A raw material quality filtering method and system

Abstract

The embodiment of the present invention discloses a raw material quality filtering method and system. It belongs to the field of material quality filtering, and includes: obtaining product quality data of raw materials; inputting the product quality data into a pre-established quality filtering model to obtain a dynamic raw material production process that matches the product quality data; and using the dynamic raw material production process to produce the raw materials. The present invention does not require narrow range control of raw materials and subsequent production processes, and while eliminating quality volatility, improves the yield rate and reduces production costs.

Description

Technical Field

[0001] The invention belongs to the technical field of material quality filtering, and in particular relates to a raw material quality filtering method and system. Background Art

[0002] The most direct external manifestation of product quality is whether it is qualified or not. China's manufacturing industry, especially high-end manufacturing industry, has entered the stage of pursuing product quality from pursuing qualified rate. An important part of product quality is the stability of quality. This is particularly prominent in the field of metal materials represented by steel.

[0003] The volatility of product quality is one of the problems that plague the industrial products sector. The volatility of raw material quality is an important factor in the subsequent quality fluctuations of parts and equipment products. Take steel as an example. After the qualified rate of steel products is close to 100%, steel companies still have a large number of quality disputes. Except for a small number of special reasons, most of them can be attributed to the quality fluctuations of qualified products, which leads to a series of usage problems in subsequent applications such as welding and processing.

[0004] From the perspective of the production process of raw materials, the quality fluctuations of metal materials represented by steel products are often caused by fluctuations in raw material quality and process control itself. The coupling of the two leads to fluctuations in the quality of the final product.

[0005] For the production of metal raw materials such as steel, the core process of its production process is a combination of multiple links in "blank smelting-casting-forging / rolling-heat treatment-nondestructive testing-surface treatment", and the reasons for its volatility include but are not limited to the inevitable quality volatility of the smelted billet itself, such as the fluctuation of the billet alloy composition factors. In the subsequent forging / rolling and other production processes, the traditional method basically adopts a static production process, but is limited by the equipment control capability, and basically cannot avoid the volatility of the control of production process parameters such as forging / rolling. Therefore, when the volatility of the billet is coupled with the process, the quality of the finished metal material is the inheritance of the quality fluctuation of the aforementioned smelting process and the further amplification in the forging / rolling process, which will directly lead to qualified raw materials + qualified processes outputting unqualified products or qualified products with large quality deviations.

[0006] For parts and equipment made of metal materials, during the manufacturing process, metal materials are used as raw materials for parts or equipment, and are formed into parts or equipment through processes such as deformation and welding. When the quality of metal materials products fluctuates, the metal materials products are coupled with the fluctuations in the manufacturing process of parts or equipment, which will also cause the same type of quality fluctuations of parts and equipment products to be amplified, resulting in qualified raw materials + qualified processes, but producing unqualified parts or equipment with large quality deviations.

[0007] The quality of industrial products, with quality stability as the core, is one of the focuses of the current manufacturing industry, especially the steel industry. The application of "narrow range" control technology usually results in sacrificing the yield of steel products, which will inevitably cause a substantial increase in the cost of steel products. For steel products, especially for "large-volume and wide-ranging" products represented by rebar, threaded steel, and coils, cost is an important consideration for competitiveness, and "narrow range" control is almost difficult to implement.

[0008] In recent years, the steel industry has proposed that a narrow process range control method can be used to solve the quality stability problem. For example, in the production process of steel materials, the composition of the raw material for steel plate rolling, the slab, is controlled within a narrow range, and the process range of subsequent rolling, heat treatment and other production processes is also controlled within a narrow range to produce steel products with better quality stability. This technology is difficult, and different control requirements are designed for different products, resulting in low yields and increased production costs. Summary of the invention

[0009] In order to solve the problem in the prior art that different designs are made for different control requirements of products, resulting in low yield and increased production cost of products, the embodiment of the present invention provides the following technical solutions:

[0010] In one aspect, a raw material quality filtering method comprises:

[0011] Obtain product quality data of raw materials;

[0012] Inputting the product quality data into a pre-established quality filtering model to obtain a dynamic raw material production process matching the product quality data;

[0013] The dynamic raw material production process is adopted to produce the raw materials.

[0014] Furthermore, it also includes:

[0015] Establish quality-process-performance relationship model;

[0016] The quality-process-performance relationship model is reversely applied and constraint conditions are solved to establish a quality filtering model, wherein the input and output of the quality filtering model are product quality data and a dynamic raw material production process, respectively.

[0017] Furthermore, the establishment of the quality-process-performance relationship model includes:

[0018] Collecting historical data of historical batches of the raw materials, the historical data including: raw material quality batch data, production process batch data, and performance batch data;

[0019] Combining mathematical methods with relevant expert knowledge, we screen the quality factor set, select the quality advantage zone, and establish the quality-process-performance relationship model.

[0020] Furthermore, it also includes:

[0021] A data alignment operation is performed on the product quality data to input the aligned product quality data into the quality filtering model.

[0022] Furthermore, it also includes:

[0023] For specific materials, specific varieties, brands and uses, a corresponding specific quality filtering model is established to input the product quality data of the raw materials into the specific quality filtering model corresponding to the raw materials.

[0024] Furthermore, the raw materials are raw materials of a single production link, or raw materials of all production links in the entire industrial chain.

[0025] Furthermore, the various production links of the entire industrial chain include: raw materials, materials, components, structures, and equipment.

[0026] Furthermore, the raw material is a metal material.

[0027] Furthermore, the metal material includes a steel billet.

[0028] In another aspect, a raw material quality filtering system includes:

[0029] A memory for storing a computer-readable storage medium;

[0030] The processor is used to read and execute the storage medium in the memory to execute any one of the above methods.

[0031] After adopting the above technical solution, the present invention has the following advantages: by obtaining the product quality data of the raw materials; inputting the product quality data into the pre-established quality filtering model, a dynamic raw material production process matching the product quality data is obtained; and the raw materials are produced using the dynamic raw material production process. The present invention does not require narrow range control of the raw materials and subsequent production processes, and while eliminating quality volatility, it improves the yield rate and reduces production costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1 This is a step diagram of a raw material quality filtering method in one embodiment of the present invention.

[0034] Figure 2 This is a step diagram of a raw material quality filtering method in another embodiment of the present invention.

[0035] Figure 3 This is a step diagram of a raw material quality filtering method in another embodiment of the present invention.

[0036] Figure 4 This is a step diagram of a raw material quality filtering method in another embodiment of the present invention.

[0037] Figure 5 It is a diagram of the solution result of the quality filtering model in one embodiment of the present invention.

[0038] Figure 6 This is a rendering of the ellipsoidal model used in the present invention. DETAILED DESCRIPTION

[0039] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0040] Embodiment 1,

[0041] like Figure 1 The figure shows a step diagram of a raw material quality filtering method according to an embodiment of the present invention, comprising:

[0042] Step S101, obtaining product quality data of raw materials;

[0043] The data of raw material products is the data used to build the model, which is the historical data of a product of the enterprise. Data source: generally, it is necessary to retrieve batch data from the relevant system of the enterprise's manufacturing management, such as the MES system, which can be retrieved automatically through the system interface or manually retrieved and input into the system of the present invention.

[0044] Step S102: input the product quality data into a pre-established quality filtering model to obtain a dynamic raw material production process that matches the product quality data;

[0045] The relationship model is established by using mathematical methods. According to the user's usage effect, the optimal zone of product performance is established. The area inside the optimal zone is the advantage, and the area outside the optimal zone is the disadvantage. The setting of the optimal zone is user-oriented, such as the optimal zone range of strength, the optimal zone range of plasticity index, etc., to find the optimal regional boundary function of quality advantages and disadvantages in high-dimensional space, and also includes the determination of factors. This function is the "raw materials-process-performance" model. Through the Habitat model, the solution function model of the process is reversely deduced.

[0046] Specifically, for specific materials, specific varieties, brands and uses, a corresponding specific quality filtering model is established to input the product quality data of the raw materials into the specific quality filtering model corresponding to the raw materials.

[0047] Step S103: adopt a dynamic raw material production process to produce raw materials.

[0048] Based on the results of quality classification analysis, a quality filtering model is established to input billet / raw material quality data and output quality filtering to hedge the raw material production process and / or product production process. This method can be used not only in a single production link of raw materials and subsequent equipment manufacturing, but also in any upstream and downstream manufacturing links of the entire equipment manufacturing industry chain with the circulation process of raw materials as the carrier, to achieve the convergence of product quality volatility, that is, to achieve the quality filtering effect.

[0049] The raw materials may be metal materials, including steel billets, or raw materials from a single production link, or raw materials from all production links in the entire industrial chain.

[0050] Quality data, including:

[0051] Billet data, including billet composition data

[0052] Production process data, including rolling process data, heat treatment process data

[0053] Performance data, including mechanical properties and other data

[0054] Quality data can be collected from the company's production and related systems.

[0055] The embodiments provided by the present invention obtain product quality data of raw materials, input the product quality data into a pre-established quality filtering model, and obtain a dynamic raw material production process that matches the product quality data; the dynamic raw material production process is used to produce the raw materials. It can be understood that the quality data corresponds to the "steel billet → production → finished product" process. For example, medium and thick steel plate products can be corresponded according to "smelting furnace number → slab number → plate number / inspection batch number". According to different varieties, combined with expert experience and principal component analysis methods, the data fields involved in modeling and calculation in the above quality data are selected. The present invention does not require narrow range control of raw materials and subsequent production processes. While eliminating quality volatility, it improves the yield rate and reduces production costs.

[0056] Embodiment 2,

[0057] As a further improvement of the above method, Figure 2 As shown, it is a step diagram of a raw material quality filtering method according to another embodiment of the present invention.

[0058] include:

[0059] Step S201, obtaining product quality data of raw materials;

[0060] Step 202: Establish a quality-process-performance relationship model;

[0061] Step 203: reversely apply the quality-process-performance relationship model and solve the constraints to establish a quality filtering model, the input and output of the quality filtering model are product quality data and dynamic raw material production process respectively.

[0062] Step S204: input the product quality data into a pre-established quality filtering model to obtain a dynamic raw material production process that matches the product quality data;

[0063] The characteristic of quality filtering is that it is a quality control process for specific products, brands, and different specifications in a specific production and manufacturing link. Its core is to use the product quality big data to conduct dynamic analysis and modeling. After the system completes the quality filtering modeling, the quality data of the blank / raw material is input, and the appropriate production / manufacturing process is intelligently selected.

[0064] Among them, the process of establishing the quality filtering model is:

[0065]

[0066] in:

[0067] P i For product i Performance

[0068] C is the raw material matrix, C= c 1 ,c 2 ...c m

[0069] T is the process matrix, T= t 1 ,t 2 ...t q

[0070] f i For performance P i Mathematical model of the composition-process-performance relationship

[0071] g j Professional constraints

[0072] T min , T max Process range

[0073] δ is a global constraint solver (i.e., quality filter model)

[0074] f i It is necessary to fit and optimize the collected data on billet composition, production process, and actual performance, that is, to establish a function of the "composition-process-performance" relationship model.

[0075] The quality filtering model can use linear or nonlinear model functions according to the complexity of the field and the number of optimization targets, and use the collected data for fitting and iterative optimization.

[0076] In the quality filtering, the δ solver (quality filter model) is used to solve the process, that is, the composition data of the steel billet is input, and the corresponding process parameters are solved by the solver.

[0077] Step S205: adopt a dynamic raw material production process to produce raw materials.

[0078] The present invention establishes a quality-process-performance relationship model; applies the quality-process-performance relationship model in reverse and solves the constraints to establish a quality filtering model, the input and output of which are product quality data and dynamic raw material production process, respectively. Aiming at the pain point of quality fluctuation control of the above-mentioned raw materials and their products, the present invention abandons the traditional static process method and adopts the dynamic process matching method to achieve automatic adaptation of the production / manufacturing process that hedges against the quality fluctuation of the blank / raw material for specific metal material products and parts and equipment products manufactured using the raw materials, while obtaining the quality level of the blank / raw material, so as to ensure that the quality fluctuation of the final product is greatly reduced and stably controlled.

[0079] Embodiment 3,

[0080] As a further improvement of the above method, Figure 3 As shown, it is a step diagram of a raw material quality filtering method according to another embodiment of the present invention.

[0081] Step S301, obtaining product quality data of raw materials;

[0082] Step 302: Collect historical data of historical batches of raw materials, including: raw material quality batch data, production process batch data, and performance batch data;

[0083] Step 303: Combine mathematical methods with relevant expert knowledge to screen the quality factor set, select the quality optimization area, and establish a quality-process-performance relationship model.

[0084] Step 304: reversely apply the quality-process-performance relationship model and solve the constraints to establish a quality filtering model, the input and output of the quality filtering model are product quality data and dynamic raw material production process respectively.

[0085] Step S305: input the product quality data into a pre-established quality filtering model to obtain a dynamic raw material production process that matches the product quality data;

[0086] Step S306: adopt a dynamic raw material production process to produce raw materials.

[0087] Embodiment 4,

[0088] As a further improvement of the above method, Figure 4 As shown, it is a step diagram of a raw material quality filtering method according to another embodiment of the present invention.

[0089] Step S401, obtaining product quality data of raw materials;

[0090] Step S402: Perform a data alignment operation on the product quality data to input the aligned product quality data into a quality filtering model.

[0091] Step S403: adopt a dynamic raw material production process to produce raw materials.

[0092] Embodiment 5,

[0093] Preferably, the present invention also provides a raw material quality filtering system, comprising:

[0094] A memory for storing a computer-readable storage medium;

[0095] The processor is used to read and execute the storage medium in the memory to execute any one of the above methods.

[0096] For ease of understanding, the present invention also provides embodiments in practical applications.

[0097] Take the yield strength quality filtering of steel products (medium and thick plates) as an example:

[0098] Data collection:

[0099] Table 1 shows 500 data samples queried and retrieved from the MES system of a steel enterprise. The fields include the composition, rolling process and performance (yield strength) of the steel billet. The raw material data (composition) includes: C, Mn, V elements, and the process data includes: final rolling temperature and red-returning temperature.

[0100] Table 1

[0101]

[0102] Build a quality filter model:

[0103] Using a linear model, the model function of "ingredients-process-performance" is fitted as follows:

[0104] YS = 962−8.97C+164.9Mn+431.5V−0.3487T1−0.5831T2

[0105] Where:

[0106] YS is the yield strength of the steel plate

[0107] C is the carbon content of the steel plate

[0108] Mn is the manganese content of the steel plate

[0109] V is the vanadium content of the steel plate

[0110] T1 is the final rolling temperature

[0111] T2 is the red return temperature

[0112] According to material properties and expert experience, the constraints for solving are set as follows:

[0113] T1 = TAr3 (C, Mn, V) + 75

[0114] T1 = {800, 960}

[0115] T2 = {600, 760}

[0116] Where:

[0117] TAr3 (C, Mn, V) is the starting temperature of ferrite transformation of the steel plate under the combination of C, Mn and V elements, which is given by experts.

[0118] Get the global constraint solver (quality filter model):

[0119]

[0120] Quality filtering effect verification:

[0121] To facilitate calculation, the model and calculation process can be programmed. The following is the calculation result after running the filtering calculation program:

[0122]

[0123] After inputting the billet data, the process combination of the target performance (target yield strength) is calculated. The comparison between the first and second groups shows that the model can calculate the corresponding processes when different billet qualities (combinations of components) correspond to the same target performance; the comparison between the second and third groups shows that the model can calculate the corresponding processes when the same billet corresponds to different target performances.

[0124] Another embodiment is also provided, taking the multi-performance target quality filtering of steel products (medium and thick plates) as an example,

[0125] Data collection:

[0126] As shown in Table 2, Table 2 is about 3,000 pieces of smelting, rolling and inspection data of steel plates extracted from the MES system of a steel enterprise. The variety is steel plates for wind turbine tower barrels. The billet data includes 15 component field data such as C, Mn, S, P, Cu, Cr, Ni, Nb, V, Ti, B, N, and billet specification data. The rolling process data includes 5 fields of data including heating temperature, second stage temperature, waiting thickness ratio, final rolling temperature, and red return temperature, as well as the thickness data of the finished product after rolling. The performance data includes 4 fields of data including yield strength, tensile strength, elongation, and impact energy.

[0127] Table 2

[0128]

[0129] Build a quality filter model:

[0130] The above single-objective model uses a linear method to establish the "composition-process-performance" relationship function, but multi-objective optimization targets a combination of multiple properties rather than a single property. In the actual production process, multi-objective optimization, taking steel materials as an example, often requires an optimized performance range composed of multiple properties such as yield strength, tensile strength, elongation, and impact energy.

[0131] Therefore, the discriminant function method is used to establish a discriminant function model, which can maximize the collected data according to the set performance combination requirements, distinguishing the good area that meets the requirements and the bad area that does not meet the requirements. Based on the discriminant function, a quality filter model is established. With this model, the process combination that meets the set performance good area can be solved under the given billet composition.

[0132] In summary, the discriminant function is the key in multi-objective optimization. In this case of medium and thick plate quality filtering, two discriminant function models, the plane model and the ellipsoidal model, were used. The plane model method is to establish a plane that distinguishes the good area and the bad area in the high-dimensional space, and establish a quality filtering model based on the function of the plane (i.e., the discriminant function).

[0133] Selected fields:

[0134] According to the collected data, combined with principal component analysis and expert experience, 6 billet quality fields including C, Mn, S, Nb, N, and billet thickness and 5 process fields including second-stage temperature, waiting temperature thickness ratio, final rolling temperature, red-returning temperature, and furnace temperature were selected, and 12 fields and steel plate thickness were combined, and their values ​​were respectively given by x 1 ~ x 12 express.

[0135] Create a discriminant function:

[0136] f ( x 1 ,x 2 …x L ) = 0

[0137] For any sample ( x 1 ,x 2…x L )have:

[0138] f ( x 1 ,x 2 …x L )>0 x ∈ φ +

[0139] f ( x 1 ,x 2 …x L )<0 x ∈ φ -

[0140] Where:

[0141] f is the discriminant function

[0142] x 1 ,x 2 …x L The field value for modeling

[0143] φ + Quality Zone

[0144] φ - Poor quality area

[0145] Then the surface represented by this function is the discriminant surface, and this function is the discriminant function.

[0146] In addition, in order to eliminate the errors caused by the difference in the order of magnitude of data in different fields, all data are normalized.

[0147] x’ =( x - x min ) / ( x max - x min )

[0148] The calculation formula for inverse restoration of normalized data is:

[0149] x = x '( x max - x min)+ x min

[0150] Where:

[0151] x is the original value

[0152] x’ is the normalized value

[0153] x max is the maximum value of x

[0154] x min is the minimum value of x

[0155] After normalization:

[0156]

[0157] (1) Model 1: Linear model

[0158] After fitting and iterative optimization, the discriminant function polynomial is obtained as follows:

[0159] -3.4664x1 - 6.3230x2 - 13.8688x3 - 23.669x4 + 8.6435x5 - 4.2352x6 -8.7x7 -57.0517x8 + 17.4542x9 - 9.2454x10 + 3.6587x11 + 6.3205x12+42.100

[0160] Where:

[0161] x1 is the C element content

[0162] x2 is the Mn element content

[0163] x3 is the S element content

[0164] x4 is the Nb element content

[0165] x5 is the N element content

[0166] x6 is the thickness of the steel plate

[0167] x7 is the thickness of the blank

[0168] x8 is the second stage temperature

[0169] x9 is the thickness ratio of the waiting temperature

[0170] x10 is the final rolling temperature

[0171] x11 is the red return temperature

[0172] x12 is the oven temperature.

[0173] The quality filter model function of this steel grade is established as:

[0174]

[0175] Figure 5 Shown is a diagram of the solution results of the quality filtering model in one embodiment of the present invention.

[0176] In one embodiment, an ellipsoidal model is also established, and the discriminant function polynomial of the ellipsoidal model is:

[0177]

[0178] Where:

[0179] x 1 C element content

[0180] x 2 Mn element content

[0181] x 3 is the S element content

[0182] x 4 Nb element content

[0183] x 5 N element content

[0184] x 6 Steel plate thickness

[0185] x 7 is the thickness of the blank

[0186] x 8 The second stage temperature

[0187] x 9 is the thickness ratio of the waiting temperature

[0188] x 10 The final rolling temperature

[0189] x 11 Red-return temperature

[0190] x 12 The oven temperature

[0191] like Figure 6 As shown, it is a diagram showing the effect of distinguishing an ellipsoidal model in one embodiment of the present invention.

[0192] Taking the production of steel materials as an example, in this embodiment, the batch data of steel plate production is used to model the "billet-process-performance" relationship and determine the quality optimal zone, thereby realizing quality filtering modeling with process output as the target. After the modeling is completed, the quality characteristic data of a single steel billet or a batch of multiple steel billets, such as composition data, are input. The system can automatically adapt the rolling process of each steel billet according to the optimal zone target of the mechanical properties of the steel plate, so that the mechanical properties of the rolled steel plate are in the optimal zone. For batches of steel plates, dynamic process adaptation is used to make the performance in the optimal zone and the volatility within a smaller range.

[0193] Taking the manufacturing process of parts and equipment as an example, the raw materials of parts and equipment products are metal materials, such as steel materials, and the manufacturing process is to use steel to perform deformation, welding, heat treatment and other processing. Therefore, for the processing process in the equipment manufacturing process, the mass data of the steel plate, such as composition, microstructure, performance data, etc., are input into the system, and the deformation process, welding process and other processing processes of the steel plate are adapted by the system to narrow the quality fluctuation of the parts or equipment after deformation and welding.

[0194] The method disclosed by the present invention has the scalability of the whole industrial chain. For the whole industrial chain process of equipment manufacturing with raw materials as carriers, including from ore and charge end to raw material production, equipment manufacturing and application process, the output of each link in the industrial chain is the raw material of the next link. Therefore, in the industrial chain, the quality filtering method is adopted between each link up and down, which can achieve the quality filtering effect of the whole industrial chain and achieve the convergence of the quality volatility of the final finished product from the dimension of the whole industrial chain.

[0195] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0196] It should be noted that, in the description of this application, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0197] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0198] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0199] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0200] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0201] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0202] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0203] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A raw material quality filtering method, characterized in that: include: Obtain product quality data of raw materials; Establish quality-process-performance relationship model; Reversely apply the quality-process-performance relationship model and solve the constraint conditions to establish a quality filtering model, wherein the input and output of the quality filtering model are product quality data and a dynamic raw material production process, respectively; Inputting the product quality data into a pre-established quality filtering model to obtain a dynamic raw material production process matching the product quality data; Using the dynamic raw material production process to produce the raw materials; The process of establishing the quality filtering model is as follows: in, P i For product i performance, C is the raw material matrix, C= c 1 ,c 2 ...c m ， T is the process matrix, T= t 1 ,t 2 ...t q ，f i is the quality-process-performance relationship model, g j is the professional constraint, δ is the quality filter model; Wherein, the establishment of the quality-process-performance relationship model includes: Collecting historical data of historical batches of the raw materials, the historical data including: raw material quality batch data, production process batch data, and performance batch data; Combining mathematical methods with relevant expert knowledge, we screen the quality factor set, select the quality advantage zone, and establish the quality-process-performance relationship model.

2. The method according to claim 1, characterized in that Also includes: A data alignment operation is performed on the product quality data to input the aligned product quality data into the quality filtering model.

3. The method according to claim 1, characterized in that Also includes: For specific materials, specific varieties, brands and uses, a corresponding specific quality filtering model is established to input the product quality data of the raw materials into the specific quality filtering model corresponding to the raw materials.

4. The method according to any one of claims 1 to 3, characterized in that: The raw materials mentioned are raw materials for a single production link, or raw materials for all production links in the entire industrial chain.

5. The method according to claim 4, characterized in that The various production links of the entire industrial chain include: raw materials, materials, components, structures, and equipment.

6. The method according to any one of claims 1 to 3, characterized in that: The raw material is a metal material.

7. The method according to claim 6, characterized in that The metal material includes a steel billet.

8. A raw material quality filtering system, characterized in that: include: A memory for storing computer-readable storage media; A processor, configured to read and enable the storage medium in the memory to execute the method according to any one of claims 1 to 7.

Images