameters include, for example, color uniformity 85 points, glossiness 45 points, and surface defect quantity 6. The mapping relationship between the pre-established appearance process label and the appearance feature quantization threshold range is (color uniformity: 80-100 points is qualified, 60-79 points is a warning, and less than 60 points is abnormal; glossiness: 70-100 points is qualified, 40-69 points is a warning, and less than 40 points is abnormal; and surface defect quantity: 0-2 is qualified, 3-5 is a warning, and more than 5 is abnormal), and at this time, the color uniformity 85 points is marked as abnormal, and the surface defect quantity 6 is marked as abnormal. The preset process knowledge guide chart shows that the color uniformity abnormality is related to the "stirring speed" and the "mixing time". At this time, the "stirring speed" and the "mixing time" are candidate production process parameters.

[0059] In the embodiment of the present application, the specific process of generating the preset process knowledge guide chart includes: obtaining historical production video streams of a production environment in a preset period; synchronizing each key frame in the historical production video streams with a first production process parameter set corresponding to the time for reflecting the reaction medium stage of the reaction kettle, the material feeding moment, and the key visual information reflecting the appearance of the chemical product; inputting each key frame into a pre-trained AI visual detection model to separate a plurality of first appearance feature quantization values; mining the correlation relationship between each first appearance feature quantization value and the first production process parameter set; and establishing the preset process knowledge guide chart according to the correlation relationship.

[0060] The reaction medium stage of the reaction kettle is the intermediate stage of the state of the material in the reaction kettle during the chemical reaction process. The material feeding moment is the moment when the raw material is fed into the reaction kettle or other production equipment during the production process. The appearance of the chemical product is the external characteristic of the final product, such as color, shape, and texture. The first production process parameter set is the initial production process parameter directly related to the production process, such as temperature, pressure, and stirring speed. The first appearance feature quantization value is the quantization data related to the appearance of the product extracted from the video, such as color saturation and shape size. The correlation relationship is the logical connection between the first appearance feature quantization value and the production process parameter.

[0061] In the embodiment of the present application, through the preset process knowledge guide chart, the production process parameters related to the quality of the appearance of the product can be quickly identified and adjusted, thereby improving the production efficiency.

[0062] In some embodiments of the present application, the specific process of mining the correlation between each first appearance feature quantitative value and the first set of production process parameters includes: analyzing the linear relationship coefficient and the nonlinear monotonic relationship coefficient between each first appearance feature quantitative value and each first production process parameter in the set; screening out the first production process parameter whose absolute value of the linear relationship coefficient is greater than a first preset threshold and whose nonlinear monotonic relationship coefficient is less than a second preset threshold as the production process parameter that has a correlation with each first appearance feature quantitative value, wherein the preset multiple value of the second preset threshold is equal to the first preset threshold; and binding the production process parameter that has a correlation with each first appearance feature quantitative value and each first appearance feature quantitative value to obtain the correlation between each first appearance feature quantitative value and the first set of production process parameters.

[0063] For example, the linear relationship coefficient can be a Pearson correlation coefficient, or a Spearman rank correlation coefficient. The first preset threshold is preferably 0.5, and the second preset threshold is preferably 0.05.

[0064] The correlation includes the production process parameter that has a correlation with each first appearance feature quantitative value.

[0065] In some embodiments of the present application, the specific process of establishing the preset process knowledge directed graph according to the correlation includes: assigning a first appearance process label to each first appearance feature quantitative value; taking the first appearance process label assigned to each first appearance feature quantitative value as a root node, and taking the production process parameter that has a correlation with each first appearance feature quantitative value as a plurality of child nodes; establishing a knowledge directed graph of the root node and the plurality of child nodes; traversing all root nodes to identify a shared child node as a nested node from the knowledge directed graph; and nesting and linking the established knowledge directed graphs based on the nested node to obtain the preset process knowledge directed graph.

[0066] The knowledge directed graph of the root node and the plurality of child nodes is shown in FIG. 8, for example. Figure 6 Some content in the preset process knowledge directed graph is shown in FIG. 9, for example. Figure 7

[0067] S104, adjusting the parameter value of the production process parameter that does not meet the preset product production condition according to the preset step size.

[0068] ​In some embodiments of the present application, in the preset step size, the color uniformity improvement step size can be: 5 units / adjustment. The glossiness improvement step size can be: 5 units / adjustment. For example, the current quantization value is 75, which is lower than the preset qualified value 80. Calculate the adjustment step size: (80-75) / 5=1. Adjust the stirring speed: 300 rpm + 1x5=305 rpm.

[0069] In an embodiment of the present application, on the one hand, the pre-trained AI visual detection model can extract a plurality of appearance characteristic parameters including physical appearance parameters (such as color, shape) and non-physical appearance parameters (such as texture, gloss) from the production video stream collected by the 5G explosion-proof camera. These parameters can comprehensively reflect the appearance information of the chemical product and provide more accurate and detailed data support for quality control. Through the analysis of these appearance characteristic parameters, any slight changes in the appearance of the product can be found in time, which can meet the high requirements of quality control. On the other hand, by obtaining a plurality of appearance characteristic parameters reflecting the appearance of the chemical product in real time, these parameters are associated with a set of production process parameters recorded in the preset process knowledge guide map, to trace the non-compliant production process parameters, the system can automatically adjust the values of these parameters to adapt to the current production conditions. The dynamic online adjustment mechanism can timely correct the deviation in the production process and reduce the product quality problems caused by improper parameter setting, thereby significantly reducing the product error probability.

[0070] Please refer to Figure 8 A flowchart of a model training method of an AI visual detection model is provided for an embodiment of the present application. As Figure 8 shown, the method of the embodiment of the present application can include the following steps:

[0071] S201, obtaining a sample production video stream of a production environment in a preset period;

[0072] In some embodiments of the present application, an explosion-proof 4K industrial camera (such as Basler ace acA2000-50gc) is used to collect the whole process video of the reaction kettle from feeding to discharging at 30fps for 7 days (covering different batches and working conditions).

[0073] S202, obtaining each sample key frame for reflecting the key visual information of the reaction kettle reflecting the middle stage, the material feeding moment and the appearance of the chemical product from the sample production video stream;

[0074] In some embodiments of the present application, the feeding moment can be the second second after the material enters the reaction kettle (triggering the pressure sensor signal to synchronize the interception), the reaction middle stage can be 5 frames continuously extracted when the temperature reaches the set value (150℃±5℃), and the discharging stage can be the surface flow state video (intercepting the last 3 seconds) when the product flows out.

[0075] S203, for each sample key frame, label the physical appearance parameters about color and shape and the non-physical appearance parameters about texture and gloss, to obtain a model training sample;

[0076] In some embodiments of the application, the OpenCV color detection script is used to take the RGB mean (Lab color space) of the center area of the reaction product as the physical appearance parameter of color, and the LabelMe polygon labeling method is used to label the bubble number / diameter (unit: mm) as the physical appearance parameter of shape. Experts determine the texture classification according to the ASTM D7869 standard (0-5 levels: smooth to severe cracking). The reflectivity is measured by the synchronous calibration method of the gloss meter GL-200, and the non-physical appearance parameter of gloss is obtained.

[0077] S204, an appearance parameter recognition algorithm is used to construct a basic appearance feature extraction module;

[0078] Among them, OpenCV+Pytorch can realize the construction of the basic appearance feature extraction module.

[0079] S205, a neural network algorithm is used to construct a deep appearance feature extraction module;

[0080] Among them, the deep appearance feature extraction module can use EfficientNet-B3 suitable for edge computing as the backbone network.

[0081] S206, according to the basic appearance feature extraction module and the deep appearance feature extraction module, an AI visual detection model is built;

[0082] In some embodiments of the application, the basic appearance feature extraction module is integrated into the deep appearance feature extraction module, and an AI visual detection model can be obtained.

[0083] S207, according to the model training sample, machine learning is performed on the AI visual detection model to obtain a pre-trained AI visual detection model.

[0084] In the embodiments of the present application, on the one hand, the pre-trained AI visual detection model can extract a plurality of appearance characteristic parameters including physical appearance parameters (such as color, shape) and non-physical appearance parameters (such as texture, luster) from the production video stream collected by the 5G explosion-proof camera. These parameters can comprehensively reflect the appearance information of the chemical product and provide more accurate and detailed data support for quality control. Through the analysis of these appearance characteristic parameters, any slight changes in the appearance of the product can be found in time, which can meet the high requirements of quality control. On the other hand, by obtaining a plurality of appearance characteristic parameters reflecting the appearance of the chemical product in real time, these parameters are associated with a set of production process parameters recorded in the preset process knowledge guide map to trace the non-compliant production process parameters. The system can automatically adjust the values of these parameters to adapt to the current production conditions. The dynamic online adjustment mechanism can timely correct the deviation in the production process and reduce the product quality problems caused by improper parameter setting, thereby significantly reducing the product error probability.

[0085] The following is an embodiment of the device of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

[0086] Please refer to Figure 9 which shows the structure of the AI visual detection-based production process parameter dynamic adjustment device provided by an exemplary embodiment of the present application. The AI visual detection-based production process parameter dynamic adjustment device can be realized by software, hardware or a combination of the two to become all or part of an electronic device. The device 1 includes a production video stream acquisition module 10, an appearance characteristic parameter output module 20, a production process parameter analysis module 30, and a parameter value adjustment module 40.

[0087] The production video stream acquisition module 10 is used to acquire the production video stream collected by the 5G explosion-proof camera pre-deployed in the production environment within a preset time period. The production video stream is used to represent the internal information of the reaction kettle, the material feeding information, and the appearance information of the chemical product.

[0088] The appearance characteristic parameter output module 20 is used to input the production video stream into the pre-trained AI visual detection model and output a plurality of appearance characteristic parameters corresponding to the production video stream for reflecting the appearance of the chemical product.

[0089] The production process parameter analysis module 30 is used to analyze the production process parameters in the production equipment of the production environment that do not meet the preset product production conditions based on the plurality of appearance characteristic parameters. The production process parameters that do not meet the preset product production conditions are obtained based on the preset process knowledge guide map. The preset process knowledge guide map is used to record the correlation between the appearance process marks and the production process parameter set of the production equipment in the production environment.

[0090] The parameter value adjustment module 40 is configured to adjust the parameter value of the production process parameter that does not meet the preset product production condition by a preset step size.

[0091] It should be noted that the AI vision detection based production process parameter dynamic adjustment apparatus provided in the above embodiments is used to execute the AI vision detection based production process parameter dynamic adjustment method, and only the division of the above functional modules is used as an example for illustration. In actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the above described functions. In addition, the AI vision detection based production process parameter dynamic adjustment apparatus and the AI vision detection based production process parameter dynamic adjustment method provided in the above embodiments belong to the same concept, and the implementation process is detailed in the method embodiments. Here, it is not repeated.

[0092] The serial numbers of the above embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0093] In the embodiments of the present application, on the one hand, the pre-trained AI vision detection model can extract a plurality of appearance characteristic parameters including physical appearance parameters (such as color, shape) and non-physical appearance parameters (such as texture, luster) from the production video stream collected by the 5G explosion-proof camera. These parameters can comprehensively reflect the appearance information of the chemical product, and provide more accurate and detailed data support for quality control. Through analysis of these appearance characteristic parameters, any slight change in the appearance of the product can be found in time, and the high requirements for quality control can be met. On the other hand, by real-time acquisition of a plurality of appearance characteristic parameters reflecting the appearance of the chemical product, the parameters are associated with a production process parameter set recorded in a preset process knowledge guide map, to trace the non-compliant production process parameters, and the system can automatically adjust the values of the parameters to adapt to the current production conditions. The dynamic online adjustment mechanism can timely correct the deviation in the production process, reduce the product quality problems caused by improper parameter setting, and thus significantly reduce the product error probability.

[0094] The present application also provides a computer readable medium having program instructions stored thereon, which, when executed by a processor, implement the AI vision detection based production process parameter dynamic adjustment method provided by each of the above method embodiments.

[0095] The present application also provides a computer program product containing instructions, which, when executed on a computer, causes the computer to perform the AI vision detection based production process parameter dynamic adjustment method of each of the above method embodiments.

[0096] Please refer to Figure 10 The present application provides a structural schematic diagram of an electronic device. As shown inFigure 10 As shown, the electronic device 1000 can include at least one processor 1001, at least one network interface 1004, a user interface 1003, a memory 1005, at least one communication bus 1002.

[0097] The communication bus 1002 is configured to realize the connection communication between the components.

[0098] The user interface 1003 can include a display, a camera, and optionally a standard wired interface and a wireless interface.

[0099] The network interface 1004 can optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0100] The processor 1001 can include one or more processing cores. The processor 1001 connects various parts of the entire electronic device 1000 through various interfaces and lines, executes various functions of the electronic device 1000 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 1005, and calling data stored in the memory 1005. Optionally, the processor 1001 can be implemented in at least one of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 1001 can be integrated with a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU is mainly used to process an operating system, a user interface, and an application program; the GPU is used to render and draw the content to be displayed on the display; and the modem is used to process wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 1001, but can be realized by a separate chip.

[0101] The memory 1005 can include a random access memory (RAM) and can also include a read-only memory (ROM). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets, or instruction sets. The memory 1005 can include a program storage area and a data storage area, where the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the various method embodiments described above, and the like; the data storage area can store data involved in the various method embodiments described above, and the like. The memory 1005 can also be at least one storage system located away from the aforementioned processor 1001. As shown in Figure 10 The memory 1005 as a computer storage medium can include an operating system, a network communication module, a user interface module, and an AI vision detection based production process parameter dynamic adjustment application program.

[0102] In the electronic device 1000 shown in Figure 10 In the electronic device 1000 shown in

[0103] The production video stream collected by the 5G explosion-proof camera pre-deployed in the production environment is obtained according to a preset period, and the production video stream is used to represent internal information of the reaction kettle, material feeding information, and appearance information of the chemical product;

[0104] The production video stream is input into the pre-trained AI vision detection model, and a plurality of appearance characteristic parameters corresponding to the production video stream and used to reflect the appearance of the chemical product are output;

[0105] According to the plurality of appearance characteristic parameters, production process parameters in the production equipment of the production environment that do not meet the preset product production conditions are analyzed; the production process parameters that do not meet the preset product production conditions are obtained based on a preset process knowledge oriented graph, and the preset process knowledge oriented graph is used to record the correlation between the appearance process mark and the production process parameter set of the production equipment in the production environment;

[0106] The parameter value of the production process parameter that does not meet the preset product production condition is adjusted according to a preset step size.

[0107] In one embodiment, the processor 1001, in operation of analyzing production process parameters that do not meet preset product production conditions in production equipment of a production environment according to a plurality of appearance feature parameters, specifically performs the following operations:

[0108] According to the appearance process mark, the corresponding appearance feature quantization threshold range is obtained from the mapping relationship between the pre-established appearance process mark and the appearance feature quantization threshold range;

[0109] Comparing the appearance feature quantization value with the appearance feature quantization threshold range, the appearance process mark corresponding to the appearance feature quantization value not in the appearance feature quantization threshold range is determined as a plurality of abnormal appearance process marks;

[0110] From the preset process knowledge guide map, production process parameter tracing is performed on each abnormal appearance process mark to obtain a plurality of candidate production process parameters;

[0111] The current value of each candidate production process parameter and its preset standard range are compared to determine the production process parameters that do not meet the preset product generation conditions in the production equipment of the production environment.

[0112] In one embodiment, the processor 1001, in operation of generating a preset process knowledge guide map, specifically performs the following operations:

[0113] Obtain the historical production video stream of the production environment in the preset period;

[0114] Synchronize each key frame in the historical production video stream that reflects the key visual information of the reaction kettle in the reaction period, the material feeding moment, and the chemical product appearance with the first production process parameter set at the corresponding time according to the size of the timestamp;

[0115] Input each key frame into a pre-trained AI visual detection model to separate a plurality of first appearance feature quantization values;

[0116] Mine the correlation relationship between each first appearance feature quantization value and the first production process parameter set;

[0117] According to the correlation relationship, a preset process knowledge guide map is established.

[0118] In one embodiment, the processor 1001, in operation of mining the correlation relationship between each first appearance feature quantization value and the first production process parameter set, specifically performs the following operations:

[0119] Analyze the linear relationship coefficient and the nonlinear monotonic relationship coefficient between each first appearance feature quantization value and each first production process parameter in the set;

[0120] Screen out the first production process parameters with an absolute value of a linear relationship coefficient greater than a first preset threshold and a nonlinear monotonic relationship coefficient less than a second preset threshold as production process parameters correlated with each first appearance feature quantitative value, and a preset multiple value of the second preset threshold is equal to the first preset threshold;

[0121] Bind the production process parameters correlated with each first appearance feature quantitative value and each first appearance feature quantitative value to obtain a correlation relationship between each first appearance feature quantitative value and the first production process parameter set.

[0122] In one embodiment, the processor 1001, when performing the establishment of the preset process knowledge directed graph according to the correlation relationship, specifically performs the following operations:

[0123] Assign a first appearance process label to each first appearance feature quantitative value;

[0124] Assign the first appearance process label to each first appearance feature quantitative value as a root node, and assign the production process parameters correlated with each first appearance feature quantitative value as a plurality of child nodes;

[0125] Establish a knowledge directed graph of the root node and the plurality of child nodes;

[0126] Traverse all root nodes to identify shared child nodes as nested nodes from the knowledge directed graph;

[0127] Based on the nested nodes, perform nested linking on the established knowledge directed graphs to obtain the preset process knowledge directed graph.

[0128] In one embodiment, the processor 1001, when performing the production process parameter tracing of each abnormal appearance process label from the preset process knowledge directed graph to obtain a plurality of candidate production process parameters, specifically performs the following operations:

[0129] From the preset process knowledge directed graph, traverse and find candidate root nodes that meet the tracing conditions under the condition of each abnormal appearance process label as a tracing condition;

[0130] Obtain all child nodes on the candidate root nodes;

[0131] Take the union set of all child nodes to obtain a plurality of candidate production process parameters.

[0132] In one embodiment, the processor 1001, when performing the input of the production video stream into the pre-trained AI visual detection model to output a plurality of appearance feature parameters corresponding to the production video stream for reflecting the appearance of the chemical product, specifically performs the following operations:

[0133] The basic appearance feature extraction module is adopted to extract physical appearance parameters of the chemical product in the production video stream, to obtain original appearance feature parameters.

[0134] The deep appearance feature extraction module is adopted to extract non-physical appearance parameters of the chemical product in the production video stream, to obtain deep appearance feature parameters.

[0135] The original appearance feature parameters and the deep appearance feature parameters are used as a plurality of appearance feature parameters of the production video stream for reflecting the appearance of the chemical product.

[0136] In one embodiment, the processor 1001 specifically performs the following operations in the process of generating the pre-trained AI visual detection model:

[0137] Obtain a sample production video stream of the production environment in a preset period;

[0138] From the sample production video stream, obtain each sample key frame for reflecting the key visual information of the reaction kettle in the reflection middle stage, the material feeding moment, and the appearance of the chemical product;

[0139] For each sample key frame, label the physical appearance parameters about color and shape and the non-physical appearance parameters about texture and luster, to obtain a model training sample;

[0140] The appearance parameter recognition algorithm is adopted to construct the basic appearance feature extraction module;

[0141] The neural network algorithm is adopted to construct the deep appearance feature extraction module;

[0142] According to the basic appearance feature extraction module and the deep appearance feature extraction module, the AI visual detection model is built;

[0143] According to the model training sample, the AI visual detection model is subjected to machine learning, to obtain the pre-trained AI visual detection model.

[0144] In one embodiment, the processor 1001 specifically performs the following operations in the process of obtaining the pre-trained AI visual detection model according to the model training sample:

[0145] The model training sample is input into the AI visual detection model, to output a model loss value;

[0146] In the case where the model loss value reaches the minimum, the pre-trained AI visual detection model is generated; or in the case where the model loss value does not reach the minimum, the model parameters of the AI visual detection model are updated, and the step of inputting the model training sample into the AI visual detection model is continuously performed until the model loss value reaches the minimum.

[0147] In the embodiments of the present application, on the one hand, the pre-trained AI visual detection model can extract a plurality of appearance characteristic parameters including physical appearance parameters (such as color, shape) and non-physical appearance parameters (such as texture, luster) from the production video stream collected by the 5G explosion-proof camera. These parameters can comprehensively reflect the appearance information of the chemical product and provide more accurate and detailed data support for quality control. Through the analysis of these appearance characteristic parameters, any slight changes in the appearance of the product can be found in time, and the high requirements for quality control can be met. On the other hand, by real-time acquisition of a plurality of appearance characteristic parameters reflecting the appearance of the chemical product, the parameters are associated with a production process parameter set recorded in a preset process knowledge guide map to trace the non-compliant production process parameters, and the system can automatically adjust the values of these parameters to adapt to the current production conditions. The dynamic online adjustment mechanism can timely correct the deviation in the production process and reduce the product quality problems caused by improper parameter setting, thereby significantly reducing the product error probability.

[0148] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The program for dynamically adjusting the production process parameters based on AI visual detection can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. The storage medium of the program for dynamically adjusting the production process parameters based on AI visual detection can be a disk, an optical disc, a read-only memory, a random access memory, etc.

[0149] The above only discloses the preferred embodiments of the present application, and of course cannot limit the scope of the rights of the present application. Therefore, equivalent changes made according to the claims of the present application still fall within the scope of the present application.

Claims

1. An AI vision detection-based production process parameter dynamic adjustment method, characterized in that, Applied to a server, the method comprises: According to a preset period, a production video stream collected by a 5G explosion-proof camera pre-deployed in a production environment is obtained, and the production video stream is used to represent internal information of a reaction kettle, material feeding information, and appearance information of a chemical product; The production video stream is input into a pre-trained AI visual detection model, and a plurality of appearance characteristic parameters corresponding to the production video stream and used to reflect the appearance of the chemical product are output; the plurality of appearance characteristic parameters include physical appearance parameters and non-physical appearance parameters; According to the plurality of appearance characteristic parameters, production process parameters in a production device of the production environment that do not meet preset product production conditions are analyzed; the production process parameters that do not meet the preset product production conditions are obtained based on a preset process knowledge guide graph, and the preset process knowledge guide graph is used to record the correlation between appearance process marks and a production process parameter set of the production device in the production environment; the preset process knowledge guide graph is generated according to the following steps, including: obtaining historical production video streams of the production environment within a preset period; synchronizing each key frame in the historical production video stream and a first production process parameter set at a corresponding time according to the size of the time stamp, the key frame being used to reflect key visual information of a reaction intermediate stage of the reaction kettle, a material feeding moment, and an appearance of a chemical product; inputting each key frame into a pre-trained AI visual detection model to separate a plurality of first appearance characteristic quantization values; mining the correlation between each first appearance characteristic quantization value and the first production process parameter set; assigning a first appearance process mark to each first appearance characteristic quantization value; taking the first appearance process mark assigned to each first appearance characteristic quantization value as a root node, and taking production process parameters that have a correlation with each first appearance characteristic quantization value as a plurality of child nodes; establishing a knowledge guide graph of the root node and the plurality of child nodes; traversing all root nodes, identifying shared child nodes as nested nodes from the knowledge guide graph; based on the nested nodes, nesting and linking each established knowledge guide graph to obtain a preset process knowledge guide graph; According to a preset step, the parameter value of the production process parameter that does not meet the preset product production condition is adjusted.

2. The method of claim 1, wherein, Each appearance characteristic parameter includes an appearance process mark and an appearance characteristic quantization value; The analysis of the production process parameters in the production device of the production environment that do not meet the preset product production conditions according to the plurality of appearance characteristic parameters comprises: According to the appearance process mark, a corresponding appearance characteristic quantization threshold range is obtained from a mapping relationship between the appearance process mark and the appearance characteristic quantization threshold range which is pre-established; The appearance characteristic quantization value and the appearance characteristic quantization threshold range are compared to determine an appearance process mark corresponding to an appearance characteristic quantization value that is not in the appearance characteristic quantization threshold range as a plurality of abnormal appearance process marks; Each abnormal appearance process mark is subjected to production process parameter tracing from the preset process knowledge guide graph to obtain a plurality of candidate production process parameters; The current value of each candidate production process parameter is compared with its preset standard range to determine the production process parameters in the production equipment of the production environment that do not meet the preset product generation condition.

3. The method of claim 1, wherein, The correlation between each first appearance feature quantitative value and the first production process parameter set is mined, including: The linear relationship coefficient and the nonlinear monotonic relationship coefficient between each first appearance feature quantitative value and each first production process parameter in the set are analyzed; The first production process parameter whose absolute value of the linear relationship coefficient is greater than a first preset threshold and whose nonlinear monotonic relationship coefficient is less than a second preset threshold is screened out as the production process parameter that has a correlation with the first appearance feature quantitative value, and a preset multiple value of the second preset threshold is equal to the first preset threshold; The production process parameter that has a correlation with each first appearance feature quantitative value and each first appearance feature quantitative value are bound to obtain the correlation between each first appearance feature quantitative value and the first production process parameter set.

4. The method of claim 3, wherein, The correlation includes the production process parameter that has a correlation with each first appearance feature quantitative value.

5. The method of claim 2, wherein, The production process parameter of each abnormal appearance process mark is traced from the preset process knowledge guide graph to obtain a plurality of candidate production process parameters, including: The each abnormal appearance process mark is taken as a tracing condition, and a candidate root node that meets the tracing condition is searched from the preset process knowledge guide graph; All child nodes on the candidate root node are obtained; The union of all child nodes is obtained to obtain a plurality of candidate production process parameters.

6. The method according to any one of claims 1 to 5, characterized in that, The pre-trained AI visual detection model includes a basic appearance feature extraction module and a deep appearance feature extraction module; the basic appearance feature extraction module is used to extract physical appearance parameters of the chemical product in the production video stream, and the deep appearance feature extraction module is used to extract non-physical appearance parameters of the chemical product in the production video stream; The production video stream is input into the pre-trained AI visual detection model, and a plurality of appearance feature parameters corresponding to the production video stream for reflecting the appearance of the chemical product are output, including: The basic appearance feature extraction module is used to extract the physical appearance parameters of the chemical product in the production video stream to obtain original appearance feature parameters; The deep appearance feature extraction module is used to extract the non-physical appearance parameters of the chemical product in the production video stream to obtain deep appearance feature parameters; The original appearance feature parameters and the deep appearance feature parameters are taken as the plurality of appearance feature parameters corresponding to the production video stream for reflecting the appearance of the chemical product.

7. The method of claim 6, wherein, The pre-trained AI visual detection model is generated according to the following steps, including: Sample production video streams of the production environment in a preset period are obtained; Each sample key frame for reflecting key visual information of the reaction intermediate stage, material feeding moment, and appearance of the chemical product of the reaction kettle is obtained from the sample production video stream; For each sample key frame, physical appearance parameters about color and shape and non-physical appearance parameters about texture and gloss are labeled to obtain model training samples; An appearance parameter recognition algorithm is used to construct a basic appearance feature extraction module; A neural network algorithm is used to construct a deep appearance feature extraction module; An AI visual detection model is built according to the basic appearance feature extraction module and the deep appearance feature extraction module; Machine learning is performed on the AI visual detection model according to the model training samples to obtain a pre-trained AI visual detection model.

8. The method of claim 7, wherein, Machine learning is performed on the AI visual detection model according to the model training samples to obtain a pre-trained AI visual detection model, including: The model training samples are input into the AI visual detection model, and a model loss value is output; When the model loss value reaches a minimum, a pre-trained AI visual detection model is generated; or when the model loss value does not reach a minimum, the model parameters of the AI visual detection model are updated, and the step of inputting the model training samples into the AI visual detection model is continued until the model loss value reaches a minimum.

9. An AI vision detection-based production process parameter dynamic adjustment device, characterized in that, The device comprises: A production video stream acquisition module is configured to acquire a production video stream collected by a 5G explosion-proof camera pre-deployed in a production environment within a preset time period, the production video stream being used to represent internal information of a reaction kettle, material feeding information, and appearance information of a chemical product; An appearance feature parameter output module is configured to input the production video stream into a pre-trained AI visual detection model and output a plurality of appearance feature parameters corresponding to the production video stream and reflecting the appearance of the chemical product; the plurality of appearance feature parameters include physical appearance parameters and non-physical appearance parameters. The production process parameter analysis module is configured to analyze production process parameters that do not meet preset product production conditions in the production equipment of the production environment according to the plurality of appearance feature parameters. The production process parameters that do not meet preset product production conditions are obtained based on a preset process knowledge guide map. The preset process knowledge guide map is used to record the correlation between appearance process marks and a production process parameter set of the production equipment in the production environment. The preset process knowledge guide map is generated by the following steps, including: obtaining historical production video streams of the production environment in a preset period; synchronizing each key frame of the historical production video streams that reflects key visual information of a reaction intermediate stage, a material feeding moment, and an appearance of a chemical product of the reaction kettle and a first production process parameter set of a corresponding moment according to the size of a timestamp; inputting each key frame into a pre-trained AI visual detection model to separate a plurality of first appearance feature quantization values; mining a correlation between each first appearance feature quantization value and the first production process parameter set; assigning a first appearance process mark to each first appearance feature quantization value; taking the first appearance process mark assigned to each first appearance feature quantization value as a root node, and taking production process parameters that have a correlation with each first appearance feature quantization value as a plurality of child nodes; establishing a knowledge guide map of the root node and the plurality of child nodes; traversing all root nodes, identifying shared child nodes as nested nodes from the knowledge guide map; and based on the nested nodes, nesting and linking each established knowledge guide map to obtain a preset process knowledge guide map. The parameter value adjustment module is configured to adjust the parameter value of the production process parameter that does not meet the preset product production condition according to a preset step size.

Images