Online batching system and method for recycled aluminum based on global optimal algorithm

Abstract

The present invention discloses an online batching system for recycled aluminum based on a global optimal algorithm, including a server, a client, an industrial computer and a spectrum analyzer, wherein the server establishes a communication connection with the industrial computer, the industrial computer establishes a communication connection with the spectrum analyzer, and the client is installed on the industrial computer; the server performs the following operations: obtaining the first parameter information to be batched input by the user through the client; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value; obtaining a batching scheme based on the brand information and / or furnace group information; determining the type of raw materials and the quantity of each type of raw materials based on the brand target value and the batching scheme. The present invention manages different raw materials and proportions corresponding to the brand-associated furnace group, and automatically analyzes the batching data in combination with the finished product output, production energy consumption, input-output balance sheet, product quality, production detection, etc., to obtain the optimal batching list of different raw materials in the same furnace group, thereby reducing production costs.

Description

Technical Field

[0001] The present invention relates to a system and a method for using the same, and in particular to an online batching system and method for recycled aluminum based on a global optimal algorithm, belonging to the technical field of recycled aluminum. Background Art

[0002] Technical status: At present, most domestic recycled aluminum enterprises rely on experience to make batching, and finally rely entirely on the results of chemical composition analysis in front of the furnace for major adjustments. The process operation is very passive, and the slightest carelessness will result in substandard products or even waste products. In addition, the cost of the furnace charge is estimated manually. In order to comprehensively integrate various scrap aluminum conditions for relatively accurate batching, only considering the five basic control elements of No., Si, Cu, Zn, and Fe, the manual calculation amount is quite large. After further applicability sorting and classification, the incoming aluminum scrap is further distinguished in detail. Each type has different chemical composition and surface quality, and the cost value also changes further. The cost issue is the most concerned issue for recycled aluminum enterprises. Every link must be carefully considered, especially the first link of production. The burn-out of aluminum scrap is large, and the burn-out of different scrap aluminum also varies greatly. The burn-out rate of alloy elements also changes accordingly. These are the main areas where recycled aluminum batching calculations must be made.

[0003] The inventory balance of each raw material provided by the statisticians in the raw material warehouse to the batching personnel also has a great impact on the batching. At present, most recycled aluminum companies rely on manual statistics of the remaining raw materials to provide to the batching personnel. The batching personnel carry out batching based on the manually counted raw material data. The remaining quantity of raw materials cannot be updated in time, resulting in a large difference between the batching quantity on the batching list and the actual quantity in the warehouse. The batching personnel need to readjust the batching materials and quantities, which increases the workload of the batching personnel.

[0004] Therefore, there is an urgent need to improve the online batching system and method of recycled aluminum based on the global optimal algorithm to solve the above-mentioned problems. Summary of the invention

[0005] The purpose of the present invention is to provide an online batching system and method for recycled aluminum based on a global optimal algorithm. The system calculates the optimal theoretical cost, automatically analyzes the batching data in combination with finished product output, production time, production energy consumption, input-output balance sheet, product quality, production inspection, etc., and obtains the optimal batching list for different raw materials of the same brand and the same furnace group, thereby reducing production costs and maximizing profits.

[0006] In order to achieve the above object, the main technical solutions adopted by the present invention include:

[0007] A recycled aluminum online batching system based on a global optimal algorithm, comprising a server, a client, an industrial computer and a spectrum analyzer, wherein the server establishes a communication connection with the industrial computer, the industrial computer establishes a communication connection with the spectrum analyzer, and the client is installed on the industrial computer;

[0008] The server performs the following operations:

[0009] Acquiring first parameter information of ingredients to be mixed input by a user through the client; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value;

[0010] Based on the brand information and / or the furnace group information, obtaining a batching plan;

[0011] Based on the brand target value and the batching plan, determine the type of raw materials and the quantity of each type of raw materials;

[0012] The spectrum analyzer is used to analyze and detect the materials produced in the workshop;

[0013] The server is used to calculate and generate ingredient data from the spectrum analyzer;

[0014] The client is used to receive and display the ingredient data from the server;

[0015] The industrial computer is used to store the batching data from the server.

[0016] Preferably, the batching scheme includes a preset batching model or historical batching records; the batching model includes: brand name, customer name, type name, product form, specification, furnace group, material name, material type, percentage and single furnace, and the batching model is used for automatic batching; when the brand name and the furnace group are the same, the batching model can only be established once; the historical batching records are used to call out materials associated with the relevant brand name and filter out the materials that need to be used.

[0017] Preferably, the determining of the types of raw materials and the quantities of various types of raw materials based on the brand target value and the batching scheme includes: modeling the furnace group by the brand to form the brand furnace group target value and quantity, the brand corresponds to the furnace group, and the furnace group is a blast furnace melting furnace; selecting the corresponding blast furnace melting furnace of the furnace group, taking the brand target value as the automatic batching calculation result, recording it as the first feeding, the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation;

[0018] The furnace group is modeled by the brand to form the brand furnace group target value and quantity. The brand corresponds to the furnace group, and the furnace group is a low-furnace alloy furnace. The corresponding low-furnace alloy furnace of the furnace group is selected, and the brand target value is used as the automatic batching calculation result, recorded as the second feeding. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

[0019] A method for online batching of recycled aluminum based on a global optimal algorithm comprises the following steps:

[0020] S1: Obtaining first parameter information of ingredients to be mixed input by a user; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value;

[0021] S2: Based on the brand information and / or the furnace group information, obtaining a batching plan;

[0022] S3: Determine the types of raw materials and the quantities of each type of raw materials based on the brand target value and the batching plan;

[0023] S4: Model the furnace group according to the brand to form the brand furnace group target value and quantity, select the corresponding furnace group, and use the brand target value as the automatic batching calculation result. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

[0024] Preferably, in S4, the actual production information of the workshop is transmitted to the server, the server calculates the ingredients, the server converts the ingredients into a feeding list and transmits it to the client, the client displays the feeding information, and the forklift driver weighs the materials according to the feeding information displayed by the client and sends them to the production workshop.

[0025] Preferably, in S3, a spectrometer is used to analyze and detect the materials produced in the workshop, and the detection results are transmitted to the server. The server compares the detection results with the target values ​​of the ingredients and calculates the information of the required materials.

[0026] Preferably, in S4, the determining the types of raw materials and the quantities of various raw materials based on the brand target value and the batching scheme includes: modeling the furnace group by the brand to form the brand furnace group target value and quantity, selecting the corresponding furnace group blast furnace melting furnace, taking the brand target value as the automatic batching calculation result, recording it as the first feeding, the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation;

[0027] The furnace group is modeled by the brand to form the brand furnace group target value and quantity, the corresponding low-temperature alloy furnace of the furnace group is selected, and the brand target value is used as the automatic batching calculation result and recorded as the second feeding. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

[0028] Preferably, the brand target value is used as the automatic batching calculation result, recorded as N batches of materials, the N batches of materials are sampled and tested, and matched to the corresponding batching list according to the spectral analysis result information.

[0029] Preferably, according to S2: based on the brand information and / or the furnace group information, the specific working process of obtaining the batching scheme includes:

[0030] Acquire the sentence of the brand information and / or the furnace group information, and acquire the sentence composition structure corresponding to the sentence of the brand information and / or the furnace group information;

[0031] Determine the conversion rule from the original language to the standardized language according to the sentence composition structure;

[0032] Obtaining an abstract syntax tree of the original language through a preset lexical analyzer, extracting sentence structure content required by the standardized language from the abstract syntax tree of the original language, and storing it in a standardized language data set;

[0033] Based on the standard language data set and the conversion rule, the statements of the brand information and / or the furnace group information are standardized and output to obtain standardized statements;

[0034] Converting the normalized sentence into an input normalized sentence applicable to a semantic training model;

[0035] Importing the converted input normalized sentence into multiple semantic parsing neural network models, and combining the multiple semantic parsing neural network models in parallel into a fusion parsing model;

[0036] Obtaining a text sentence that reaches a preset positive example probability score threshold in the input normalized sentence output by the fusion parsing model, and using the text sentence as the final parsed sentence;

[0037] Based on the fusion parsing model, semantically parse the final parsed statement, and obtain the logical expression of the parsed standard statement;

[0038] Acquire corresponding text data based on the logical expression;

[0039] Labeling and defining the text data, obtaining the type label of the brand information and / or the furnace group information, and obtaining label data;

[0040] Searching a mapping relationship library, and obtaining ingredient data matching the label data, and establishing a matching relationship between the label data and the ingredient data according to a regular expression;

[0041] The matching relationship is the batching scheme.

[0042] Preferably, based on S4, after the furnace group is modeled by the brand, the method further includes:

[0043] The furnace group model after furnace group modeling is obtained, the furnace group model is simulated, and the simulation result is obtained. Based on the simulation result, the adaptive value of the furnace group model is calculated, and the optimization value of the furnace group model is calculated according to the adaptive value, and the furnace group model is optimized according to the optimization value. The specific working process includes:

[0044] Modeling the furnace group according to the brand and obtaining the furnace group model;

[0045] Simulating the furnace group model in a computer and obtaining simulation results;

[0046] Analyzing the simulation results to obtain a comprehensive simulation value of the furnace group model, and calculating an adaptive value of the furnace group model according to the comprehensive simulation value;

[0047]

[0048] Wherein, A represents the adaptive value of the furnace group model; η represents the value of the simulation parameter for simulating the furnace group model in the computer; t represents the simulation time required for simulating the furnace group model in the computer; f represents the simulation frequency required for simulating the furnace group model in the computer; m represents the capacity of the space ratio occupied by the furnace group model itself; Z represents the comprehensive simulation value of the furnace group model; q represents the probability of simulation deviation when simulating the furnace group model in the computer; V represents the simulation speed required for simulating the furnace group model in the computer;

[0049] Based on the computer, an ideal furnace group model is obtained, and an ideal adaptive value corresponding to the ideal furnace group model is obtained, and the ideal adaptive value is compared with the adaptive value of the furnace group model to obtain a comparison result;

[0050] Based on the comparison result, calculating the optimization value of the furnace group model;

[0051]

[0052] Wherein, Y represents the optimization value of the furnace group model; λ represents the optimization rate of the furnace group model; L represents the ideal adaptive value corresponding to the ideal furnace group model; P represents the brand target value; Q represents the amount of information of the furnace group model; A represents the adaptive value of the furnace group model; S represents the information entropy corresponding to the information of the furnace group model; represents the optimization coefficient;

[0053] Determining the total weight of the furnace group model when optimizing in a preset neural network;

[0054] Determine the weight of each layer of the furnace group model in the preset neural network based on the total weight;

[0055] According to the optimization value of the furnace group model and the weight value of each layer, the number of iterations of the furnace group model in the preset neural network is obtained;

[0056] Determining the optimal weight of the furnace group model in the preset neural network according to the number of iterations;

[0057] Based on the optimal weight and in combination with the brand, the furnace group model is optimized.

[0058] Through the above technical solution, the burn-off value of each element is managed digitally. When the recycled aluminum raw material contains the corresponding element, the system automatically participates in the calculation. The content of each element in different raw materials is inconsistent. One grade of batching involves multiple materials. Manually considering the burn-off value of each material corresponding to different elements is large and difficult to complete accurate calculations manually. Later, it is necessary to adjust the alloy furnace with pure metal, which is costly. During the batching process, the system provides data through the burn-off value of each element to accurately calculate the burn-off value of each element, reduce the pure metal ratio of secondary batching, and thus reduce costs.

[0059] The present invention has at least the following beneficial effects:

[0060] 1. Automatic batching data is combined with finished product output, production time, production energy consumption, input-output balance sheet, product quality, production inspection, etc. for data analysis to obtain the optimal batching list, thereby reducing production costs and maximizing profits.

[0061] 2. Manage the burnout value of each element in a digital way. When the recycled aluminum raw materials contain the corresponding elements, the system automatically participates in the calculation. During the batching process, the system provides data through the "burnout value of each element" to accurately calculate the burnout value of each element, thereby reducing the proportion of pure metal in the secondary batching and reducing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0063] Figure 1 The electrical schematic diagram of an online batching system for recycled aluminum based on a global optimal algorithm of the present invention;

[0064] Figure 2 It is a schematic flow chart of an online batching method for recycled aluminum based on a global optimal algorithm of the present invention;

[0065] Reference numerals: server 1 , client 2 , industrial computer 3 and spectrum analyzer 4 . DETAILED DESCRIPTION

[0066] The following will describe the implementation methods of the present application in detail with the help of accompanying drawings and examples, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0067] like Figure 1 and Figure 2 As shown, the embodiment provides an online batching system for recycled aluminum based on a global optimal algorithm, comprising a server 1, a client 2, an industrial computer 3 and a spectrum analyzer 4, wherein the server establishes a communication connection with the industrial computer, the industrial computer establishes a communication connection with the spectrum analyzer, and the client is installed on the industrial computer;

[0068] The server performs the following operations:

[0069] Acquiring first parameter information of ingredients to be mixed input by a user through the client; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value;

[0070] Based on the brand information and / or the furnace group information, obtaining a batching plan;

[0071] Based on the brand target value and the batching plan, determine the type of raw materials and the quantity of each type of raw materials;

[0072] The spectrum analyzer is used to analyze and detect the materials produced in the workshop;

[0073] The server is used to calculate and generate ingredient data from the spectrum analyzer;

[0074] The client is used to receive and display the ingredient data from the server;

[0075] The industrial computer is used to store the batching data from the server.

[0076] The batching plan includes a preset batching model or historical batching records; the batching model includes: brand name, customer name, type name, product form, specification, furnace group, material name, material type, percentage and single furnace, and the batching model is used for automatic batching; when the brand name and the furnace group are the same, the batching model can only be established once; the historical batching records are used to call out materials associated with the relevant brand names and screen out the materials that need to be used.

[0077] The method of determining the types of raw materials and the quantities of various raw materials based on the brand target value and the batching scheme includes: modeling the furnace group by the brand to form the brand furnace group target value and quantity, wherein the brand corresponds to the furnace group, and the furnace group is a blast furnace melting furnace; selecting the corresponding blast furnace melting furnace of the furnace group, taking the brand target value as the automatic batching calculation result, recording it as the first feeding, and the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation;

[0078] The furnace group is modeled by the brand to form the brand furnace group target value and quantity. The brand corresponds to the furnace group, and the furnace group is a low-furnace alloy furnace. The corresponding low-furnace alloy furnace of the furnace group is selected, and the brand target value is used as the automatic batching calculation result, recorded as the second feeding. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

[0079] A method for online batching of recycled aluminum based on a global optimal algorithm is also provided, comprising the following steps:

[0080] S1: Obtaining first parameter information of ingredients to be mixed input by a user; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value;

[0081] S2: Based on the brand information and / or the furnace group information, obtaining a batching plan;

[0082] S3: Determine the types of raw materials and the quantities of each type of raw materials based on the brand target value and the batching plan;

[0083] S4: Model the furnace group according to the brand to form the brand furnace group target value and quantity, select the corresponding furnace group, and use the brand target value as the automatic batching calculation result. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

[0084] In S4, the actual production information of the workshop is transmitted to the server, the server calculates the ingredients, the server converts the ingredients into a feeding list and transmits it to the client, the client displays the feeding information, and the forklift driver weighs the materials according to the feeding information displayed by the client and sends them to the production workshop.

[0085] In S3, a spectrometer is used to analyze and test the materials produced in the workshop, and the test results are transmitted to the server. The server compares the test results with the target values ​​of the ingredients and calculates the information of the required materials.

[0086] In S4, based on the brand target value and the batching scheme, the types of raw materials and the quantities of various raw materials are determined, including: modeling the furnace group by the brand to form the brand furnace group target value and quantity, selecting the corresponding furnace group blast furnace melting furnace, taking the brand target value as the automatic batching calculation result, recording it as the first feeding, the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation;

[0087] The furnace group is modeled by the brand to form the brand furnace group target value and quantity, the corresponding low-temperature alloy furnace of the furnace group is selected, and the brand target value is used as the automatic batching calculation result and recorded as the second feeding. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

[0088] The brand target value is used as the automatic batching calculation result, recorded as N batches of materials, and the N batches of materials are sampled and tested, and matched to the corresponding batching list according to the spectral analysis result information.

[0089] The system is modeled based on the batching target value. The batching target value is associated with the furnace group corresponding to different raw materials and proportions with the optimal theoretical cost. Manual batching needs to consider the use of different batches of the same material, which has a large calculation workload and cannot reach the optimal theoretical cost. Automatic batching data is combined with finished product output, production time, production energy consumption, input-output balance sheet, product quality, production inspection, etc. for data analysis to obtain the optimal batching list for different raw materials in the same furnace group with the same batching target value, thereby reducing production costs and maximizing profits.

[0090] Recycled ingredient target value modeling, based on expert ingredient knowledge base, digitized brand ingredient modeling, and digitized management, reduce dependence on ingredient personnel and reduce personnel costs.

[0091] The burnout value of each element is managed digitally. When the recycled aluminum raw materials contain the corresponding elements, the system automatically calculates the burnout value of each element. During the batching process, the system provides data through the "burnout value of each element" to accurately calculate the burnout value of each element, reduce the proportion of pure metal in the secondary batching, and thus reduce costs.

[0092] like Figure 1 and Figure 2 As shown, the working steps of the server system of the online batching system for recycled aluminum based on the global optimal algorithm provided in this embodiment are as follows:

[0093] Brand step-by-step ingredients:

[0094] Brand blast furnace melting furnace target value brand to furnace group modeling and corresponding number brand to furnace group modeling;

[0095] The target value of the brand low-alloy furnace is improved in the brand modeling, and the corresponding number of brands and corresponding furnace group modeling;

[0096] Select the corresponding blast furnace in the batching main page, and use the brand target value as the automatic batching calculation result to feed the first batch. The system automatically calculates the required raw materials and the user adjusts the materials according to the actual situation.

[0097] Select the corresponding furnace group low furnace on the batching main page, and use the brand target value as the automatic batching calculation result for the second batching. The system automatically calculates the required raw materials and the user adjusts the materials according to the actual situation.

[0098] After several steps of batching, a batching list that theoretically meets the internal control values ​​of the brand is finally formed to guide workshop production.

[0099] Brand step-by-step ingredients:

[0100] Brand blast furnace melting furnace target value brand modeling of furnace group corresponding quantity brand corresponding furnace group modeling;

[0101] Select the corresponding furnace group on the batching main page, and use the brand target value as the automatic batching calculation result for the first batching. The system automatically calculates the required raw materials and the user adjusts the materials based on the actual situation.

[0102] According to the actual production situation of the workshop, the ingredients are converted into feeding lists, and the client system displays the feeding information. The forklift driver weighs the materials and sends them out of the warehouse according to the material information on the feeding list. The quality control department conducts sampling inspection on the first feeding. The system background matches the furnace number, brand number, number of tests, etc. according to the spectral analysis results, and the corresponding feeding list is matched. The spectrum analyzer results are compared with the brand target value. The system automatically calculates which materials are needed and provides them to the batching personnel for reference to select the furnace group corresponding to the brand for modeling. The quality control department conducts sampling inspection on the Nth feeding. The system background matches the furnace number, brand number, number of tests, etc. according to the spectral analysis results, and the spectrum analyzer is compared with the brand target value. The system automatically calculates which materials are needed and provides them to the batching personnel for reference and selection. The system background automatically saves each test result and the corresponding materials added.

[0103] Entering target values ​​for ingredients into the server system;

[0104] Feeding materials to workshop production; the actual production information of the workshop is transmitted to the server system, the server system calculates the ingredients, the server system converts the ingredients into a feeding list and transmits it to the client system, the client system displays the feeding information, and the forklift driver weighs the materials and sends them to the production workshop according to the feeding information displayed by the client system;

[0105] Carry out sampling inspection of workshop production materials, compare the sampling inspection results with the batching target value, and the server system automatically calculates the required material information based on the comparison results; the sampling inspection uses a spectrometer to analyze and inspect the workshop production materials, and the inspection results are transmitted to the server system. The server system compares the inspection results with the batching target value and calculates the required material information;

[0106] Re-feed materials to the workshop production according to the required material information; the server system converts the required material information into a material feeding list and transmits it to the client system. The forklift driver weighs the materials and sends them to the production workshop according to the material feeding information displayed on the client system;

[0107] Repeat the sampling and testing of workshop production materials and feed until the sampling and testing structure meets the batching target value;

[0108] The server system saves data; the server system saves data, and the server system saves each test result and the corresponding added material data to the data storage module.

[0109] The present invention provides an online batching method for recycled aluminum based on a global optimal algorithm. According to S2: based on the grade information and / or the furnace group information, the specific working process of obtaining the batching scheme includes:

[0110] Acquire the sentence of the brand information and / or the furnace group information, and acquire the sentence composition structure corresponding to the sentence of the brand information and / or the furnace group information;

[0111] Determine the conversion rule from the original language to the standardized language according to the sentence composition structure;

[0112] Obtaining an abstract syntax tree of the original language through a preset lexical analyzer, extracting sentence structure content required by the standardized language from the abstract syntax tree of the original language, and storing it in a standardized language data set;

[0113] Based on the standard language data set and the conversion rule, the statements of the brand information and / or the furnace group information are standardized and output to obtain standardized statements;

[0114] Converting the normalized sentence into an input normalized sentence applicable to a semantic training model;

[0115] Importing the converted input normalized sentence into multiple semantic parsing neural network models, and combining the multiple semantic parsing neural network models in parallel into a fusion parsing model;

[0116] Obtaining a text sentence that reaches a preset positive example probability score threshold in the input normalized sentence output by the fusion parsing model, and using the text sentence as the final parsed sentence;

[0117] Based on the fusion parsing model, semantically parse the final parsed statement, and obtain the logical expression of the parsed standard statement;

[0118] Acquire corresponding text data based on the logical expression;

[0119] Labeling and defining the text data, obtaining the type label of the brand information and / or the furnace group information, and obtaining label data;

[0120] Searching a mapping relationship library, and obtaining ingredient data matching the label data, and establishing a matching relationship between the label data and the ingredient data according to a regular expression;

[0121] The matching relationship is the batching scheme.

[0122] In this embodiment, the sentence composition structure may be composed of sentence identification characters, symbols, and main content of the sentence.

[0123] In this embodiment, the original language may be sentences of brand information and / or furnace group information, and is not processed.

[0124] In this embodiment, the sentence structure content may include identification data of the sentence and content composed of the sentence composition structure.

[0125] The beneficial effects of the above technical solution are:

[0126] By parsing the statements of the brand information and / or the furnace group information, the corresponding parsing data is obtained, and by matching the parsing data with the batching data, the batching scheme can be accurately obtained, thereby greatly improving the efficiency and accuracy of the present invention.

[0127] The present invention provides an online batching method for recycled aluminum based on a global optimal algorithm, which further comprises:

[0128] The furnace group model after furnace group modeling is obtained, the furnace group model is simulated, and the simulation result is obtained. Based on the simulation result, the adaptive value of the furnace group model is calculated, and the optimization value of the furnace group model is calculated according to the adaptive value, and the furnace group model is optimized according to the optimization value. The specific working process includes:

[0129] Modeling the furnace group according to the brand and obtaining the furnace group model;

[0130] Simulating the furnace group model in a computer and obtaining simulation results;

[0131] Analyzing the simulation results to obtain a comprehensive simulation value of the furnace group model, and calculating an adaptive value of the furnace group model according to the comprehensive simulation value;

[0132]

[0133] Wherein, A represents the adaptive value of the furnace group model; η represents the value of the simulation parameter for simulating the furnace group model in the computer; t represents the simulation time required for simulating the furnace group model in the computer; f represents the simulation frequency required for simulating the furnace group model in the computer; m represents the capacity of the space ratio occupied by the furnace group model itself; Z represents the comprehensive simulation value of the furnace group model; q represents the probability of simulation deviation when simulating the furnace group model in the computer; V represents the simulation speed required for simulating the furnace group model in the computer;

[0134] Based on the computer, an ideal furnace group model is obtained, and an ideal adaptive value corresponding to the ideal furnace group model is obtained, and the ideal adaptive value is compared with the adaptive value of the furnace group model to obtain a comparison result;

[0135] Based on the comparison result, calculating the optimization value of the furnace group model;

[0136]

[0137] Wherein, Y represents the optimization value of the furnace group model; λ represents the optimization rate of the furnace group model; L represents the ideal adaptive value corresponding to the ideal furnace group model; P represents the brand target value; Q represents the amount of information of the furnace group model; A represents the adaptive value of the furnace group model; S represents the information entropy corresponding to the information of the furnace group model; represents the optimization coefficient;

[0138] Determining the total weight of the furnace group model when optimizing in a preset neural network;

[0139] Determine the weight of each layer of the furnace group model in the preset neural network based on the total weight;

[0140] According to the optimization value of the furnace group model and the weight value of each layer, the number of iterations of the furnace group model in the preset neural network is obtained;

[0141] Determining the optimal weight of the furnace group model in the preset neural network according to the number of iterations;

[0142] Based on the optimal weight and in combination with the brand, the furnace group model is optimized.

[0143] In this embodiment, the adaptive value of the furnace group model may be the added value of the material, the inventory balance, etc., which are automatically adjusted according to the working degree of the furnace group model when the furnace group model is performing simulation.

[0144] In this embodiment, the comprehensive simulation value of the furnace group model may be the sum of the material value and the inventory balance value that need to be added to the furnace group model when the furnace group model is simulated.

[0145] In this embodiment, the optimization value may be an optimization value determined based on the difference between the ideal adaptive value and the adaptive value of the furnace group model, and its purpose is to better adjust the adaptive value of the furnace group model, thereby optimizing the furnace group model.

[0146] In this embodiment, the information of the furnace group model includes: inventory balance, raw materials, burn-down values ​​of various elements, etc.; the amount of information of the furnace group model is to quantify the information of the furnace group model; for example, the value of inventory balance is defined as 1, the value of raw materials is defined as 2, the burn-down value of various elements is defined as 3, etc. When a certain value is missing from the data, the information of the furnace group model changes, and the corresponding information value also changes accordingly.

[0147] The beneficial effects of the above technical solution are:

[0148] By obtaining the furnace group model after furnace group modeling and simulating the furnace group model, the adaptive value of the furnace group model can be accurately calculated according to the simulation results, and the optimization value of the furnace group model can be accurately calculated according to the adaptive value. The furnace group model is accurately optimized through the optimization value, thereby greatly improving the durability and effectiveness of the furnace group model.

[0149] For example, certain words are used in the specification and claims to refer to specific components. Those skilled in the art should understand that hardware manufacturers may use different terms to refer to the same component. This specification and claims do not use differences in names as a way to distinguish components, but use differences in the functions of components as the criteria for distinction. For example, "including" mentioned throughout the specification and claims is an open term, so it should be interpreted as "including but not limited to". "Approximately" means that within an acceptable error range, those skilled in the art can solve technical problems within a certain error range and basically achieve technical effects.

[0150] It should be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a product or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such product or system. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the product or system including the elements.

[0151] The above description shows and describes several preferred embodiments of the present invention, but as mentioned above, it should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the invention concept described herein through the above teachings or the technology or knowledge of the relevant field. Changes and variations made by those skilled in the art do not depart from the spirit and scope of the present invention, and should be within the scope of protection of the claims attached to the present invention.

Claims

1. An online batching system for recycled aluminum based on a global optimal algorithm, characterized in that: It comprises a server (1), a client (2), an industrial computer (3) and a spectrum analyzer (4), wherein the server establishes a communication connection with the industrial computer, the industrial computer establishes a communication connection with the spectrum analyzer, and the client is installed on the industrial computer; The server performs the following operations: Acquiring, through the client, first parameter information of the ingredient to be prepared input by the user; The first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value; Based on the brand information and / or the furnace group information, obtaining a batching plan; Based on the brand target value and the batching plan, determine the type of raw materials and the quantity of each type of raw materials; The spectrum analyzer is used to analyze and detect the materials produced in the workshop; The server is used to calculate and generate ingredient data from the spectrum analyzer; The client is used to receive and display the ingredient data from the server; The industrial computer is used to store the batching data from the server; Based on the brand information and / or the furnace group information, the specific working process of obtaining the batching scheme includes: Acquire the sentence of the brand information and / or the furnace group information, and acquire the sentence composition structure corresponding to the sentence of the brand information and / or the furnace group information; Determine the conversion rule from the original language to the standardized language according to the sentence composition structure; Obtaining an abstract syntax tree of the original language through a preset lexical analyzer, extracting sentence structure content required by the standardized language from the abstract syntax tree of the original language, and storing it in a standardized language data set; Based on the standard language data set and the conversion rule, the statements of the brand information and / or the furnace group information are standardized and output to obtain standardized statements; Converting the normalized sentence into an input normalized sentence applicable to a semantic training model; Importing the converted input normalized sentence into multiple semantic parsing neural network models, and combining the multiple semantic parsing neural network models in parallel into a fusion parsing model; Obtaining a text sentence that reaches a preset positive example probability score threshold in the input normalized sentence output by the fusion parsing model, and using the text sentence as the final parsed sentence; Based on the fusion parsing model, semantically parse the final parsed statement, and obtain the logical expression of the parsed normalized statement; Acquire corresponding text data based on the logical expression; Labeling and defining the text data, obtaining the type label of the brand information and / or the furnace group information, and obtaining label data; Searching a mapping relationship library, and obtaining ingredient data matching the label data, and establishing a matching relationship between the label data and the ingredient data according to a regular expression; Wherein, the matching relationship is the batching scheme; The method of determining the types of raw materials and the quantities of various raw materials based on the brand target value and the batching scheme includes: modeling the furnace group by the brand to form the brand furnace group target value and quantity, wherein the brand corresponds to the furnace group, and the furnace group is a blast furnace melting furnace; selecting the corresponding blast furnace melting furnace of the furnace group, taking the brand target value as the automatic batching calculation result, recording it as the first feeding, and the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation; The furnace group is modeled by the brand to form a brand furnace group target value and quantity, the brand corresponds to the furnace group, and the furnace group is a low-furnace alloy furnace; the corresponding low-furnace alloy furnace of the furnace group is selected, and the brand target value is used as the automatic batching calculation result, recorded as the second feeding, and the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation; After the furnace group is modeled by the brand, it also includes: The furnace group model after furnace group modeling is obtained, the furnace group model is simulated, and the simulation result is obtained. Based on the simulation result, the adaptive value of the furnace group model is calculated, and the optimization value of the furnace group model is calculated according to the adaptive value, and the furnace group model is optimized according to the optimization value. The specific working process includes: Modeling the furnace group according to the brand and obtaining the furnace group model; Simulating the furnace group model in a computer and obtaining simulation results; Analyzing the simulation results to obtain a comprehensive simulation value of the furnace group model, and calculating an adaptive value of the furnace group model according to the comprehensive simulation value; Wherein, A represents the adaptive value of the furnace group model; η represents the value of the simulation parameter for simulating the furnace group model in the computer; t represents the simulation time required for simulating the furnace group model in the computer; f represents the simulation frequency required for simulating the furnace group model in the computer; m represents the capacity of the space ratio occupied by the furnace group model itself; Z represents the comprehensive simulation value of the furnace group model; q represents the probability of simulation deviation when simulating the furnace group model in the computer; V represents the simulation speed required for simulating the furnace group model in the computer; Based on the computer, an ideal furnace group model is obtained, and an ideal adaptive value corresponding to the ideal furnace group model is obtained, and the ideal adaptive value is compared with the adaptive value of the furnace group model to obtain a comparison result; Based on the comparison result, calculating the optimization value of the furnace group model; Wherein, Y represents the optimization value of the furnace group model; λ represents the optimization rate of the furnace group model; L represents the ideal adaptive value corresponding to the ideal furnace group model; P represents the brand target value; Q represents the amount of information of the furnace group model; A represents the adaptive value of the furnace group model; S represents the information entropy corresponding to the information of the furnace group model; represents the optimization coefficient; Determining the total weight of the furnace group model when optimizing in a preset neural network; Determine the weight of each layer of the furnace group model in the preset neural network based on the total weight; According to the optimization value of the furnace group model and the weight value of each layer, the number of iterations of the furnace group model in the preset neural network is obtained; Determining the optimal weight of the furnace group model in the preset neural network according to the number of iterations; Based on the optimal weight and in combination with the brand, the furnace group model is optimized.

2. The online batching system for recycled aluminum based on the global optimal algorithm according to claim 1 is characterized in that: The batching plan includes a preset batching model or historical batching records; the batching model includes: brand name, customer name, type name, product form, specification, furnace group, material name, material type, percentage and single furnace, and the batching model is used for automatic batching; when the brand name and the furnace group are the same, the batching model can only be established once; the historical batching records are used to call out materials associated with the relevant brand names and screen out the materials that need to be used.

3. An online batching method for recycled aluminum based on a global optimal algorithm, characterized in that: Includes the following step: S1: Obtaining first parameter information of ingredients to be mixed input by a user; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value; S2: Based on the brand information and / or the furnace group information, obtaining a batching plan; S3: Determine the types of raw materials and the quantities of each type of raw materials based on the brand target value and the batching plan; S4: Model the furnace group according to the brand, form the brand furnace group target value and quantity, select the corresponding furnace group, use the brand target value as the automatic batching calculation result, the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation; According to S2: based on the brand information and / or the furnace group information, the specific working process of obtaining the batching scheme includes: Acquire the sentence of the brand information and / or the furnace group information, and acquire the sentence composition structure corresponding to the sentence of the brand information and / or the furnace group information; Determine the conversion rule from the original language to the standardized language according to the sentence composition structure; Obtaining an abstract syntax tree of the original language through a preset lexical analyzer, extracting sentence structure content required by the standardized language from the abstract syntax tree of the original language, and storing it in a standardized language data set; Based on the standard language data set and the conversion rule, the statements of the brand information and / or the furnace group information are standardized and output to obtain standardized statements; Converting the normalized sentence into an input normalized sentence applicable to a semantic training model; Importing the converted input normalized sentence into multiple semantic parsing neural network models, and combining the multiple semantic parsing neural network models in parallel into a fusion parsing model; Obtaining a text sentence that reaches a preset positive example probability score threshold in the input normalized sentence output by the fusion parsing model, and using the text sentence as the final parsed sentence; Based on the fusion parsing model, semantically parse the final parsed statement, and obtain the logical expression of the parsed normalized statement; Acquire corresponding text data based on the logical expression; Labeling and defining the text data, obtaining the type label of the brand information and / or the furnace group information, and obtaining label data; Searching a mapping relationship library, and obtaining ingredient data matching the label data, and establishing a matching relationship between the label data and the ingredient data according to a regular expression; Wherein, the matching relationship is the batching scheme; Based on the step S4, after the furnace group is modeled by the brand, the step further includes: The furnace group model after furnace group modeling is obtained, the furnace group model is simulated, and the simulation result is obtained. Based on the simulation result, the adaptive value of the furnace group model is calculated, and the optimization value of the furnace group model is calculated according to the adaptive value, and the furnace group model is optimized according to the optimization value. The specific working process includes: Modeling the furnace group according to the brand and obtaining the furnace group model; Simulating the furnace group model in a computer and obtaining simulation results; Analyzing the simulation results to obtain a comprehensive simulation value of the furnace group model, and calculating an adaptive value of the furnace group model according to the comprehensive simulation value; Wherein, A represents the adaptive value of the furnace group model; η represents the value of the simulation parameter for simulating the furnace group model in the computer; t represents the simulation time required for simulating the furnace group model in the computer; f represents the simulation frequency required for simulating the furnace group model in the computer; m represents the capacity of the space ratio occupied by the furnace group model itself; Z represents the comprehensive simulation value of the furnace group model; q represents the probability of simulation deviation when simulating the furnace group model in the computer; V represents the simulation speed required for simulating the furnace group model in the computer; Based on the computer, an ideal furnace group model is obtained, and an ideal adaptive value corresponding to the ideal furnace group model is obtained, and the ideal adaptive value is compared with the adaptive value of the furnace group model to obtain a comparison result; Based on the comparison result, calculating the optimization value of the furnace group model; Wherein, Y represents the optimization value of the furnace group model; λ represents the optimization rate of the furnace group model; L represents the ideal adaptive value corresponding to the ideal furnace group model; P represents the brand target value; Q represents the amount of information of the furnace group model; A represents the adaptive value of the furnace group model; S represents the information entropy corresponding to the information of the furnace group model; represents the optimization coefficient; Determining the total weight of the furnace group model when optimizing in a preset neural network; Determine the weight of each layer of the furnace group model in the preset neural network based on the total weight; According to the optimization value of the furnace group model and the weight value of each layer, the number of iterations of the furnace group model in the preset neural network is obtained; Determining the optimal weight of the furnace group model in the preset neural network according to the number of iterations; Based on the optimal weight and in combination with the brand, the furnace group model is optimized.

4. The online batching method for recycled aluminum based on the global optimal algorithm according to claim 3 is characterized in that: In S4, the actual production information of the workshop is transmitted to the server, the server calculates the ingredients, the server converts the ingredients into a feeding list and transmits it to the client, the client displays the feeding information, and the forklift driver weighs the materials according to the feeding information displayed by the client and sends them to the production workshop.

5. The online batching method for recycled aluminum based on the global optimal algorithm according to claim 3 is characterized in that: In S3, a spectrometer is used to analyze and test the materials produced in the workshop, and the test results are transmitted to the server. The server compares the test results with the target values ​​of the ingredients and calculates the information of the required materials.

6. The online batching method for recycled aluminum based on the global optimal algorithm according to claim 3 is characterized in that: In S4, based on the brand target value and the batching scheme, the types of raw materials and the quantities of various raw materials are determined, including: modeling the furnace group by the brand to form the brand furnace group target value and quantity, selecting the corresponding furnace group blast furnace melting furnace, taking the brand target value as the automatic batching calculation result, recording it as the first feeding, the recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation; The furnace group is modeled by the brand to form the brand furnace group target value and quantity, the corresponding low-temperature alloy furnace of the furnace group is selected, and the brand target value is used as the automatic batching calculation result and recorded as the second feeding. The recycled aluminum online batching system automatically calculates the required raw materials, and the user adjusts the materials according to the actual situation.

7. The online batching method for recycled aluminum based on the global optimal algorithm according to claim 6 is characterized in that: The brand target value is used as the automatic batching calculation result, recorded as N batches of materials, and the N batches of materials are sampled and tested, and matched to the corresponding batching list according to the spectral analysis result information.

Images