Intelligent efficient automatic production system for color steel plates

Abstract

The invention relates to the technical field of color steel plate production, and discloses an intelligent efficient automatic production system for color steel plates. An acquisition module of the system acquires multi-source heterogeneous data in a production process in real time, wherein the multi-source heterogeneous data comprises raw material attribute parameters and the like. The analysis module performs multi-dimensional feature extraction on the data, such as physical attribute features of raw materials. The decision module generates a production control instruction set based on a multi-objective optimization algorithm. The scheduling module constructs a process knowledge base and stores related production data. And the execution module drives the production equipment to perform automatic processing through a distributed cooperative control algorithm according to the instruction set and the process knowledge base data. The system also has the functions of dealing with sudden interference, verifying the compatibility of the processing path and the like, can reconstruct the processing path in real time, and ensures that the production meets the process requirements. The system realizes intelligentization and high efficiency of color steel plate production, improves the product quality and reduces the production cost.

Description

Technical Field

[0001] The present invention relates to the technical field of color steel plate production, and specifically to an intelligent, efficient and automated production system for color steel plates. Background Art

[0002] In modern industry, color steel plates are widely used in many fields such as construction, household appliances, and transportation due to their excellent properties, such as beauty, corrosion resistance, high strength, etc. With the continuous growth of market demand, higher requirements are put forward for the production efficiency and quality of color steel plates. The traditional production method of color steel plates mainly relies on manual operation and empirical judgment, with many drawbacks, seriously restricting the development of the color steel plate production industry.

[0003] In terms of raw material management, it is difficult for traditional production to accurately control the attribute parameters of raw materials. The quality of raw materials is uneven, and there are differences in physical properties such as thickness, hardness, and chemical composition among different batches of raw materials. Manual inspection often cannot comprehensively and accurately obtain these parameters, resulting in difficulty in making reasonable adjustments according to the actual situation of raw materials during the production process, and affecting the stability of product quality. For example, if the thickness of the raw materials is inconsistent, it may cause uneven thickness during the pressing process of the color steel plate, reducing the qualified rate of the product.

[0004] In terms of equipment operation management, the traditional method cannot monitor the operation status of equipment in real time. During the long-term operation of equipment, due to component wear, aging, etc., faults are likely to occur. Manual inspection can only be carried out when the equipment stops running, and potential problems during equipment operation cannot be detected in time. Once the equipment breaks down suddenly, it will not only cause production interruption, increase maintenance costs and time, but also may damage the semi-finished products that have been produced, affecting the production progress and product quality. In addition, the vibration situation of the equipment directly reflects its operation stability, but it is difficult for traditional methods to deeply analyze the vibration spectrum of the equipment and predict equipment faults in advance.

[0005] There are also problems in production environment management. Factors such as temperature, humidity, and dust in the production environment have an important impact on the production quality of color steel plates. For example, in an environment with high humidity, raw materials are easily affected by moisture, affecting the adhesion of the coating, resulting in defects such as coating peeling on the surface of the color steel plate; while too high or too low temperature may affect the effect of the processing technology, causing the physical properties of the color steel plate to decline. However, the traditional production method lacks real-time monitoring and effective control means for the production environment and cannot respond in time to the impact of environmental changes on production.

[0006] In the quality inspection process, the traditional manual inspection method is inefficient and inaccurate. For the tiny defects on the surface of color steel plates, such as scratches and bubbles, it is easy to miss them during manual inspection, resulting in unqualified products flowing into the market, damaging the reputation of the enterprise and the interests of customers. At the same time, manual inspection cannot conduct quantitative analysis on product quality, making it difficult to trace the root cause of quality problems, which is not conducive to the continuous improvement of the enterprise's production process.

[0007] In terms of production scheduling, traditional production scheduling mainly relies on manual experience and lacks scientific and reasonable planning. Due to the inability to accurately grasp the historical data of production tasks and the actual load capacity of equipment, situations such as equipment idleness or overloaded operation are likely to occur, resulting in low production efficiency and serious energy waste. For example, when arranging production tasks, too many tasks may be assigned to a certain piece of equipment, causing it to be in a high-load operation state for a long time, not only shortening the service life of the equipment but also potentially affecting product quality, while other equipment is in an idle state, causing resource waste. Summary of the Invention

[0008] The purpose of the present invention is to provide an intelligent, efficient and automated production system for color steel plates to solve the problems raised in the above background technology.

[0009] To achieve the above purpose, the present invention provides the following technical solution: An intelligent, efficient and automated production system for color steel plates, the system includes:

[0010] A collection module, used to collect multi-source heterogeneous data in the production process of color steel plates in real time, and the multi-source heterogeneous data includes raw material attribute parameters, equipment operation status data, environmental monitoring data, high-precision image data and process execution logs;

[0011] An analysis module, used to extract multi-dimensional features from the multi-source heterogeneous data, and the multi-dimensional features include raw material physical property features, equipment vibration spectrum features, environmental temperature and humidity dynamic features, surface defect visual features and process time sequence correlation features;

[0012] A decision module, which maps the multi-dimensional features to a unified decision space based on a multi-objective optimization algorithm to generate a production control instruction set;

[0013] A scheduling module, which constructs a process knowledge base according to historical production task data, and the fourth data is stored in the process knowledge base, and the fourth data includes the standard process flow of each color steel plate model, equipment load threshold and historical energy consumption record;

[0014] An execution module, based on the production control instruction set and the fourth data in the process knowledge base, drives the production equipment to complete the automated processing of color steel plates through a distributed collaborative control algorithm.

[0015] Preferably, the automated processing of color steel plates by driving production equipment through a distributed collaborative control algorithm includes:

[0016] Generate an initial equipment scheduling plan according to the process relevance in the production control instruction set and the process knowledge base;

[0017] Construct a production task topology network based on the equipment load threshold, where the nodes in the topology network represent processing equipment and the edges represent the material flow or timing dependence relationship between equipment;

[0018] Optimize the initial equipment scheduling plan through a resource allocation algorithm according to the production task topology network and the real-time equipment operation status data, and output the final processing path and equipment collaboration instructions.

[0019] Preferably, the multi-source heterogeneous data multidimensional feature extraction includes:

[0020] Generate physical attribute classification features for the raw material attribute parameters by using a fuzzy clustering algorithm;

[0021] Extract non-linear dynamic features for the equipment vibration spectrum features by using an adaptive resonance theory model;

[0022] Extract the logical dependence relationship between processes for the process timing correlation features by using a causal inference network.

[0023] Preferably, the mapping of the multidimensional features to a unified decision space based on a multi-objective optimization algorithm includes:

[0024] Calculate the priority weights of each feature through weighted calculation by a multi-attribute decision-making model;

[0025] Input the normalized features into a chaotic particle swarm optimization algorithm to solve the optimal control parameter combination under multiple constraints.

[0026] Preferably, the distributed collaborative control algorithm further includes:

[0027] Construct an equipment cooperation equilibrium model based on game theory. The equilibrium model dynamically adjusts the task allocation ratio between equipment through the Nash bargaining mechanism, and the Nash bargaining mechanism includes energy consumption cost penalty, task delay penalty, and equipment utilization reward.

[0028] Preferably, the resource allocation algorithm is an improved ant colony algorithm, including:

[0029] Encode the production task topology network into a pheromone distribution map, and the pheromone concentration represents the task processing efficiency of equipment nodes;

[0030] Iteratively update the pheromone intensity of equipment nodes through the path selection probability formula;

[0031] Screen the globally optimal machining path according to the convergence condition, where the convergence condition is calculated based on the total path time-consuming and the equipment load balance degree.

[0032] Preferably, the system further includes:

[0033] Adopt a dynamic threshold segmentation algorithm for the environmental monitoring data to extract the characteristics of sudden disturbances in the production environment; when a sudden disturbance is detected, trigger the real-time reconstruction of the machining path, and update the equipment cooperation instruction according to the reconstructed path.

[0034] Preferably, the real-time reconstruction adopts a grey prediction control strategy, and the specific method includes:

[0035] Define the grey correlation degree between the equipment state transition matrix and the environmental disturbance variable;

[0036] Generate candidate reconstruction paths through grey differential equations, and select the path scheme with the highest stability based on the residual check algorithm.

[0037] Preferably, the system further includes:

[0038] Construct a process constraint knowledge graph, where the nodes in the knowledge graph represent process parameters, and the edges represent the constraint relationships or linkage rules between the parameters; during the equipment cooperation process, verify the compatibility between the machining path and the knowledge graph through the constraint propagation algorithm.

[0039] Preferably, the constraint propagation algorithm is a two-way constraint reasoning algorithm based on interval algebra, including:

[0040] Extract the local constraint subgraph from the current machining path;

[0041] Deduce the feasible parameter range forward in the knowledge graph, and verify the solvability of the constraint conflict backward.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] In modern industry, color steel plates are widely used in many fields such as construction, home appliances, and transportation due to their excellent properties, such as beauty, corrosion resistance, high strength, etc. With the continuous growth of market demand, higher requirements are put forward for the production efficiency and quality of color steel plates. The traditional production method of color steel plates mainly relies on manual operation and empirical judgment, with many drawbacks, seriously restricting the development of the color steel plate production industry.

[0044] In terms of raw material management, it is difficult for traditional production to accurately control the attribute parameters of raw materials. The quality of raw materials is uneven, and there are differences in physical properties such as thickness, hardness, and chemical composition among raw materials of different batches. Manual inspection often fails to comprehensively and accurately obtain these parameters, making it difficult to make reasonable adjustments according to the actual situation of raw materials during the production process, which affects the stability of product quality. For example, if the thickness of the raw materials is inconsistent, it may cause uneven thickness during the pressing process of color steel plates, reducing the qualified rate of products.

[0045] In terms of equipment operation management, the traditional method cannot monitor the operation status of equipment in real time. During the long-term operation of equipment, due to reasons such as component wear and aging, it is prone to failures. Manual inspection can only be carried out when the equipment stops running, and potential problems during equipment operation cannot be detected in a timely manner. Once the equipment suddenly fails, it will not only cause production interruption, increase maintenance costs and time, but also may damage the semi-finished products that have been produced, affecting production progress and product quality. In addition, the vibration situation of the equipment directly reflects its operation stability, but it is difficult to deeply analyze the vibration spectrum of the equipment by traditional methods, and it is impossible to predict equipment failures in advance.

[0046] There are also problems in production environment management. Factors such as temperature, humidity, and dust in the production environment have an important impact on the production quality of color steel plates. For example, in an environment with high humidity, raw materials are easily affected by moisture, which affects the adhesion of the coating, resulting in defects such as coating peeling on the surface of color steel plates; while too high or too low temperature may affect the effect of the processing technology, causing the physical properties of color steel plates to decline. However, the traditional production method lacks real-time monitoring and effective control means for the production environment and cannot respond in a timely manner to the impact of environmental changes on production.

[0047] In the quality inspection link, the traditional manual inspection method is inefficient and inaccurate. Minor defects on the surface of color steel plates, such as scratches and bubbles, are easily missed by manual inspection, resulting in unqualified products flowing into the market, damaging the reputation of the enterprise and the interests of customers. At the same time, manual inspection cannot conduct quantitative analysis of product quality, making it difficult to trace the root cause of quality problems and being unfavorable for the enterprise to continuously improve the production process.

[0048] In terms of production scheduling, traditional production scheduling mainly relies on manual experience and lacks scientific and reasonable planning. Due to the inability to accurately grasp the historical data of production tasks and the actual load capacity of equipment, situations such as equipment idleness or overloading are likely to occur, resulting in low production efficiency and serious energy waste. For example, when arranging production tasks, too many tasks may be assigned to a certain piece of equipment, causing it to be in a high-load operation state for a long time, not only shortening the service life of the equipment, but also possibly affecting product quality, while other equipment is in an idle state, causing resource waste. Description of the Drawings

[0049] Figure 1 This is the working principle diagram of the intelligent, efficient and automated production system for color steel plates of the present invention;

[0050] Figure 2 This is the working principle diagram of the real-time reconstruction of the production and processing path of color steel plates based on environmental monitoring;

[0051] Figure 3 This is the working principle diagram of the verification of the collaborative processing path of color steel plate equipment based on the process constraint knowledge graph;

[0052] Figure 4 This is the working principle diagram of the verification and adjustment of the constraint conflict of the color steel plate processing path. Specific embodiments

[0053] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0054] Please refer to Figures 1 - 4 , the present invention provides an intelligent, efficient and automated production system for color steel plates, and its overall implementation plan is as follows:

[0055] The acquisition module is responsible for real-time acquisition of multi-source heterogeneous data. In terms of raw materials, accurately obtain the attribute parameters of raw materials, covering the thickness, width, material composition, mechanical properties, etc. of raw materials, which directly affect the final quality of color steel plates. For the equipment operation status data, collect information such as the rotation speed, temperature, pressure, current, etc. of the equipment to comprehensively master the operation status of the equipment. Environmental monitoring data cannot be underestimated, including the temperature, humidity, air quality, etc. of the production workshop, because environmental factors may affect the production process and product quality. High-precision image data is obtained through devices such as industrial cameras for detecting whether there are defects on the surface of color steel plates, such as scratches, bubbles, unevenness, etc. In addition, the process execution log details information such as the start time, end time, and operator of each production process.

[0056] After receiving the data transmitted by the acquisition module, the analysis module performs multi-dimensional feature extraction. For the raw material attribute parameters, the fuzzy clustering algorithm is used to classify the raw materials according to their physical attributes. For example, the raw materials are divided into different categories according to different materials and thicknesses, so as to carry out more targeted production control in the follow-up. For the equipment vibration spectrum characteristics, the adaptive resonance theory model is adopted. This model can extract non-linear dynamic characteristics from complex equipment vibration data and help analyze the potential fault risks of the equipment. When dealing with the process time sequence correlation characteristics, with the help of the causal inference network, the logical dependence relationship between processes is deeply explored. For example, it is determined that a certain process must start after another process is completed, providing a basis for the optimization of the production process.

[0057] Based on the multi-objective optimization algorithm, the decision-making module maps the multi-dimensional features to a unified decision-making space. The multi-objective optimization algorithm comprehensively considers multiple objectives such as product quality, production efficiency, and production cost. Through the multi-attribute decision-making model, priority weights are assigned to each feature. For example, a higher weight is given to the visual feature of surface defects that has a greater impact on product quality. Then the normalized features are input into the chaotic particle swarm optimization algorithm to solve the optimal control parameter combination under multiple constraints, and finally a production control instruction set is generated to achieve precise control of the production process.

[0058] The scheduling module constructs a process knowledge base based on historical production task data. The process knowledge base stores rich fourth data. For each color steel plate model, the standard process flow is detailedly recorded, clearly stating which processes are required for the production of each model of color steel plate and the sequence of the processes. The equipment load threshold limits the maximum load that each production equipment can bear within the safe operating range, avoiding equipment damage due to overload. The historical energy consumption records are used to analyze the energy consumption under different production tasks and equipment operating states, providing a reference for optimizing production scheduling and reducing energy consumption.

[0059] According to the production control instruction set and the fourth data in the process knowledge base, the execution module drives the production equipment to complete the automated processing of color steel plates with the help of the distributed cooperative control algorithm. The execution module initially plans the task allocation and processing sequence of each equipment based on the process relevance in the production control instruction set and the process knowledge base, and generates an initial equipment scheduling plan. Then a production task topology network is constructed based on the equipment load threshold. The nodes in the network represent processing equipment, and the edges represent the material flow or time sequence dependence relationship between equipment, clearly presenting the relationship between equipment in the entire production process. Then, according to the production task topology network and the real-time equipment operation state data, the initial equipment scheduling plan is optimized through the resource allocation algorithm, and finally the best processing path and equipment cooperation instructions are determined to ensure the efficient and stable progress of the production process.

[0060] The following further illustrates the implementation of the present invention in combination with Embodiments 1 to 5.

[0061] Example 1:

[0062] This example details the process of driving production equipment to complete the automated processing of color steel plates through a distributed collaborative control algorithm in the execution module. The execution module first generates an initial equipment scheduling plan based on the process relevance in the production control instruction set and the process knowledge base. Suppose a certain type of color steel plate is to be produced, and the standard process flow specified in the process knowledge base is: first, pre-treat the raw materials on equipment A, then perform color coating operations on equipment B, and finally perform cutting and forming on equipment C. The initial equipment scheduling plan will initially arrange for equipment A to start working according to this process information. After completing the pre-treatment, the materials will be transported to equipment B, and after equipment B completes the color coating, they will be transported to equipment C.

[0063] Construct a production task topology network based on the equipment load thresholds in the process knowledge base. Let the equipment load thresholds be T A 、T B 、T C , and these thresholds are set according to the performance parameters and safe operation standards of the equipment. In the topology network, the nodes are equipment A, equipment B, and equipment C respectively, and the edges represent the material flow relationship. For example, the edge from equipment A to equipment B means that the material flows from equipment A to equipment B, and the edge from equipment B to equipment C means that the material flows from equipment B to equipment C. At the same time, the edges also contain time sequence dependencies, meaning that equipment B must start working after equipment A completes its work and the materials are transported in place, and the same applies to equipment C.

[0064] According to the production task topology network and the real-time equipment operation status data, optimize the initial equipment scheduling plan through a resource allocation algorithm. The real-time equipment operation status data includes the current load conditions of the equipment. Let the current load of equipment A be L A , the current load of equipment B be L B , and the current load of equipment C be L C . The resource allocation algorithm will comprehensively consider this data. If the current load of equipment A is close to its load threshold T A , while the load of equipment B is relatively low, the algorithm may adjust the initial plan to let other parallel tasks be carried out on equipment B first to balance the load between the equipment. Through continuous adjustment and optimization, finally output the final processing path and equipment collaboration instructions that meet the production requirements to ensure the efficient progress of production.

[0065] Example 2:

[0066] This embodiment focuses on the specific operations of multi-dimensional feature extraction for multi-source heterogeneous data in the analysis module. For the raw material attribute parameters, a fuzzy clustering algorithm is used to generate physical attribute classification features. Suppose the raw material attribute parameters include thickness t, material composition ratio m, etc. The fuzzy clustering algorithm first determines the number of clustering categories k. For example, k = 3 categories are set according to common raw material types. Then, the membership degree of each raw material sample to each clustering center is calculated. The membership degree calculation formula is:

[0067]

[0068] where u ij represents the membership degree of the i-th raw material sample belonging to the j-th cluster, d(x i ,c j ) represents the distance between the i-th raw material sample and the j-th clustering center, and m is a parameter greater than 1, usually taken as 2. By continuously iterating and updating the clustering center and membership degree, the raw materials are finally classified into different categories according to their physical attributes.

[0069] For the equipment vibration spectrum features, an adaptive resonance theory model is used to extract non-linear dynamic features. The equipment vibration data contains multiple frequency components. Let the vibration signal be x(t), and the spectrum X(f) is obtained through Fourier transform. The adaptive resonance theory model will process the spectrum X(f). The key parameter in the model is the vigilance parameter ρ, which determines the similarity judgment criterion for the input pattern. When the similarity between the input spectrum feature and the existing category pattern is lower than ρ, the model will create a new category to store this feature, thus effectively extracting the non-linear dynamic features in the equipment vibration spectrum and helping to judge whether the operation state of the equipment is normal.

[0070] For the process time sequence correlation features, a causal inference network is used to extract the logical dependence relationship between processes. Suppose the process set is {P1, P2, …, P n}. The causal inference network analyzes the sequence of process execution times in the process execution log to judge the causal relationship between processes. For example, if process P i always starts after process P j is completed, and the time interval between them has a certain pattern, then it can be inferred that process P j is the pre-process of process P i , and there is a logical dependence relationship. This relationship provides an important basis for optimizing the production process.

[0071] Example 3:

[0072] This embodiment details the process of mapping multi-dimensional features to a unified decision space in the decision-making module based on a multi-objective optimization algorithm. First, the priority weights of each feature are calculated through a multi-attribute decision-making model. Let the multi-dimensional feature set be {F1, F2, …, F n}, for example, F1 represents the physical property features of raw materials, F2 represents the vibration spectrum features of equipment, etc. The importance of each feature is represented by the weight w i , and the calculation of the weight is based on multi-attribute decision-making methods such as the analytic hierarchy process. Suppose there are three objectives: product quality G1, production efficiency G2, and production cost G3. A judgment matrix is constructed through methods such as expert scoring to calculate the relative importance of each feature for different objectives, and then the weight w i of each feature is obtained. The calculation formula is:

[0073]

[0074] where a ij is the relative importance coefficient of feature F i for objective G j , and is the weight of objective G j .

[0075] The normalized features are input into the chaotic particle swarm optimization algorithm to solve the optimal control parameter combination under multiple constraints. By continuously iterating and updating the positions and velocities of the particles, the optimal control parameter combination is found under the satisfaction of multiple constraints (such as equipment load constraints, process parameter constraints, etc.), and a production control instruction set is generated.

[0076] Embodiment 4:

[0077] This embodiment mainly elaborates on the implementation process of constructing an equipment cooperation equilibrium model based on game theory and using an improved ant colony algorithm as a resource allocation algorithm in the distributed cooperative control algorithm. An equipment cooperation equilibrium model is constructed based on game theory, and this model dynamically adjusts the task allocation ratio between devices through the Nash bargaining mechanism. Let the device set be {E1, E2, …, E m}. In the Nash bargaining mechanism, energy consumption cost penalty, task delay penalty, and device utilization reward are considered. Suppose the penalty value of the energy consumption cost for device E i to complete a task is , the penalty value of the task delay time is , and the reward value of the device utilization is U i . The revenue function of device E i can be expressed as:

[0078]

[0079] Among them, α, β, and γ are weight coefficients used to adjust the influence degree of different factors on the benefit. Through the Nash bargaining mechanism, each device negotiates according to its own benefit function and dynamically adjusts the task allocation ratio to achieve the overall optimal cooperation effect.

[0080] An improved ant colony algorithm is adopted as the resource allocation algorithm. First, the production task topology network is encoded into a pheromone distribution map, and the pheromone concentration represents the task processing efficiency of the device nodes. Let the pheromone concentration be τ ij , representing the pheromone concentration on the path from device i to device j. The path selection probability formula is:

[0081]

[0082] Among them, α and β are influence factors used to control the relative importance of the pheromone concentration and the heuristic information. η ij is the heuristic information, which is usually related to factors such as the distance between devices and the task processing time. allowed represents the set of paths that the ant can currently choose. The pheromone intensity of the device nodes is iteratively updated through the path selection probability formula. When the convergence condition (calculated based on the total path time T total and the device load balance degree B) is satisfied, the global optimal processing path is screened out to achieve the reasonable allocation of production tasks and the efficient cooperation of devices.

[0083] Embodiment 5:

[0084] This embodiment covers the processing of environmental monitoring data in the system, the real-time reconstruction of the processing path, and the construction of a process constraint knowledge graph and the application of a constraint propagation algorithm. The dynamic threshold segmentation algorithm is used for the environmental monitoring data to extract the sudden interference characteristics in the production environment. Let the environmental monitoring data (such as temperature T, humidity H, etc.) be time series data x(t). The dynamic threshold segmentation algorithm first calculates the mean μ and standard deviation σ of the data under normal conditions based on historical data. The dynamic threshold T th is calculated by the formula:

[0085] T th = μ + k × σ

[0086] Among them, k is a coefficient set according to the actual situation. When the monitoring data x(t) exceeds the dynamic threshold T th , it is determined that a sudden interference is detected, and at this time, the real-time reconstruction of the processing path is triggered.

[0087] The real-time reconstruction adopts a grey prediction control strategy. The grey correlation degree between the defined device state transition matrix A and the environmental interference variable is defined. Candidate reconstruction paths are generated through grey differential equations, and the path scheme with the highest stability is selected based on the residual check algorithm.

[0088] Construct a process constraint knowledge graph, where the nodes in the knowledge graph represent process parameters and the edges represent the constraint relationships or linkage rules between the parameters. During the equipment collaboration process, verify the compatibility between the processing path and the knowledge graph through a two-way constraint reasoning algorithm based on interval algebra. Extract a local constraint subgraph from the current processing path. Suppose the local constraint subgraph contains process parameters such as P1, P2, etc. Deduce the feasible parameter range forward in the knowledge graph. For example, according to the value range of parameter P1 and its constraint relationship with P2, calculate the feasible value range of P2. Then verify the resolvability of constraint conflicts in reverse. If no conflict is found during the reverse verification process, it indicates that the current processing path is compatible with the knowledge graph; otherwise, the processing path needs to be adjusted to ensure that the production process meets the process requirements.

[0089] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0090] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent and efficient automated production system for color steel plates, characterized in that, Including: A data acquisition module for real-time acquisition of multi-source heterogeneous data in the production process of color steel plates. The multi-source heterogeneous data includes raw material attribute parameters, equipment operation status data, environmental monitoring data, high-precision image data, and process execution logs. An analysis module for performing multi-dimensional feature extraction on the multi-source heterogeneous data. The multi-dimensional features include raw material physical attribute features, equipment vibration spectrum features, environmental temperature and humidity dynamic features, surface defect visual features, and process time series correlation features. A decision-making module that maps the multi-dimensional features to a unified decision space based on a multi-objective optimization algorithm to generate a production control instruction set. A scheduling module that constructs a process knowledge base according to historical production task data. The fourth data stored in the process knowledge base includes the standard process flow of each color steel plate model, equipment load thresholds, and historical energy consumption records. An execution module that drives production equipment to complete the automated processing of color steel plates based on the production control instruction set and the fourth data in the process knowledge base through a distributed collaborative control algorithm.

2. The intelligent and efficient automated production system for color steel plates according to claim 1, wherein The driving of production equipment to complete the automated processing of color steel plates through a distributed collaborative control algorithm includes: Generating an initial equipment scheduling plan according to the process relevance between the production control instruction set and the process knowledge base. Constructing a production task topology network based on the equipment load thresholds. The nodes in the topology network represent processing equipment, and the edges represent the material flow or time series dependency relationship between the equipment. Optimizing the initial equipment scheduling plan through a resource allocation algorithm according to the production task topology network and real-time equipment operation status data, and outputting the final processing path and equipment collaboration instructions.

3. The intelligent and efficient automated production system for color steel plates according to claim 1, characterized in that, The multi-dimensional feature extraction of the multi-source heterogeneous data includes: Using a fuzzy clustering algorithm for the raw material attribute parameters to generate physical attribute classification features. Using an adaptive resonance theory model to extract non-linear dynamic features for the equipment vibration spectrum features. Using a causal inference network to extract the logical dependency relationship between processes for the process time series correlation features.

4. The intelligent and efficient automated production system for color steel plates according to claim 3, wherein, The mapping of the multi-dimensional features to a unified decision space based on a multi-objective optimization algorithm includes: Calculating the priority weights of each feature through a multi-attribute decision-making model with weighted calculation. Inputting the normalized features into a chaotic particle swarm optimization algorithm to solve the optimal control parameter combination under multiple constraints.

5. The intelligent and highly efficient automated production system for color steel plates according to claim 2, wherein The distributed collaborative control algorithm further includes: Constructing an equipment cooperation equilibrium model based on game theory. The equilibrium model dynamically adjusts the task allocation ratio between equipment through a Nash bargaining mechanism. The Nash bargaining mechanism includes energy consumption cost penalty, task delay penalty, and equipment utilization reward.

6. The intelligent and efficient automated production system for color steel plates according to claim 5, characterized in that The resource allocation algorithm is an improved ant colony algorithm, including: Encoding the production task topology network into a pheromone distribution map, where the pheromone concentration represents the task processing efficiency of equipment nodes. Iteratively updating the pheromone intensity of equipment nodes through a path selection probability formula. Selecting the global optimal processing path according to the convergence condition, and the convergence condition is calculated based on the total path time and equipment load balance.

7. The intelligent and efficient automated production system for color steel plates according to claim 1, characterized in that The system further includes: Apply the dynamic threshold segmentation algorithm to the environmental monitoring data to extract the characteristics of sudden disturbances in the production environment; when a sudden disturbance is detected, trigger the real-time reconstruction of the processing path and update the device cooperation instructions according to the reconstructed path.

8. The intelligent and efficient automated production system for color steel plates according to claim 7, characterized in that, The real-time reconstruction adopts the grey prediction control strategy, and the specific method includes: Define the grey correlation degree between the device state transition matrix and the environmental disturbance variable; Generate candidate reconstruction paths through grey differential equations, and select the path scheme with the highest stability based on the residual check algorithm.

9. The intelligent and efficient automated production system for color steel plates according to claim 1, wherein The system also includes: Construct a process constraint knowledge graph, where the nodes in the knowledge graph represent process parameters, and the edges represent the constraint relationships or linkage rules between the parameters; during the device cooperation process, verify the compatibility between the processing path and the knowledge graph through the constraint propagation algorithm.

10. The intelligent and efficient automated production system for color steel plates according to claim 9, wherein, The constraint propagation algorithm is a two-way constraint reasoning algorithm based on interval algebra, including: Extract the local constraint subgraph from the current processing path; Deduce the feasible parameter range forward in the knowledge graph and verify the solvability of constraint conflicts backward.

Images