An intelligent scheduling system for metal sheet processing

Through the intelligent scheduling system, the scheduling scheme is optimized using particle swarm algorithm and gray wolf algorithm, and the device status is monitored in the Internet of Things, the problem that traditional scheduling methods are difficult to cope with complex production environments is solved, and the metal sheet processing is realized Efficiency, rational use of resources and cost reduction.

CN118466429BActive Publication Date: 2024-11-08NANTONG SUHE LAMP PARTS
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410909492.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-08
Publication Date
2024-11-08
Estimated Expiration
2044-07-08

AI Technical Summary

Technical Problem

The traditional metal sheet processing and scheduling method relies on manual experience and simple rules, making it difficult to cope with complex and changeable production environments, resulting in long production cycles, low equipment utilization rate and high energy consumption, especially when facing multiple varieties and small batch orders, it is difficult to quickly. response.

Method used

An intelligent scheduling system for metal sheet processing is designed, including data collection module, data preprocessing module, solution output module, evaluation and sorting module and real-time monitoring module. Data is collected in real time through sensors and coordinated evolution of particle swarm algorithm and gray wolf algorithm. Intelligent scheduling model generates the optimal scheduling solution, and monitors the device status in real time based on the Internet of Things, and adjusts parameters to deal with abnormal situations.

Benefits of technology

It significantly improves the efficiency of metal sheet processing, realizes the rational allocation and efficient utilization of resources, reduces processing costs, and avoids waste of resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118466429B_ABST
    Figure CN118466429B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of intelligent scheduling, and discloses an intelligent scheduling system for metal plate processing, which comprises a data collection module, a data preprocessing module, a scheme output module, an evaluation and sorting module and a real-time monitoring module, and constructs a processing data set based on initial working state data and processing order information; preprocesses data in the processing data set to obtain a preprocessed processing data set; uses a particle swarm algorithm and a grey wolf algorithm for co-evolution to optimize parameters of an intelligent scheduling model, and outputs a processing scheduling scheme; evaluates and sorts the processing scheduling scheme based on preset processing indicators to obtain an optimal scheduling scheme, generates operation instructions, and sends the operation instructions to processing equipment for metal plates; monitors equipment status in real time, and when an abnormality occurs in the processing equipment, adjusts the processing equipment based on the Internet of Things to cope with the abnormality until the processing order for the metal plate is completed; the present invention significantly improves the efficiency of metal plate processing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent scheduling, and in particular to an intelligent scheduling system for metal plate processing. Background Art

[0002] Metal sheets refer to flat sheets made of metal. Common metal materials include aluminum, steel, copper, zinc, magnesium, etc. According to different thicknesses, metal sheets can be divided into thin sheets (less than 3mm thick), medium sheets (3-60mm thick) and thick sheets (greater than 60mm thick). According to different metal materials and usage environments, metal sheets can be further subdivided into anti-corrosion sheets, container sheets, heat insulation sheets, decorative sheets, etc.

[0003] There are many types of metal sheets, each with its own unique properties and uses. For example, electrogalvanized steel sheets (SECC) have good corrosion resistance and maintain the processability of cold-rolled sheets, and are widely used in home appliances, computer cases, door panels, etc.; copper sheets have good mechanical properties and corrosion resistance, but are prone to corrosion cracking; aluminum alloy sheets are lightweight, corrosion-resistant and easy to process, and are often used in construction, furniture and decoration projects; stainless steel sheets have high strength and corrosion resistance, and can be used for a long time in various harsh environments.

[0004] The sheet metal processing industry is an important part of the manufacturing industry and is widely used in the automotive, construction, aerospace and other fields. Traditional sheet metal processing scheduling methods mainly rely on manual experience and simple rules, which are difficult to cope with complex and changing production environments. For example, when faced with orders of multiple varieties and small batches, traditional methods often find it difficult to respond quickly and schedule effectively, resulting in long production cycles, low equipment utilization, and high energy consumption. Summary of the invention

[0005] The purpose of the present invention is to solve the above problems and to design an intelligent scheduling system for metal sheet processing.

[0006] The present invention provides an intelligent scheduling system for metal sheet processing, which includes a data collection module, a data preprocessing module, a solution output module, an evaluation and sorting module, and a real-time monitoring module, wherein:

[0007] A data collection module, used to collect initial working state data of processing equipment in real time through sensors, collect processing order information of metal sheets, and construct a processing data set based on the initial working state data and the processing order information;

[0008] A data preprocessing module, used for preprocessing the data in the processed data set to obtain a preprocessed processed data set;

[0009] A scheme output module is used to construct an intelligent scheduling model, and use the particle swarm algorithm and the gray wolf algorithm to co-evolve to optimize the parameters of the intelligent scheduling model, and output the processing scheduling scheme through the intelligent scheduling model and the pre-processed processing data set;

[0010] An evaluation and sorting module is used to evaluate and sort the processing scheduling scheme based on preset processing indicators to obtain an optimal scheduling scheme, generate an operation instruction based on the optimal scheduling scheme, and send the operation instruction to the processing equipment of the metal sheet;

[0011] The real-time monitoring module is used to monitor the equipment status in real time during the processing of the processing equipment. When there is an abnormal situation in the processing equipment, the processing equipment is adjusted based on the Internet of Things to deal with the abnormal situation until the processing order of the metal sheet is completed.

[0012] Optionally, in a first implementation of the present invention, the data preprocessing module includes a mean processing submodule, a denoising processing submodule, a deep optimization submodule and a classification submodule, wherein:

[0013] A mean value processing submodule is used to obtain data in the processed data set and perform mean value processing on the time series data in the processed data set by discrete Fourier transform;

[0014] A denoising processing submodule is used to calculate the expected value of the energy of the actual data and noise data of the time series data in the processed data set after the averaging processing. When the expected value of the energy of the actual data is greater than the expected value of the energy of the noise data, the current time series is maintained. Otherwise, the actual data will be submerged in the noise data, and the data in the processed data set is denoised.

[0015] The deep optimization submodule is used to input the processed data set after denoising into the two-layer neural network, extract the data feature information of the processed data set after denoising through the bidirectional LSTM layer, and perform deep optimization on the extracted data feature information;

[0016] The classification submodule is used to transfer the data feature information after deep optimization to the output layer, and transfer the data feature information to Softmax for classification to obtain the preprocessed processed data set.

[0017] Optionally, in a second implementation of the present invention, the solution output module includes a parameter setting submodule, a first calculation submodule, a local search submodule, a selection submodule and a decoding submodule, wherein:

[0018] The parameter setting submodule is used to obtain n processing orders and m processing equipment from the preprocessed processing data set, encode the production of n processing orders on m processing equipment and initialize the population, randomly generate the initial population, set the maximum number of iterations, population size, external archive set capacity, and local search mode change times

[0019] The first calculation submodule is used to copy the initialization population to generate a second population, use the particle swarm algorithm and the gray wolf algorithm as search modes for the individuals of the initialization population and the second population, generate a sub-population, and calculate the corresponding objective function values ​​of the individuals of the sub-population;

[0020] The local search submodule is used to perform a local search on each individual when the number of local search mode changes has not been reached. If the new individual found dominates the original individual, the new individual will replace the original individual. If the two do not dominate each other, the probability of whether to replace is 1 / 2.

[0021] The selection submodule is used to select the non-inferior solution set from the offspring individuals by using the dominance relationship, delete the dominated individuals in the non-inferior solution set, and put the remaining individuals into the external archive set. In the external archive set, the global optimal particle of the group is selected, and the α, β and δ wolves in the gray wolf search;

[0022] The decoding submodule is used to repeat iterations until the maximum number of iterations is reached, sort and decode, obtain the sequence of processing orders on the processing equipment, obtain the processing equipment number, and output the processing scheduling plan.

[0023] Optionally, in a third implementation of the present invention, the evaluation and ranking module includes a setting submodule, a second calculation submodule and a generating instruction submodule, wherein:

[0024] The setting submodule is used to set the evaluation function according to the preset processing indicators:

[0025] F(x)=w1*C(x)+w2*T(x)+w3*Q(x)+w4*U(x)

[0026] Among them, F(x) is the comprehensive evaluation score, x represents a set of processing scheduling solutions, C(x) represents the sum of costs, T(x) represents the total processing time, Q(x) represents the quality index, U(x) represents the equipment utilization rate, w1, w 2、 w 3、 w4 are the weights of cost, time, quality, and equipment utilization;

[0027] A second calculation submodule is used to calculate a comprehensive evaluation score for all processing scheduling solutions based on the evaluation function, sort all processing scheduling solutions according to the comprehensive evaluation score, and select the solution with the highest score as the optimal scheduling solution;

[0028] The instruction generation submodule is used to generate operation instructions based on the optimal scheduling solution and send the operation instructions to the metal sheet processing equipment.

[0029] Optionally, in a fourth implementation of the present invention, the real-time monitoring module includes a data acquisition submodule, a health assessment submodule, a parameter adjustment submodule and a monitoring progress submodule, wherein:

[0030] The data acquisition submodule is used to connect the equipment sensors of all processing equipment to the Internet of Things platform, collect data at preset time intervals, and obtain equipment status data;

[0031] A health assessment submodule, used to assess the current health status of the processing equipment based on the equipment status data and output a corresponding equipment health level;

[0032] A parameter adjustment submodule is used to trigger an exception handling process based on the health level of the equipment, and the IoT monitoring automatically adjusts the parameters of the processing equipment through the actuator;

[0033] The progress monitoring submodule is used to generate an alarm message and notify the operator if the Internet of Things cannot adjust the parameters, and reselect the scheduling plan based on the processing scheduling plan. The Internet of Things monitors the processing progress of the metal sheet until the processing order of the metal sheet is completed.

[0034] Optionally, in a fifth implementation of the present invention, the health assessment submodule includes a filling unit, an initialization unit, an input unit, a training unit, an evaluation unit and a trigger unit, wherein:

[0035] A filling unit is used to obtain the health data samples of the metal sheet processing equipment, pre-process the acquired health data samples, remove invalid data points, and fill the missing data using the interpolation method to obtain a training set;

[0036] Initialization unit, used to initialize AdaBoost parameters, set the number of weak classifiers, and initialize the distribution weights of training samples in the training set;

[0037] The input unit is used to input the training set, use the multi-granularity scanning stage strategy to maximize the sequential relationship of data features, input the feature vector processed by the multi-granularity scanning stage, use the cascade forest stage to analyze and process the feature vector processed by the multi-granularity scanning stage layer by layer, average the classification vector generated by the last level, and take the classification result of the category corresponding to the maximum value;

[0038] The training unit is used to calculate the weight coefficient of the base classifier and update the sample distribution, train the weak classifier until the iteration round reaches the maximum value, and output the strong classifier;

[0039] An evaluation unit, used to input the equipment status data into a strong classifier, evaluate the current health status of the processing equipment using the trained strong classifier, and output a corresponding equipment health level;

[0040] The trigger unit is used to trigger the exception handling process when the equipment health level is abnormal. The IoT monitoring automatically adjusts the parameters of the processing equipment through the actuator.

[0041] Optionally, in a sixth implementation of the present invention, the parameter adjustment submodule includes a healthy state unit, a sub-healthy state unit, a mild state unit and an abnormal state unit, wherein:

[0042] A health status unit is used for when the equipment health level is in a healthy state, the processing equipment indicators are normal, and the processing equipment works normally until the processing order of the metal sheet is completed;

[0043] Sub-health status unit, which is used to trigger the exception handling process when the equipment health level is in a sub-healthy state, and automatically adjust the parameters of the processing equipment until the processing order of the metal sheet is completed;

[0044] The minor status unit is used to strengthen the health monitoring of the processing equipment when the equipment health level is in a minor state and there is a minor abnormal fault in the processing equipment, and automatically adjust the parameters of the processing equipment until the processing order of the metal sheet is completed;

[0045] The abnormal status unit is used to stop the operation of the processing equipment when the equipment health level is in an abnormal state, generate an alarm message and notify the operator.

[0046] Optionally, in a seventh implementation of the present invention, a method for implementing an intelligent scheduling system for metal sheet processing includes the following steps:

[0047] Collecting initial working state data of processing equipment in real time through sensors, collecting processing order information of metal plates, and constructing a processing data set based on the initial working state data and the processing order information;

[0048] Preprocessing the data in the processing data set to obtain a preprocessed processing data set, constructing an intelligent scheduling model, and using a particle swarm algorithm and a gray wolf algorithm for co-evolution to optimize the parameters of the intelligent scheduling model, and outputting a processing scheduling plan through the intelligent scheduling model and the preprocessed processing data set;

[0049] Based on the preset processing indicators, the processing scheduling schemes are evaluated and sorted to obtain the optimal scheduling scheme, an operation instruction is generated based on the optimal scheduling scheme, and the operation instruction is sent to the processing equipment of the metal sheet;

[0050] During the processing of the processing equipment, the equipment status is monitored in real time. When there is an abnormal situation in the processing equipment, the processing equipment is adjusted based on the Internet of Things to deal with the abnormal situation until the processing order of the metal sheet is completed.

[0051] Optionally, in an eighth implementation of the present invention, preprocessing the data in the processed data set to obtain the preprocessed processed data set includes the following steps:

[0052] Acquire data in the processed data set, and perform averaging processing on the time series data in the processed data set by discrete Fourier transform;

[0053] Calculate the expected value of the energy of the actual data and noise data of the time series data in the processed data set after averaging processing. When the expected value of the energy of the actual data is greater than the expected value of the energy of the noise data, maintain the current time series. Otherwise, the actual data will be submerged in the noise data. Perform denoising processing on the data in the processed data set.

[0054] The denoised processed data set is input into a two-layer neural network, and the data feature information of the denoised processed data set is extracted through a bidirectional LSTM layer, and the extracted data feature information is deeply optimized;

[0055] The data feature information after deep optimization is transmitted to the output layer, and the data feature information is transmitted to Softmax for classification to obtain the preprocessed processed data set.

[0056] Optionally, in a ninth implementation of the present invention, constructing an intelligent scheduling model, and using a particle swarm algorithm and a gray wolf algorithm for co-evolution to optimize parameters of the intelligent scheduling model, and outputting a processing scheduling plan through the intelligent scheduling model and a preprocessed processing data set includes the following steps:

[0057] Obtain n processing orders and m processing equipment from the preprocessed processing data set, encode the production of n processing orders on m processing equipment and initialize the population, randomly generate the initial population, set the maximum number of iterations, population size, external archive set capacity, and local search mode change times

[0058] The initialization population is copied to generate a second population. The individuals of the initialization population and the second population are searched using the particle swarm algorithm and the gray wolf algorithm to generate a sub-population and calculate the corresponding objective function values ​​of the individuals.

[0059] When the number of local search mode changes has not been reached, a local search is performed on each individual. If the new individual found dominates the original individual, the new individual replaces the original individual. If the two do not dominate each other, the probability of whether to replace is 1 / 2.

[0060] The dominance relationship is used to select the non-inferior solution set among the offspring individuals, and the dominated individuals in the non-inferior solution set are deleted. The remaining individuals are placed in the external archive set. In the external archive set, the global optimal particle of the group is selected, and the α, β and δ wolves in the gray wolf search;

[0061] Repeat the iteration until the maximum number of iterations is reached, sort and decode, obtain the order of processing orders on the processing equipment, obtain the processing equipment number, and output the processing scheduling plan.

[0062] In the technical solution provided by the present invention, the metal sheet processing intelligent scheduling system includes a data collection module, a data preprocessing module, a solution output module, an evaluation and sorting module and a real-time monitoring module. The initial working state data of the processing equipment is collected in real time through sensors, and the processing order information of the metal sheet is collected. A processing data set is constructed based on the initial working state data and the processing order information; the data in the processing data set is preprocessed to obtain a preprocessed processing data set; an intelligent scheduling model is constructed, and the particle swarm algorithm and the gray wolf algorithm are used for co-evolution to optimize the parameters of the intelligent scheduling model, and a processing scheduling scheme is output through the intelligent scheduling model and the preprocessed processing data set; based on the preset processing indicators, the processing scheduling scheme is evaluated and sorted to obtain the optimal scheduling scheme, and an operation instruction is generated based on the optimal scheduling scheme, and the operation instruction is issued to the processing equipment of the metal sheet; during the processing of the processing equipment, the equipment status is monitored in real time. When the processing equipment has an abnormal situation, the processing equipment is adjusted based on the Internet of Things to cope with the abnormal situation until the processing order of the metal sheet is completed; the present invention significantly improves the efficiency of metal sheet processing, realizes the reasonable allocation and efficient utilization of various resources, avoids the waste of resources, and reduces the processing cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.

[0064] Figure 1 A schematic diagram of the structure of the intelligent scheduling system for metal sheet processing provided by an embodiment of the present invention;

[0065] Figure 2 A schematic diagram of the structure of a data preprocessing module provided in an embodiment of the present invention;

[0066] Figure 3 A schematic diagram of the structure of a solution output module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0067] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0068] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 The schematic diagram of the structure of the metal sheet processing intelligent scheduling system provided by the embodiment of the present invention includes a data collection module, a data preprocessing module, a solution output module, an evaluation and sorting module and a real-time monitoring module, wherein:

[0069] The data collection module 101 is used to collect the initial working state data of the processing equipment in real time through sensors, collect the processing order information of the metal sheet, and construct a processing data set based on the initial working state data and the processing order information;

[0070] The data preprocessing module 102 is used to preprocess the data in the processed data set to obtain a preprocessed processed data set;

[0071] The solution output module 103 is used to construct an intelligent scheduling model, and use the particle swarm algorithm and the gray wolf algorithm to optimize the parameters of the intelligent scheduling model, and output the processing scheduling solution through the intelligent scheduling model and the pre-processed processing data set;

[0072] An evaluation and sorting module 104 is used to evaluate and sort the processing scheduling schemes based on preset processing indicators to obtain an optimal scheduling scheme, generate operation instructions based on the optimal scheduling scheme, and send the operation instructions to the processing equipment of the metal sheet;

[0073] The real-time monitoring module 105 is used to monitor the equipment status in real time during the processing of the processing equipment. When there is an abnormal situation in the processing equipment, the processing equipment is adjusted based on the Internet of Things to deal with the abnormal situation until the processing order of the metal plate is completed.

[0074] In this embodiment, the evaluation and ranking module includes a setting submodule, a second calculation submodule and a generating instruction submodule, wherein:

[0075] The setting submodule is used to set the evaluation function according to the preset processing indicators:

[0076] F(x)=w1*C(x)+w2*T(x)+w3*Q(x)+w4*U(x)

[0077] Among them, F(x) is the comprehensive evaluation score, x represents a set of processing scheduling solutions, C(x) represents the sum of costs, T(x) represents the total processing time, Q(x) represents the quality index, U(x) represents the equipment utilization rate, w1, w 2、 w 3、 w4 are the weights of cost, time, quality, and equipment utilization;

[0078] The second calculation submodule is used to calculate the comprehensive evaluation score of all processing scheduling schemes based on the evaluation function, sort all processing scheduling schemes according to the comprehensive evaluation score, and select the scheme with the highest score as the optimal scheduling scheme;

[0079] The instruction generation submodule is used to generate operation instructions based on the optimal scheduling plan and send the operation instructions to the metal sheet processing equipment.

[0080] In this embodiment, the real-time monitoring module includes a data acquisition submodule, a health assessment submodule, a parameter adjustment submodule and a monitoring progress submodule, wherein:

[0081] The data acquisition submodule is used to connect the equipment sensors of all processing equipment to the Internet of Things platform, collect data at preset time intervals, and obtain equipment status data;

[0082] The health assessment submodule is used to assess the current health status of the processing equipment based on the equipment status data and output the corresponding equipment health level;

[0083] The parameter adjustment submodule is used to trigger the exception handling process based on the equipment health level, and the IoT monitoring automatically adjusts the parameters of the processing equipment through the actuator;

[0084] The progress monitoring submodule is used to generate an alarm message and notify the operator if the Internet of Things cannot adjust the parameters, and reselect the scheduling plan based on the processing scheduling plan. The Internet of Things monitors the processing progress of the metal sheet until the processing order of the metal sheet is completed.

[0085] In this embodiment, the health assessment submodule includes a filling unit, an initialization unit, an input unit, a training unit, an evaluation unit and a trigger unit, wherein:

[0086] A filling unit is used to obtain the health data samples of the metal sheet processing equipment, pre-process the acquired health data samples, remove invalid data points, and fill the missing data using the interpolation method to obtain a training set;

[0087] Initialization unit, used to initialize AdaBoost parameters, set the number of weak classifiers, and initialize the distribution weights of training samples in the training set;

[0088] The input unit is used to input the training set, use the multi-granularity scanning stage strategy to maximize the sequential relationship of data features, input the feature vector processed by the multi-granularity scanning stage, use the cascade forest stage to analyze and process the feature vector processed by the multi-granularity scanning stage layer by layer, average the classification vector generated by the last level, and take the classification result of the category corresponding to the maximum value;

[0089] The training unit is used to calculate the weight coefficient of the base classifier and update the sample distribution, train the weak classifier until the iteration round reaches the maximum value, and output the strong classifier;

[0090] An evaluation unit is used to input the equipment status data into a strong classifier, use the trained strong classifier to evaluate the current health status of the processing equipment, and output the corresponding equipment health level;

[0091] The trigger unit is used to trigger the exception handling process when the equipment health level is abnormal. The IoT monitoring automatically adjusts the parameters of the processing equipment through the actuator.

[0092] In this embodiment, the parameter adjustment submodule includes a healthy state unit, a sub-healthy state unit, a mild state unit and an abnormal state unit, wherein:

[0093] A health status unit is used for when the equipment health level is in a healthy state, the processing equipment indicators are normal, and the processing equipment works normally until the processing order of the metal sheet is completed;

[0094] Sub-health status unit, which is used to trigger the exception handling process when the equipment health level is in a sub-healthy state, and automatically adjust the parameters of the processing equipment until the processing order of the metal sheet is completed;

[0095] The minor status unit is used to strengthen the health monitoring of the processing equipment when the equipment health level is in a minor state and there is a minor abnormal fault in the processing equipment, and automatically adjust the parameters of the processing equipment until the processing order of the metal sheet is completed;

[0096] The abnormal status unit is used to stop the operation of the processing equipment when the equipment health level is in an abnormal state, generate an alarm message and notify the operator.

[0097] See also Figure 2 , a schematic diagram of the structure of a data preprocessing module in a metal sheet processing intelligent scheduling system provided by an embodiment of the present invention, wherein the data preprocessing module 102 includes a mean processing submodule 1021, a denoising processing submodule 1022, a deep optimization submodule 1023 and a classification submodule 1024, wherein,

[0098] The averaging processing submodule is used to obtain data in the processing data set and perform averaging processing on the time series data in the processing data set through discrete Fourier transform;

[0099] The denoising submodule is used to calculate the expected value of the energy of the actual data and noise data of the time series data in the processed data set after averaging. When the expected value of the energy of the actual data is greater than the expected value of the energy of the noise data, the current time series is maintained. Otherwise, the actual data will be submerged in the noise data, and the data in the processed data set is denoised.

[0100] The deep optimization submodule is used to input the processed data set after denoising into the two-layer neural network, extract the data feature information of the processed data set after denoising through the bidirectional LSTM layer, and perform deep optimization on the extracted data feature information;

[0101] The classification submodule is used to transfer the data feature information after deep optimization to the output layer, and transfer the data feature information to Softmax for classification to obtain the preprocessed processed data set.

[0102] See also Figure 3 , a schematic diagram of the structure of a solution output module in a metal sheet processing intelligent scheduling system provided by an embodiment of the present invention, wherein the solution output module 103 includes a parameter setting submodule 1031, a first calculation submodule 1032, a local search submodule 1033, a selection submodule 1034 and a decoding submodule 1035, wherein,

[0103] The parameter setting submodule is used to obtain n processing orders and m processing equipment from the preprocessed processing data set, encode the production of n processing orders on m processing equipment and initialize the population, randomly generate the initial population, set the maximum number of iterations, population size, external archive set capacity, and local search mode change times

[0104] The first calculation submodule is used to copy the initialization population to generate a second population, use the particle swarm algorithm and the gray wolf algorithm as search modes for the individuals of the initialization population and the second population, generate a sub-population, and calculate the corresponding objective function values ​​of the individuals of the sub-population;

[0105] The local search submodule is used to perform a local search on each individual when the number of local search mode changes has not been reached. If the new individual found dominates the original individual, the new individual will replace the original individual. If the two do not dominate each other, the probability of whether to replace is 1 / 2.

[0106] The selection submodule is used to select the non-inferior solution set from the offspring individuals by using the dominance relationship, delete the dominated individuals in the non-inferior solution set, and put the remaining individuals into the external archive set. In the external archive set, the global optimal particle of the group is selected, and the α, β and δ wolves in the gray wolf search;

[0107] The decoding submodule is used to repeat iterations until the maximum number of iterations is reached, sort and decode, obtain the sequence of processing orders on the processing equipment, obtain the processing equipment number, and output the processing scheduling plan.

[0108] Through the implementation of the above scheme, the metal sheet processing intelligent scheduling system includes a data collection module, a data preprocessing module, a scheme output module, an evaluation sorting module and a real-time monitoring module. The present invention significantly improves the efficiency of metal sheet processing, realizes the rational allocation and efficient utilization of various resources, avoids waste of resources, and reduces processing costs.

[0109] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An intelligent scheduling system for metal sheet processing, characterized in that: The metal sheet processing intelligent scheduling system includes a data collection module, a data preprocessing module, a solution output module, an evaluation and sorting module and a real-time monitoring module, wherein: A data collection module, used to collect initial working state data of processing equipment in real time through sensors, collect processing order information of metal sheets, and construct a processing data set based on the initial working state data and the processing order information; A data preprocessing module, used for preprocessing the data in the processed data set to obtain a preprocessed processed data set; A scheme output module is used to construct an intelligent scheduling model, and use the particle swarm algorithm and the gray wolf algorithm to co-evolve to optimize the parameters of the intelligent scheduling model, and output the processing scheduling scheme through the intelligent scheduling model and the pre-processed processing data set; An evaluation and sorting module is used to evaluate and sort the processing scheduling scheme based on preset processing indicators to obtain an optimal scheduling scheme, generate an operation instruction based on the optimal scheduling scheme, and send the operation instruction to the processing equipment of the metal sheet; The real-time monitoring module is used to monitor the equipment status in real time during the processing of the processing equipment. When there is an abnormality in the processing equipment, the processing equipment is adjusted based on the Internet of Things to deal with the abnormality until the processing order of the metal sheet is completed; The data preprocessing module includes a mean processing submodule, a denoising processing submodule, a deep optimization submodule and a classification submodule, wherein: A mean value processing submodule is used to obtain data in the processed data set and perform mean value processing on the time series data in the processed data set by discrete Fourier transform; A denoising processing submodule is used to calculate the expected value of the energy of the actual data and noise data of the time series data in the processed data set after the averaging processing. When the expected value of the energy of the actual data is greater than the expected value of the energy of the noise data, the current time series is maintained. Otherwise, the actual data will be submerged in the noise data, and the data in the processed data set is denoised. The deep optimization submodule is used to input the processed data set after denoising into the two-layer neural network, extract the data feature information of the processed data set after denoising through the bidirectional LSTM layer, and perform deep optimization on the extracted data feature information; The classification submodule is used to transfer the data feature information after deep optimization to the output layer, and transfer the data feature information to Softmax for classification to obtain the preprocessed processed data set; The scheme output module includes a parameter setting submodule, a first calculation submodule, a local search submodule, a selection submodule and a decoding submodule, wherein: The parameter setting submodule is used to obtain n processing orders and m processing equipment from the preprocessed processing data set, encode the production of n processing orders on m processing equipment and initialize the population, randomly generate the initial population, set the maximum number of iterations, population size, external archive set capacity, and local search mode change times The first calculation submodule is used to copy the initialization population to generate a second population, use the particle swarm algorithm and the gray wolf algorithm as search modes for the individuals of the initialization population and the second population, generate a sub-population, and calculate the corresponding objective function values ​​of the individuals of the sub-population; The local search submodule is used to perform a local search on each individual when the number of local search mode changes has not been reached. If the new individual found dominates the original individual, the new individual will replace the original individual. If the two do not dominate each other, the probability of whether to replace is 1 / 2. The selection submodule is used to select the non-inferior solution set from the offspring individuals by using the dominance relationship, delete the dominated individuals from the non-inferior solution set, and put the remaining individuals into the external archive set. In the external archive set, the global optimal particle of the group is selected, and the α, β and δ wolves in the gray wolf search; The decoding submodule is used to repeat the iteration until the maximum number of iterations is reached, sort and decode, obtain the order of processing orders on the processing equipment, obtain the processing equipment number, and output the processing scheduling plan; The evaluation and ranking module includes a setting submodule, a second calculation submodule and a generating instruction submodule, wherein: The setting submodule is used to set the evaluation function according to the preset processing indicators: F(x)=w1*C(x)+w2*T(x)+w3*Q(x)+w4*U(x) Among them, F(x) is the comprehensive evaluation score, x represents a set of processing scheduling solutions, C(x) represents the sum of costs, T(x) represents the total processing time, Q(x) represents the quality index, U(x) represents the equipment utilization rate, w1, w 2、 w 3、 w4 are the weights of cost, time, quality, and equipment utilization; A second calculation submodule is used to calculate a comprehensive evaluation score for all processing scheduling solutions based on the evaluation function, sort all processing scheduling solutions according to the comprehensive evaluation score, and select the solution with the highest score as the optimal scheduling solution; A command generation submodule, used for generating an operation command based on the optimal scheduling solution, and sending the operation command to the metal sheet processing equipment; The real-time monitoring module includes a data acquisition submodule, a health assessment submodule, a parameter adjustment submodule and a monitoring progress submodule, wherein: The data acquisition submodule is used to connect the equipment sensors of all processing equipment to the Internet of Things platform, collect data at preset time intervals, and obtain equipment status data; A health assessment submodule, used to assess the current health status of the processing equipment based on the equipment status data and output a corresponding equipment health level; A parameter adjustment submodule is used to trigger an exception handling process based on the health level of the equipment, and the IoT monitoring automatically adjusts the parameters of the processing equipment through the actuator; The progress monitoring submodule is used to generate an alarm message and notify the operator if the Internet of Things cannot adjust the parameters, and reselect the scheduling plan based on the processing scheduling plan. The Internet of Things monitors the processing progress of the metal sheet until the processing order of the metal sheet is completed.

2. The metal sheet processing intelligent scheduling system according to claim 1, characterized in that: The health assessment submodule includes a filling unit, an initialization unit, an input unit, a training unit, an assessment unit and a trigger unit, wherein: A filling unit is used to obtain the health data samples of the metal sheet processing equipment, pre-process the acquired health data samples, remove invalid data points, and fill the missing data using the interpolation method to obtain a training set; Initialization unit, used to initialize AdaBoost parameters, set the number of weak classifiers, and initialize the distribution weights of training samples in the training set; The input unit is used to input the training set, use the multi-granularity scanning stage strategy to maximize the sequential relationship of data features, input the feature vector processed by the multi-granularity scanning stage, use the cascade forest stage to analyze and process the feature vector processed by the multi-granularity scanning stage layer by layer, average the classification vector generated by the last level, and take the classification result of the category corresponding to the maximum value; The training unit is used to calculate the weight coefficient of the base classifier and update the sample distribution, train the weak classifier until the iteration round reaches the maximum value, and output the strong classifier; An evaluation unit, used to input the equipment status data into a strong classifier, evaluate the current health status of the processing equipment using the trained strong classifier, and output a corresponding equipment health level; The trigger unit is used to trigger the exception handling process when the equipment health level is abnormal. The IoT monitoring automatically adjusts the parameters of the processing equipment through the actuator.

3. The metal sheet processing intelligent scheduling system according to claim 1, characterized in that: The parameter adjustment submodule includes a healthy state unit, a sub-healthy state unit, a mild state unit and an abnormal state unit, wherein: A health status unit is used for when the equipment health level is in a healthy state, the processing equipment indicators are normal, and the processing equipment works normally until the processing order of the metal sheet is completed; Sub-health status unit, which is used to trigger the exception handling process when the equipment health level is in a sub-healthy state, and automatically adjust the parameters of the processing equipment until the processing order of the metal sheet is completed; The minor status unit is used to strengthen the health monitoring of the processing equipment when the equipment health level is in a minor state and there is a minor abnormal fault in the processing equipment, and automatically adjust the parameters of the processing equipment until the processing order of the metal sheet is completed; The abnormal status unit is used to stop the operation of the processing equipment when the equipment health level is in an abnormal state, generate an alarm message and notify the operator.

4. The metal sheet processing intelligent scheduling system according to claim 1, characterized in that: The method for realizing the intelligent scheduling system for metal sheet processing comprises the following steps: Collecting initial working state data of processing equipment in real time through sensors, collecting processing order information of metal plates, and constructing a processing data set based on the initial working state data and the processing order information; Preprocessing the data in the processing data set to obtain a preprocessed processing data set, constructing an intelligent scheduling model, and using a particle swarm algorithm and a gray wolf algorithm for co-evolution to optimize the parameters of the intelligent scheduling model, and outputting a processing scheduling plan through the intelligent scheduling model and the preprocessed processing data set; Based on the preset processing indicators, the processing scheduling schemes are evaluated and sorted to obtain the optimal scheduling scheme, an operation instruction is generated based on the optimal scheduling scheme, and the operation instruction is sent to the processing equipment of the metal sheet; During the processing of the processing equipment, the equipment status is monitored in real time. When there is an abnormal situation in the processing equipment, the processing equipment is adjusted based on the Internet of Things to deal with the abnormal situation until the processing order of the metal sheet is completed.

5. The metal sheet processing intelligent scheduling system according to claim 4, characterized in that: Preprocessing the data in the processed data set to obtain the preprocessed processed data set includes the following steps: Acquire data in the processed data set, and perform averaging processing on the time series data in the processed data set by discrete Fourier transform; Calculate the expected value of the energy of the actual data and noise data of the time series data in the processed data set after averaging processing. When the expected value of the energy of the actual data is greater than the expected value of the energy of the noise data, maintain the current time series. Otherwise, the actual data will be submerged in the noise data. Perform denoising processing on the data in the processed data set. The denoised processed data set is input into a two-layer neural network, and the data feature information of the denoised processed data set is extracted through a bidirectional LSTM layer, and the extracted data feature information is deeply optimized; The data feature information after deep optimization is transmitted to the output layer, and the data feature information is transmitted to Softmax for classification to obtain the preprocessed processed data set.

6. The metal sheet processing intelligent scheduling system according to claim 4, characterized in that: The construction of the intelligent scheduling model, and the use of particle swarm algorithm and gray wolf algorithm co-evolution to optimize the parameters of the intelligent scheduling model, and outputting the processing scheduling plan through the intelligent scheduling model and the pre-processed processing data set include the following steps: Obtain n processing orders and m processing equipment from the preprocessed processing data set, encode the production of n processing orders on m processing equipment and initialize the population, randomly generate the initial population, set the maximum number of iterations, population size, external archive set capacity, and local search mode change times The initialization population is copied to generate a second population. The individuals of the initialization population and the second population are searched using the particle swarm algorithm and the gray wolf algorithm to generate a sub-population and calculate the corresponding objective function values ​​of the individuals. When the number of local search mode changes has not been reached, a local search is performed on each individual. If the new individual found dominates the original individual, the new individual replaces the original individual. If the two do not dominate each other, the probability of whether to replace is 1 / 2. The dominance relationship is used to select the non-inferior solution set among the offspring individuals, and the dominated individuals in the non-inferior solution set are deleted. The remaining individuals are placed in the external archive set. In the external archive set, the global optimal particle of the group is selected, and the α, β and δ wolves in the gray wolf search; Repeat the iteration until the maximum number of iterations is reached, sort and decode, obtain the order of processing orders on the processing equipment, obtain the processing equipment number, and output the processing scheduling plan.