Die steel heat treatment performance prediction system based on big data

Abstract

The invention provides a die steel heat treatment performance prediction system based on big data, and belongs to the technical field of intelligent prediction, and the system comprises a data collection module which is used for obtaining detailed chemical component data, production process setting data and production process monitoring data of different batches of steel from a die steel manufacturer; the determining module is used for determining a component difference array and an affected array of the chemical components, the process setting parameters and the process monitoring parameters based on the preprocessed data, and meanwhile determining a first factor set influencing the heat treatment performance of each batch of steel in the historical actual production process; the result prediction module is used for inputting the actual acquired parameters of the current die steel into the performance prediction model and determining an initial performance prediction result; and the result optimization module is used for optimizing the initial performance prediction result based on the component difference array, the influenced array and the first factor set to obtain a current performance prediction result. And the prediction accuracy of the heat treatment performance of the die steel is ensured.

Description

Technical Field

[0001] The present invention relates to the field of intelligent prediction technology, and in particular to a die steel heat treatment performance prediction system based on big data. Background Art

[0002] Mold steel heat treatment is a process that changes the structure and performance of mold steel to meet the requirements of mold use by heating, keeping warm and cooling the mold steel. In the mold manufacturing industry, the precise control of mold steel heat treatment performance plays a decisive role in product quality, production efficiency and cost control. The traditional method of relying on experience and repeated trials is not only inefficient, but also difficult to meet the growing demand for high-precision and high-performance molds.

[0003] Therefore, the present invention proposes a die steel heat treatment performance prediction system based on big data. Summary of the invention

[0004] The present invention provides a mold steel heat treatment performance prediction system based on big data, which is used to optimize the initial performance prediction results by collecting chemical composition data, production process setting data and production process monitoring data to establish a composition difference array, an affected array and a factor set, thereby ensuring the accuracy of the mold steel heat treatment performance prediction.

[0005] The present invention provides a die steel heat treatment performance prediction system based on big data, comprising: The data acquisition module is used to obtain detailed chemical composition data, production process setting data and production process monitoring data of different batches of steel from mold steel manufacturers, and to clean and preprocess the collected data; A determination module is used to determine the chemical composition, the composition difference array of the process setting parameters and the process monitoring parameters, and the affected array based on the pre-processed data, and at the same time, determine the first factor set that affects the heat treatment performance of each batch of steel in the actual historical production process; The result prediction module is used to input the actual acquired parameters of the current mold steel into the performance prediction model to determine the initial performance prediction results; The result optimization module is used to optimize the initial performance prediction result based on the component difference array, the affected array and the first factor set to obtain the current performance prediction result.

[0006] Preferably, the determining module includes: A comparison and analysis unit, used to compare and analyze the historical actual production process with the historical theoretical production process to determine the abnormal factors corresponding to the abnormal values; The statistical unit is used to count all abnormal factors to obtain a first factor set.

[0007] Preferably, the determining module further includes: A composition determination unit is used to determine the production model of the die steel, and determine the production purpose based on the production model, and then obtain key analysis parameters based on the production purpose from the purpose-analysis comparison table to obtain the specified composition data corresponding to the production model, wherein the key analysis parameters are related to the chemical composition; The process deviation unit is used to match and divide the production process monitoring data of each batch of mold steel with the production process setting data according to the production stage to determine the actual process deviation set of the second monitoring data. ,in, It represents the difference of the jth process parameter of the corresponding mold steel at the corresponding production stage; Represents the actual value of the jth process parameter of the corresponding mold steel at the corresponding production stage; It represents the standard value of the jth process parameter of the corresponding mold steel in the corresponding production stage; m2 is the total number of process parameters involved in the corresponding production stage; A deviation impact unit, configured to determine a deviation impact set of the first monitoring data at the corresponding production stage on the standard process data based on the actual process deviation set; ,in, They respectively represent the influence coefficients of the production environment temperature and the production environment humidity involved in the corresponding production stage on the jth process parameter; An array construction unit, used to construct an affected array of all process parameters involved in each production stage according to the deviation impact set; The difference construction unit is used to extract the standard component array under the corresponding production stage from the specified component data to compare with the historical component array of the corresponding production stage obtained by actual monitoring to obtain a component difference array.

[0008] Preferably, the deviation influencing unit comprises: An array determination subunit, used to determine an initial array of the first monitoring data corresponding to the production stage { , };

[0009]

[0010]

[0011]

[0012] in, Indicates the evaluation influence factor of all production environment temperatures on process parameters corresponding to the first monitoring data; Indicates the evaluation influence factor of all production environment humidity on process parameters corresponding to the first monitoring data; represents the error judgment function of the i1th production environment temperature in the corresponding production stage; represents the error judgment function of the humidity of the i2th production environment in the corresponding production stage; Indicates the number of production environment temperatures involved in the corresponding production stage, and is consistent with the number of production environment humidity; Indicates that the temperature of the i1th production environment is continuously at the current location The number of consecutive states or the duration of The number of consecutive states or the number of consecutive The number of consecutive states; Indicates that the humidity of the i2nd production environment is continuously at the current position The number of consecutive states or the duration of The number of consecutive states or the number of consecutive The number of consecutive states; Indicates the temperature of the i1th production environment; Indicates the humidity of the i2th production environment; , They represent the temperature threshold and humidity threshold of the corresponding production stage respectively; , Respectively represent the i1th production environment temperature based on , The influence of parameter values ​​in different situations; , Respectively represent the humidity of the i2th production environment based on , The parameter value affects the amount of influence in the case.

[0013] Preferably, the deviation influencing unit further includes: The error adjustment subunit is used to perform system error adjustment on the difference of the jth process parameter in the corresponding production stage, and combines the initial array { , }Respectively determine the first influence coefficient and the second influence coefficient of the production environment temperature and the production environment humidity on the adjusted difference.

[0014] Preferably, the error adjustment subunit includes: like , assign 1 to the first influence coefficient and the second influence coefficient in the corresponding production stage respectively, and calibrate them to the jth process parameter, where, Indicates the ratio threshold of the jth process parameter based on temperature Indicates the ratio threshold of the jth process parameter based on humidity It represents the value after the system error adjustment of the difference of the jth process parameter in the corresponding production stage; like , assign 1 to the first influence coefficient and 0 to the second influence coefficient in the corresponding production stage, and calibrate them to the jth process parameter; like , assign 0 to the first influence coefficient and 1 to the second influence coefficient in the corresponding production stage, and calibrate them to the jth process parameter; like , assign 0 to the first influence coefficient and the second influence coefficient in the corresponding production stage respectively, and calibrate them to the jth process parameter.

[0015] Preferably, the result optimization module includes: A coefficient determination unit, used for inputting the component difference array, the affected array, the first factor set and the corresponding historical performance results into different performance indicator models, obtaining coefficients to be optimized for different performance indicators at different production stages, and extracting the optimal coefficient and the first variance; The optimization unit is used to optimize and adjust the initial performance value under each performance indicator according to the minimum coefficient.

[0016] Preferably, the performance indicators include: a hardness index, a toughness index and a wear resistance index.

[0017] Compared with the prior art, the present invention has the following beneficial effects: By collecting chemical composition data, production process setting data and production process monitoring data, a composition difference array and an affected array are established and combined with a factor set to optimize the initial performance prediction results and ensure the accuracy of the mold steel heat treatment performance prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are 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.

[0019] Figure 1 It is a structural diagram of a mold steel heat treatment performance prediction system based on big data provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] The present invention provides a die steel heat treatment performance prediction system based on big data, such as Figure 1 As shown, including: The data acquisition module is used to obtain detailed chemical composition data, production process setting data and production process monitoring data of different batches of steel from mold steel manufacturers, and to clean and preprocess the collected data; A determination module is used to determine the chemical composition, the composition difference array of the process setting parameters and the process monitoring parameters, and the affected array based on the pre-processed data, and at the same time, determine the first factor set that affects the heat treatment performance of each batch of steel in the actual historical production process; The result prediction module is used to input the actual acquired parameters of the current mold steel into the performance prediction model to determine the initial performance prediction results; The result optimization module is used to optimize the initial performance prediction result based on the component difference array, the affected array and the first factor set to obtain the current performance prediction result.

[0022] In this embodiment, the chemical composition data includes the content of each trace element in the mold steel, such as the content of carbon, chromium, molybdenum, vanadium and other elements.

[0023] The production process setting data includes: theoretical temperature, theoretical time, theoretical rate and other process parameters in the heating, insulation and cooling stages; The production process monitoring data includes process parameters such as actual temperature, actual time, actual rate, etc. during the heating, insulation and cooling stages, and also includes environmental data, such as humidity data and temperature data of mold steel at different production stages.

[0024] In this implementation, the performance prediction results are related to the hardness, toughness, wear resistance, etc. of the die steel.

[0025] In this embodiment, data cleaning and preprocessing is to use data cleaning algorithms to identify and correct erroneous values, missing values, and duplicate values ​​in the data. Data from different sources and formats are standardized, and data units and encoding methods are unified to lay the foundation for subsequent analysis and modeling.

[0026] In this embodiment, the performance prediction model is obtained by training the neural network model with the composition of the new mold steel and the proposed process parameters as input samples, and the performance prediction values ​​such as hardness, toughness, and wear resistance as output samples. Since the model does not consider factors related to the environment, it is necessary to determine the relevant composition difference array, the affected array, and the first factor set to improve the accuracy of the performance result prediction.

[0027] In this embodiment, the die steel heat treatment includes the following processes: Annealing: Heating the steel to a certain temperature and keeping it warm for a period of time, then slowly cooling it down. The purpose is to eliminate structural defects, improve the structure to make the composition uniform and refine the grains, improve the mechanical properties of the steel, and reduce residual stress; at the same time, it can reduce hardness, increase plasticity and toughness, and improve cutting performance.

[0028] Normalizing: Heat the steel to above the critical temperature to transform the steel into uniform austenite, and then cool it naturally in the air. It can eliminate the network cementite of hypereutectoid steel. For hypoeutectoid steel, normalizing can refine the lattice and improve the comprehensive mechanical properties. It is more economical to use normalizing instead of annealing process for parts with low requirements.

[0029] Quenching: Heat the steel to above the critical temperature, keep it warm for a period of time, and then quickly put it into the quenchant to make its temperature drop suddenly, and cool it rapidly at a rate greater than the critical cooling rate to obtain an unbalanced structure dominated by martensite. Quenching can increase the strength and hardness of steel, but will reduce its plasticity.

[0030] Tempering: reheating the quenched steel to a certain temperature and then cooling it by a certain method. The purpose is to eliminate the internal stress caused by quenching, reduce hardness and brittleness, and obtain the expected mechanical properties. Tempering is divided into three categories: high temperature tempering, medium temperature tempering and low temperature tempering, and is often used in conjunction with quenching and normalizing.

[0031] Surface heat treatment: including surface quenching and chemical heat treatment. Surface quenching is to quickly heat the surface of the steel to above the critical temperature, but quickly cool it before the heat can be transferred to the core, so that the outer layer is quenched into the martensite structure, while the core remains unchanged; chemical heat treatment refers to the atoms of chemical elements, relying on the ability of atomic diffusion at high temperature, entering the surface layer of the workpiece to change the chemical composition and structure of the surface layer of the workpiece, including carburizing, nitriding, cyaniding and metal infiltration.

[0032] It should be noted that different types of mold steel, such as cold working mold steel, hot working mold steel, etc., have different specific heat treatment process parameters, which need to be reasonably selected and adjusted according to factors such as the type of steel and the use requirements of the mold.

[0033] The beneficial effect of the above technical solution is: by collecting chemical composition data, production process setting data and production process monitoring data to establish a composition difference array, an affected array and combining a factor set, the initial performance prediction results are optimized to ensure the accuracy of the mold steel heat treatment performance prediction.

[0034] The present invention provides a die steel heat treatment performance prediction system based on big data, wherein the determination module comprises: A comparison and analysis unit, used to compare and analyze the historical actual production process with the historical theoretical production process to determine the abnormal factors corresponding to the abnormal values; The statistical unit is used to count all abnormal factors to obtain a first factor set.

[0035] In this embodiment, for example, the working parameters set by the process involved in the heat treatment of mold steel are compared with the actual parameters. If the difference between the two is too large, the parameter description of the corresponding set working parameters will be regarded as an abnormal factor. For example, during the annealing process, the steel needs to be kept warm for 10 minutes after being heated to a certain temperature, but the actual insulation time is 8 minutes, resulting in an increase in hardness. At this time, the insulation time involved in the annealing process is regarded as an abnormal factor.

[0036] In this embodiment, the first factor set = {all abnormal factors}.

[0037] The beneficial effect of the above technical solution is: by comparing the actual historical production process with the theoretical production process, abnormal factors are determined, providing a basis for subsequent optimization of the initial predicted performance results.

[0038] The present invention provides a die steel heat treatment performance prediction system based on big data, wherein the determination module further comprises: A composition determination unit is used to determine the production model of the die steel, and determine the production purpose based on the production model, and then obtain key analysis parameters based on the production purpose from the purpose-analysis comparison table to obtain the specified composition data corresponding to the production model, wherein the key analysis parameters are related to the chemical composition; The process deviation unit is used to match and divide the production process monitoring data of each batch of mold steel with the production process setting data according to the production stage to determine the actual process deviation set of the second monitoring data. ,in, It represents the difference of the jth process parameter of the corresponding mold steel at the corresponding production stage; Represents the actual value of the jth process parameter of the corresponding mold steel at the corresponding production stage; It represents the standard value of the jth process parameter of the corresponding mold steel in the corresponding production stage; m2 is the total number of process parameters involved in the corresponding production stage; A deviation impact unit, configured to determine a deviation impact set of the first monitoring data at the corresponding production stage on the standard process data based on the actual process deviation set; ,in, They respectively represent the influence coefficients of the production environment temperature and the production environment humidity involved in the corresponding production stage on the jth process parameter; An array construction unit, used to construct an affected array of all process parameters involved in each production stage according to the deviation impact set; The difference construction unit is used to extract the standard component array under the corresponding production stage from the specified component data to compare with the historical component array of the corresponding production stage obtained by actual monitoring to obtain a component difference array.

[0039] In this embodiment, for example, the standard component array extracted in production stage 1 is {u1, u2}, and the historical component array actually monitored is {r1, r2}, and the component difference array is: {u1-r1, u2-r2}.

[0040] In this embodiment, the usage-analysis comparison table contains parameters related to chemical composition under different production purposes. For example, the carbon composition of steel used for house construction (production purpose) is u1, the chromium composition is u2, the molybdenum composition is u3, and the vanadium composition is u4. The remaining chemical compositions can be ignored (specified composition data), among which the key analysis parameters are: carbon composition, chromium composition, molybdenum composition and vanadium composition.

[0041] In this embodiment, the production model refers to a unique code corresponding to the mold steel.

[0042] In this embodiment, since the production of a batch of mold steel will go through different process flows, there will be multiple production stages.

[0043] In this embodiment, the process parameters under the production process of different batches of mold steel all have corresponding standard values, which are all pre-set and can be directly called up and used, because the firing standards of different types of mold steel are unified.

[0044] In this embodiment, the actual value refers to the value actually monitored during the historical process.

[0045] In this embodiment, the number of process parameters is at least greater than 3.

[0046] In this embodiment, That is, the first influence coefficient and the second influence coefficient obtained subsequently.

[0047] In this embodiment, the affected array={the deviation impact set of all process parameters in each production stage}.

[0048] The beneficial effects of the above technical solution are: by determining the production model to obtain the specified composition data, a comparison basis is provided for obtaining the composition difference data, and by determining the actual process deviation set and the deviation impact set on the standard process data to construct an affected array. The combination of the two provides a basis for subsequent prediction adjustments to ensure prediction accuracy.

[0049] The present invention provides a die steel heat treatment performance prediction system based on big data, wherein the deviation influence unit comprises: An array determination subunit, used to determine an initial array of the first monitoring data corresponding to the production stage { , };

[0050]

[0051]

[0052]

[0053] in, Indicates the evaluation influence factor of all production environment temperatures on process parameters corresponding to the first monitoring data; Indicates the evaluation influence factor of all production environment humidity on process parameters corresponding to the first monitoring data; represents the error judgment function of the i1th production environment temperature in the corresponding production stage; represents the error judgment function of the humidity of the i2th production environment in the corresponding production stage; Indicates the number of production environment temperatures involved in the corresponding production stage, and is consistent with the number of production environment humidity; Indicates that the temperature of the i1th production environment is continuously at the current location The number of consecutive states or the duration of The number of consecutive states or the number of consecutive The number of consecutive states; Indicates that the humidity of the i2nd production environment is continuously at the current position The number of consecutive states or the duration of The number of consecutive states or the number of consecutive The number of consecutive states; Indicates the temperature of the i1th production environment; Indicates the humidity of the i2th production environment; , They represent the temperature threshold and humidity threshold of the corresponding production stage respectively; , Respectively represent the i1th production environment temperature based on , The influence of parameter values ​​in different situations; , Respectively represent the humidity of the i2th production environment based on , The parameter value affects the amount of influence in the case.

[0054] In this embodiment, for example, The moments are: moment 1, moment 2, moment 5, The moments are moment 3 and moment 4. At this time, the continuous quantity state for moment 2 is =2.

[0055] The beneficial effect of the above technical solution is: starting from the ambient temperature and ambient humidity, to determine the influencing factors of temperature and humidity in the production stage on the evaluation of process parameters, to compare and analyze the actual impact, to further verify the true and false impact of temperature and humidity on the corresponding parameters in different production stages, and to provide an analytical basis for subsequent predictions.

[0056] The present invention provides a die steel heat treatment performance prediction system based on big data, wherein the deviation influence unit further comprises: The error adjustment subunit is used to perform system error adjustment on the difference of the jth process parameter in the corresponding production stage, and combines the initial array { , }Respectively determine the first influence coefficient and the second influence coefficient of the production environment temperature and the production environment humidity on the adjusted difference.

[0057] Preferably, the error adjustment subunit includes: like , assign 1 to the first influence coefficient and the second influence coefficient in the corresponding production stage respectively, and calibrate them to the jth process parameter, where, Indicates the ratio threshold of the jth process parameter based on temperature Indicates the ratio threshold of the jth process parameter based on humidity It represents the value after the system error adjustment of the difference of the jth process parameter in the corresponding production stage; like , assign 1 to the first influence coefficient and 0 to the second influence coefficient in the corresponding production stage, and calibrate them to the jth process parameter; like , assign 0 to the first influence coefficient and 1 to the second influence coefficient in the corresponding production stage, and calibrate them to the jth process parameter; like , assign 0 to the first influence coefficient and the second influence coefficient in the corresponding production stage respectively, and calibrate them to the jth process parameter.

[0058] In this embodiment, = / (1-system error), where the system error is an inherent error and a fixed value that cannot be eliminated. represents the difference of the jth process parameter in the corresponding production stage; It represents the value after the difference of the j-th process parameter in the corresponding production stage is adjusted for the system error.

[0059] The beneficial effect of the above technical solution is: by calculating the ratio of temperature and humidity, the process parameters are given influence coefficients based on temperature and humidity, providing variables for constructing an affected array.

[0060] The present invention provides a die steel heat treatment performance prediction system based on big data, wherein the result optimization module comprises: A coefficient determination unit, used for inputting the component difference array, the affected array, the first factor set and the corresponding historical performance results into different performance indicator models, obtaining coefficients to be optimized for different performance indicators at different production stages, and extracting the optimal coefficient and the first variance; The optimization unit is used to optimize and adjust the initial performance value under each performance indicator according to the minimum coefficient.

[0061] Preferably, the performance indicators include: hardness index, toughness index and wear resistance index.

[0062] In this embodiment, a neural network is used to construct a complex nonlinear prediction model. Through training with a large amount of historical data, the chemical composition, the composition difference array between the process parameters and the monitoring parameters, the affected array, the first factor set and the predicted performance of the relevant history, the actual performance of the relevant history, and the coefficients to be optimized are learned to obtain different performance indicator models, and then the corresponding coefficients to be optimized can be directly obtained.

[0063] Since there is a coefficient to be optimized in each production stage, but in order to ensure the reliability of the calculation, the coefficient closest to 0 is selected and added to the first variance to optimize the initial performance, that is, the adjusted performance value = initial performance value × (1 + optimal coefficient + first variance). For example, the coefficients to be optimized are: -0.1, -0.05, 1. At this time, -0.05 is closest to 0, so -0.05 is regarded as the optimal coefficient. The first variance is the variance of the coefficients to be optimized in different production stages involved in the corresponding performance index.

[0064] The beneficial effect of the above technical solution is: by inputting the component difference array, the affected array, the first factor set, and the historical performance results into the model to obtain the optimal coefficient and the first variance for the indicator, to achieve optimization adjustment and ensure the accuracy of the performance result prediction.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A die steel heat treatment performance prediction system based on big data, characterized in that: include: The data acquisition module is used to obtain detailed chemical composition data, production process setting data and production process monitoring data of different batches of steel from mold steel manufacturers, and to clean and preprocess the collected data; A determination module is used to determine the chemical composition, the composition difference array of the process setting parameters and the process monitoring parameters and the affected array based on the pre-processed data, and at the same time, determine the first factor set that affects the heat treatment performance of each batch of steel in the historical actual production process; The result prediction module is used to input the actual acquired parameters of the current mold steel into the performance prediction model to determine the initial performance prediction results; The result optimization module is used to optimize the initial performance prediction result based on the component difference array, the affected array and the first factor set to obtain the current performance prediction result.

2. The die steel heat treatment performance prediction system based on big data according to claim 1 is characterized in that: The determining module comprises: A comparison and analysis unit, used to compare and analyze the historical actual production process with the historical theoretical production process to determine the abnormal factors corresponding to the abnormal values; The statistical unit is used to count all abnormal factors to obtain a first factor set.

3. The die steel heat treatment performance prediction system based on big data according to claim 1 is characterized in that: The determining module further includes: A composition determination unit is used to determine the production model of the die steel, and determine the production purpose based on the production model, and then obtain key analysis parameters based on the production purpose from the purpose-analysis comparison table to obtain the specified composition data corresponding to the production model, wherein the key analysis parameters are related to the chemical composition; The process deviation unit is used to match and divide the production process monitoring data of each batch of mold steel with the production process setting data according to the production stage to determine the actual process deviation set of the second monitoring data. ,in, It represents the difference of the jth process parameter of the corresponding mold steel at the corresponding production stage; Represents the actual value of the jth process parameter of the corresponding mold steel at the corresponding production stage; It represents the standard value of the jth process parameter of the corresponding mold steel in the corresponding production stage; m2 is the total number of process parameters involved in the corresponding production stage; A deviation impact unit, configured to determine a deviation impact set of the first monitoring data at the corresponding production stage on the standard process data based on the actual process deviation set; ,in, They respectively represent the influence coefficients of the production environment temperature and the production environment humidity involved in the corresponding production stage on the jth process parameter; An array construction unit, used to construct an affected array of all process parameters involved in each production stage according to the deviation impact set; The difference construction unit is used to extract the standard component array under the corresponding production stage from the specified component data to compare with the historical component array of the corresponding production stage obtained by actual monitoring to obtain a component difference array.

4. The die steel heat treatment performance prediction system based on big data according to claim 3 is characterized in that: The deviation influencing unit comprises: An array determination subunit, used to determine an initial array of the first monitoring data corresponding to the production stage { , }; in, Indicates the evaluation influence factor of all production environment temperatures on process parameters corresponding to the first monitoring data; Indicates the evaluation influence factor of all production environment humidity on process parameters corresponding to the first monitoring data; represents the error judgment function of the i1th production environment temperature in the corresponding production stage; represents the error judgment function of the humidity of the i2th production environment in the corresponding production stage; Indicates the number of production environment temperatures involved in the corresponding production stage, and is consistent with the number of production environment humidity; Indicates that the temperature of the i1th production environment is continuously at the current location The number of consecutive states or the duration of The number of consecutive states or the number of consecutive The number of consecutive states; Indicates that the humidity of the i2nd production environment is continuously at the current position The number of consecutive states or the duration of The number of consecutive states or the number of consecutive The number of consecutive states; Indicates the temperature of the i1th production environment; Indicates the humidity of the i2th production environment; , They represent the temperature threshold and humidity threshold of the corresponding production stage respectively; , Respectively represent the i1th production environment temperature based on , The influence of parameter values ​​in different situations; , Respectively represent the humidity of the i2th production environment based on , The parameter value affects the amount of influence in the case.

5. The die steel heat treatment performance prediction system based on big data according to claim 4 is characterized in that: The deviation influencing unit further includes: The error adjustment subunit is used to perform system error adjustment on the difference of the jth process parameter in the corresponding production stage, and combines the initial array { , }Respectively determine the first influence coefficient and the second influence coefficient of the production environment temperature and the production environment humidity on the adjusted difference.

6. The die steel heat treatment performance prediction system based on big data according to claim 5 is characterized in that: The error adjustment subunit comprises: like , assign 1 to the first influence coefficient and the second influence coefficient in the corresponding production stage respectively, and calibrate them to the jth process parameter, where, Indicates the ratio threshold of the jth process parameter based on temperature The adjustment coefficient of Indicates the ratio threshold of the jth process parameter based on humidity The adjustment coefficient of It represents the value after the system error adjustment of the difference of the jth process parameter in the corresponding production stage; like , assign 1 to the first influence coefficient and 0 to the second influence coefficient in the corresponding production stage, and calibrate them to the jth process parameter; like , assign 0 to the first influence coefficient and 1 to the second influence coefficient in the corresponding production stage, and calibrate them to the jth process parameter; like , assign 0 to the first influence coefficient and the second influence coefficient in the corresponding production stage respectively, and calibrate them to the jth process parameter.

7. The die steel heat treatment performance prediction system based on big data according to claim 1 is characterized in that: The result optimization module comprises: A coefficient determination unit, used for inputting the component difference array, the affected array, the first factor set and the corresponding historical performance results into different performance indicator models, obtaining coefficients to be optimized for different performance indicators at different production stages, and extracting the optimal coefficient and the first variance; The optimization unit is used to optimize and adjust the initial performance value under each performance indicator according to the minimum coefficient.

8. The die steel heat treatment performance prediction system based on big data according to claim 7 is characterized in that: The performance indicators include: hardness index, toughness index and wear resistance index.

Images