Industrial intelligent production method and system based on pre-training large model, and terminal equipment

Abstract

The invention relates to the technical field of industrial production, and discloses an industrial intelligent production method and system based on a pre-training large model, and terminal equipment, and the method comprises the steps: constructing the pre-training large model, obtaining a production index and a production parameter, obtaining a production index score, obtaining a strongly correlated production index, obtaining a correlated production parameter, and generating an early warning data set. The system and the terminal device both correspond to the method. According to the method, the industrial production data is simplified by accurately acquiring the production indexes and the production parameters, so that an administrator can conveniently acquire the production data with relatively high value, data-driven production analysis is realized by pre-training the large model, the proportion of human factors is reduced, and the production efficiency is improved. The obtained strong correlation production indexes and associated production parameters provide a data basis for industrial production problem traceability, the problem traceability is optimized, complementation of key data volume and optimization of a data processing mode are realized, and the industrial intelligence level is improved. And finally, industrial intelligent production based on the pre-training large model is realized.

Description

Technical Field

[0001] The present application relates to the field of industrial production technology, and specifically to an industrial intelligent production method, system and terminal equipment based on a pre-trained large model. Background Art

[0002] With the development of industry, the improvement of industrial intelligence level is a common problem faced by this industrial production field. However, the inventor found the following characteristics in industrial production in actual research, which led to the setback of the improvement of industrial intelligence level: 1. Massive industrial production data. Industrial production includes equipment data, capacity data, quality data, energy consumption data and environmental data. Each type of data can characterize the industrial production process to a certain extent. However, for managers, a large amount of data will lead to confusion in their management, which is not conducive to achieving intelligent effects. 2. It is impossible to reasonably evaluate industrial production from the perspective of data. The judgment of industrial production depends on experienced workers, resulting in a high proportion of human factors, which is not conducive to achieving intelligent effects; 3. It is difficult to trace the source of discovered industrial production problems, resulting in difficulty in finding the source of the problem even if the data is known to be abnormal, and even difficulty in narrowing the scope of investigation.

[0003] With the development of pre-trained large model technology, it has become possible to improve the level of industrial intelligence through pre-trained large models. However, there are still problems in the application of pre-trained large models in the industrial field due to insufficient key data or improper data processing methods.

[0004] In summary, the existing technology urgently needs a technical solution for industrial intelligent production based on pre-trained large models. Summary of the invention

[0005] The purpose of this application is to provide an industrial intelligent production method, system and terminal device based on a pre-trained large model to solve the technical problems raised in the above background technology.

[0006] To achieve the above objectives, this application discloses the following technical solutions: In a first aspect, the present application discloses an industrial intelligent production method based on a pre-trained large model, the method comprising: Construct a pre-trained large model, preset a supervision cycle, and perform the following steps based on the pre-trained large model in the supervision cycle: S1: Obtaining various production indicators, wherein the production indicators include corresponding production parameters; S2: using the production parameters to obtain the corresponding production index scores; S3: Analyze the production index scores to obtain strongly correlated production indicators; S4: Deconstructing the strongly correlated production indicators to obtain associated production parameters; S5: generating an early warning data set based on the associated production parameters; Among them: the pre-trained large model is used to supervise industrial production in all dimensions based on the production indicators; the production indicators also include equipment indicators, production capacity indicators, quality indicators, energy consumption indicators and environmental indicators and are used to characterize industrial production in all dimensions; the production parameters include equipment parameter sets, production capacity parameter sets, quality parameter sets, energy consumption parameter sets and environmental parameter sets and are used to comprehensively detect industrial production; the strongly correlated production indicators are the production indicators with strong correlation obtained through analysis; the associated production parameters are deconstructed to obtain the associated generation parameters; the early warning data set is used to provide analyzed data to administrators.

[0007] Preferably, the construction of the pre-trained large model includes training the pre-trained large model to obtain the production index score, training the pre-trained large model to obtain the strongly correlated production index, training the pre-trained large model to obtain the associated production parameters and training the pre-trained large model to generate the early warning data set.

[0008] Preferably, the device parameter set includes device parameters for characterizing device status, operating time, processing speed, device temperature, vibration amplitude, and current and voltage; The capacity parameter set includes capacity parameters for characterizing output, production cycle, process time and material consumption; The quality parameter set includes quality parameters for characterizing each item of yield rate, dimensional accuracy, surface roughness, hardness and strength; The energy consumption parameter set includes energy consumption parameters for characterizing various items of electricity consumption, water consumption and gas consumption; The environmental parameter set includes environmental parameters for characterizing each of temperature, humidity, noise level and air quality; The equipment parameters, the production capacity parameters, the quality parameters, the energy consumption parameters and the environmental parameters are all normalized data.

[0009] Preferably, the training of the pre-trained large model to obtain the production index score specifically includes: The device parameter set and the device indicator scoring formula are used to calculate the device indicator score, wherein the device indicator scoring formula is specifically: in: A first device parameter that characterizes a device state; A second device parameter characterizing the operating time; A third equipment parameter that characterizes the processing speed; A fourth device parameter characterizing device temperature; A fifth equipment parameter characterizing the vibration amplitude; A sixth device parameter characterizing current and voltage; Score equipment metrics; The capacity parameter set and the capacity index scoring formula are used to calculate the capacity index score, wherein the capacity index scoring formula is specifically: in: It is the first capacity parameter that characterizes the output; A second capacity parameter that characterizes the production cycle; is the sum of the third capacity data set representing the process time; The fourth capacity parameter characterizing material consumption; Score the capacity indicator; The quality index score is calculated using the quality parameter set and the quality index score formula, wherein the quality index score formula is specifically: in: It is the first quality parameter that characterizes the yield rate; It is the second quality parameter that characterizes dimensional accuracy; It is the third quality parameter that characterizes the surface roughness; It is the fourth quality parameter that characterizes hardness; It is the fifth quality parameter that characterizes strength; score quality indicators; The energy consumption index score is calculated using the energy consumption parameter set and the energy consumption index scoring formula, wherein the energy consumption index scoring formula is specifically: in: A first energy consumption parameter for characterizing electric energy consumption; A second energy consumption parameter to characterize water energy consumption; The third energy consumption parameter characterizing gas energy consumption; is the maximum value operator; Score energy consumption indicators; The environmental index score is calculated using the environmental parameter set and the environmental index score formula, wherein the environmental index score formula is specifically: in: It is the first environmental parameter that characterizes temperature; It is the second environmental parameter that characterizes humidity; A third environmental parameter that characterizes the noise level; It is the fourth environmental parameter that characterizes air quality; is the minimum value operator; Score environmental indicators.

[0010] Preferably, the training of the pre-trained large model to obtain the strongly correlated production index specifically includes: The strongly correlated production index is captured by using the production index of each item and the strongly correlated capture formula, wherein the strongly correlated capture formula is specifically: in: is the first constraint of the strong correlation capture formula, is the judgment result of the first constraint condition, Indicates that when the production index score is less than the preset production index threshold ,otherwise ; The second constraint for the strong correlation capture formula is, To capture the strongly correlated production indicators, Indicates when Write the corresponding production index score into the corresponding production index , otherwise it is not written and the for The placeholder in is an index symbol, indicating that The traversal index satisfies The corresponding production index score corresponds to the production index, Score production indicators, corner marks Indicates the corresponding threshold.

[0011] Preferably, the training of the pre-trained large model to obtain the associated production parameters specifically includes: Get And analyze to obtain the corresponding strongly correlated index, deconstruct the strongly correlated index to obtain the corresponding production index, deconstruct the production index to obtain the corresponding production parameter, use the preset production parameter threshold table to verify the production parameter, obtain the production parameter that fails the verification, and define all the obtained production parameters that fail the verification as the associated production parameters and then output them.

[0012] Preferably, the training of the pre-trained large model to generate the early warning data set specifically includes: Obtain the associated production parameters and obtain the previous The historical production parameters corresponding to the supervision cycles are packaged, and the packaged results are defined as the early warning data set.

[0013] Preferably, when the pre-trained large model is used to supervise industrial production, the following steps are specifically included: A1: Obtaining various real-time production indicators, wherein the real-time production indicators include corresponding real-time production parameters; A2: Using the real-time production parameters to obtain corresponding real-time production index scores; A3: Analyze the real-time production index scores to obtain real-time strongly correlated production indicators; A4: Deconstructing the real-time strongly correlated production indicators to obtain real-time correlated production parameters; A5: Generate a real-time warning data set based on the real-time associated production parameters.

[0014] In a second aspect, the present application discloses an industrial intelligent production system based on a pre-trained large model, the system is applicable to the industrial intelligent production method based on the pre-trained large model as described above, and the system includes: A pre-trained large model construction module, wherein the pre-trained large model construction module is configured to construct the pre-trained large model for full-dimensional supervision of industrial production; A pre-trained large model operation module, wherein the pre-trained large model operation module is configured to: preset a supervision cycle, and based on the supervision cycle, the pre-trained large model runs and performs the following steps: S1: Obtaining various production indicators, wherein the production indicators include corresponding production parameters; S2: using the production parameters to obtain the corresponding production index scores; S3: Analyze the production index scores to obtain strongly correlated production indicators; S4: Deconstructing the strongly correlated production indicators to obtain associated production parameters; S5: generating an early warning data set based on the associated production parameters; Among them: the production indicators also include equipment indicators, production capacity indicators, quality indicators, energy consumption indicators and environmental indicators and are used to characterize industrial production in all dimensions; the production parameters include equipment parameter sets, production capacity parameter sets, quality parameter sets, energy consumption parameter sets and environmental parameter sets and are used to comprehensively detect industrial production; the strongly correlated production indicators are the production indicators with strong correlation obtained through analysis; the associated production parameters are deconstructed to obtain the associated generation parameters; the early warning data set is used to provide the administrator with analysis data.

[0015] In a third aspect, the present application discloses a terminal device, comprising: at least one processor and at least one memory, wherein the memory is communicatively connected to the processor; the memory stores a computer program that can be executed by the processor, and when the computer program is executed by the processor, the industrial intelligent production method based on the pre-trained large model as described above is implemented.

[0016] Beneficial effects: The industrial intelligent production method, system and terminal equipment based on the pre-trained large model of the present application simplify industrial production data by accurately acquiring production indicators and production parameters, making it easier for administrators to obtain relatively high-value production data, and realizing data-driven production analysis through pre-trained large models, thereby reducing the proportion of human factors. The strongly correlated production indicators and related production parameters obtained provide a data basis for tracing the source of industrial production problems, optimize problem tracing, realize the supplementation of key data volumes and optimization of data processing methods, improve the level of industrial intelligence, and realize industrial intelligent production based on pre-trained large models. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application 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 application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A flowchart of an industrial intelligent production method based on a pre-trained large model provided in an embodiment of the present application; Figure 2 A structural block diagram of an industrial intelligent production system based on a pre-trained large model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0020] In this article, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, the elements defined by the sentence "comprising..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0021] In a first aspect, this embodiment discloses Figure 1 An industrial intelligent production method based on a pre-trained large model is shown, the method comprising: Build a pre-trained large model, preset a supervision cycle, and perform the following steps based on the pre-trained large model in the supervision cycle: S1: Obtain various production indicators, which include corresponding production parameters; S2: Use production parameters to obtain the corresponding production index scores; S3: Analyze the production index scores to obtain strongly correlated production indicators; S4: Deconstruct strongly correlated production indicators to obtain associated production parameters; S5: Generate an early warning data set based on the associated production parameters.

[0022] Among them: the pre-trained large model is used to supervise industrial production in all dimensions based on production indicators; production indicators also include equipment indicators, production capacity indicators, quality indicators, energy consumption indicators and environmental indicators and are used to characterize industrial production in all dimensions; production parameters include equipment parameter sets, production capacity parameter sets, quality parameter sets, energy consumption parameter sets and environmental parameter sets and are used to comprehensively detect industrial production; strongly correlated production indicators are production indicators with strong correlation obtained by analysis; related production parameters are deconstructed to obtain related generation parameters; early warning data sets are used to provide administrators with analyzed data.

[0023] In this regard, this embodiment simplifies the amount of data for data processing by constructing and pre-training a large model, rationalizes data processing to obtain corresponding production indicator scores, realizes quantification of industrial production processes, and provides technical support for data-driven data analysis. The strongly correlated production indicators obtained narrow the scope of investigation for problem tracing, reduce the dependence of the judgment process on human experience, and improve the level of industrial intelligence. The associated production parameters obtained further narrow the scope of investigation for problem tracing, and the early warning data set obtained provides an accurate data basis for administrators to conduct analysis, which is conducive to the development of administrators' analysis work.

[0024] Specifically, constructing the pre-trained large model includes training the pre-trained large model to obtain production indicator scores, training the pre-trained large model to obtain strongly correlated production indicators, training the pre-trained large model to obtain associated production parameters, and training the pre-trained large model to generate an early warning data set.

[0025] Specifically, the device parameter set includes device parameters for characterizing device status, operating time, processing speed, device temperature, vibration amplitude, and current and voltage.

[0026] The capacity parameter set includes capacity parameters for characterizing various items of output, production cycle, process time and material consumption.

[0027] The quality parameter set includes quality parameters for characterizing each of yield, dimensional accuracy, surface roughness, hardness and strength.

[0028] The energy consumption parameter set includes energy consumption parameters for characterizing various items of electric energy consumption, water energy consumption and gas energy consumption.

[0029] The environmental parameter set includes environmental parameters for characterizing each of temperature, humidity, noise level, and air quality.

[0030] Equipment parameters, production capacity parameters, quality parameters, energy consumption parameters and environmental parameters are all normalized data.

[0031] It should be noted that this embodiment selects representative indicators and parameters to reduce the amount of data for industrial production, but at the same time ensures the acquisition of multi-dimensional industrial production data, thereby achieving full-dimensional supervision of industrial production. It should be further noted that the normalization processing technology used in this embodiment is an existing normalization processing technology, which aims to unify the dimensions of various parameters to facilitate intelligent data-driven data analysis.

[0032] In this regard, this embodiment simplifies massive production data into production indicators and corresponding production parameters, providing a more accurate data basis for data processing.

[0033] Specifically, training the pre-trained large model obtains production indicator scores, including: The equipment index score is calculated using the equipment parameter set and the equipment index score formula, wherein the equipment index score formula is specifically: It should be noted that, in this embodiment: As the first equipment parameter to characterize the equipment status, the downtime rate is adopted; As the second equipment parameter characterizing the running time, the duration during which all the equipment on the assembly line operates normally at the same time is adopted; As the third equipment parameter characterizing the processing speed, the average processing speed of each process is adopted; For a fourth device parameter characterizing device temperature, an average value of device temperatures of normal devices within a preset temperature threshold is adopted; For the fifth device parameter characterizing the vibration amplitude, an average value of the vibration amplitude of a normal device within a preset vibration amplitude threshold is adopted; To characterize the sixth device parameter of current and voltage, an average value of current and voltage of a normal device within a preset current and voltage threshold is adopted; Score equipment metrics; In this regard, this embodiment realizes real-time evaluation of the working status of production equipment through equipment indicator scoring, ensuring the reliability and stability of the hardware foundation of industrial production.

[0034] The capacity index score is calculated using the capacity parameter set and the capacity index score formula, where the capacity index score formula is specifically: It should be noted that, in this embodiment: To characterize the first capacity parameter of output, the output value is adopted; To characterize the second capacity parameter of the production cycle, the time from the start of processing to the final product landing of a product is used; The sum of the third capacity data set representing the process time is the sum of the time taken by each process; in this embodiment, It reflects the impact of the excessive time between processes on production capacity and optimizes the data. To characterize the fourth capacity parameter of material consumption, production cost is used; Score the capacity indicator; In this regard, this embodiment realizes real-time evaluation of production capacity through production capacity index scoring, ensuring the smoothness of the industrial production process.

[0035] The quality index score is calculated using the quality parameter set and the quality index score formula, wherein the quality index score formula is specifically: It should be noted that, in this embodiment: As the first quality parameter to characterize the yield rate, the qualified rate during product acceptance is adopted; As the second quality parameter characterizing dimensional accuracy, the average value of the dimensional errors of each product is adopted; As the third quality parameter characterizing the surface roughness, the average value of the surface roughness errors of each product is adopted; To characterize the fourth quality parameter of hardness, the average value of the hardness error of each product is used; To characterize the fifth quality parameter of strength, the average value of the strength errors of each product is used; score quality indicators; In this regard, this embodiment achieves real-time evaluation of product quality through quality indicator scoring, thereby ensuring the quality of industrially produced finished products.

[0036] The energy consumption index score is calculated using the energy consumption parameter set and the energy consumption index scoring formula, wherein the energy consumption index scoring formula is specifically: It should be noted that, in this embodiment: As a first energy consumption parameter characterizing the power consumption, the difference between the power consumption of the current supervision cycle and the previous supervision cycle is adopted; To characterize the second energy consumption parameter of water energy consumption, the difference between the water energy consumption in the current supervision cycle and the previous supervision cycle is used; To characterize the third energy consumption parameter of gas energy consumption, the difference between the gas energy consumption of the current supervision cycle and the previous supervision cycle is used; is the maximum value operator; Score the energy consumption index; in actual industry, it is understandable that energy consumption is basically unchanged if the production plan does not change significantly. Therefore, collecting the difference is more conducive to reflecting possible production anomalies; In this regard, this embodiment achieves effective evaluation of energy consumption in industrial production processes through energy consumption index scoring, which is helpful in improving energy planning of industrial production processes and production supervision from an energy dimension.

[0037] The environmental index score is calculated using the environmental parameter set and the environmental index score formula, where the environmental index score formula is specifically: It should be noted that, in this embodiment: As the first environmental parameter representing the temperature, the difference between the ambient temperature of the current supervision cycle and the previous supervision cycle is used; As a second environmental parameter representing humidity, the difference between the environmental humidity of the current supervision cycle and the previous supervision cycle is used; As the third environmental parameter representing the noise level, the difference between the environmental noise level in the current supervision cycle and the previous supervision cycle is used; For the fourth environmental parameter representing air quality, the difference between the ambient air quality of the current regulatory cycle and the previous regulatory cycle is used; is the minimum value operator; Scoring environmental indicators. In actual industry, it is understandable that if there is no significant change in production technology or equipment, environmental parameters should also remain basically unchanged. Therefore, collecting differences is more conducive to reflecting possible production anomalies.

[0038] In this regard, this embodiment achieves effective assessment of the industrial production process environment through environmental indicator scoring, which is conducive to production supervision from an environmental dimension.

[0039] Specifically, the pre-trained large model is trained to obtain strongly related production indicators, including: The strongly correlated production indicators are captured using the production indicators of each item and the strongly correlated capture formula, where the strongly correlated capture formula is specifically: in: is the first constraint of the strong correlation capture formula, is the judgment result of the first constraint condition, Indicates that when the production index score is less than the preset production index threshold ,otherwise ; The second constraint for the strong correlation capture formula is, The strongly correlated production index captured is the matrix and is determined by The value of the matrix is ​​written in real time. Indicates when Write the corresponding production index score into the corresponding production index , otherwise it is not written and the for The placeholder in is an index symbol, indicating that The traversal index satisfies The corresponding production index score corresponds to the production index, Score production indicators, corner marks Indicates the corresponding threshold. Exemplarily, when the environmental index score and the equipment index score are less than the preset environmental index threshold and the equipment index threshold, It includes environmental indicators and equipment indicators.

[0040] It should be noted that, within the scope that can be understood by technical personnel in this field, for the same production process, there is a correlation between the various production indicators. Traditional industrial production often captures the correlation based on long-term experience, but this is not conducive to the improvement of the level of industrial intelligence. Therefore, this embodiment uses a strong correlation capture formula based on various production indicators to realize data-driven analysis of industrial production data, optimizes the analysis method of industrial data, reduces the proportion of human factors, and improves the level of industrial intelligence.

[0041] Specifically, the pre-trained large model is trained to obtain associated production parameters, including: Get And analyze to obtain the corresponding strongly correlated index, deconstruct the strongly correlated index to obtain the corresponding production index, deconstruct the production index to obtain the corresponding production parameter, and use the preset production parameter threshold table to verify the production parameter. The table is obtained based on the threshold of the production parameter known to those skilled in the art, and is suitable for the verification of the normalized data of this embodiment. In this embodiment, the verification method adopted is the well-known technical means of comparing with the threshold, and the production parameters that failed the verification are obtained and all the obtained production parameters that failed the verification are defined as associated production parameters and then output. Exemplarily, the strongly correlated production indicators are equipment indicators and quality indicators, and the corresponding deconstruction obtains equipment parameters and quality parameters, and the preset equipment parameter threshold and the preset quality parameter threshold are used to verify the equipment parameters and the quality parameters.

[0042] Specifically, training a pre-trained large model to generate an early warning data set includes: Get the associated production parameters and the corresponding previous The historical production parameters corresponding to each supervision cycle are packaged and the packaged results are defined as an early warning data set.

[0043] Specifically, when using a pre-trained large model to supervise industrial production, the following steps are included: A1: Obtain various real-time production indicators, which include corresponding real-time production parameters; A2: Use real-time production parameters to obtain corresponding real-time production index scores; A3: Analyze the real-time production index scores to obtain real-time strongly correlated production indicators; A4: Deconstruct real-time strongly correlated production indicators to obtain real-time correlated production parameters; A5: Generate real-time warning data sets based on real-time associated production parameters.

[0044] In this regard, this embodiment realizes full-dimensional and real-time supervision of industrial production.

[0045] In a second aspect, this embodiment discloses Figure 2 An industrial intelligent production system based on a pre-trained large model is shown, and the system is applicable to the industrial intelligent production method based on a pre-trained large model as described above, and the system includes: The pre-trained large model construction module is configured to build a pre-trained large model for full-dimensional supervision of industrial production.

[0046] The pre-trained large model operation module is configured as follows: a preset supervision cycle is set, and the pre-trained large model is run based on the supervision cycle and the following steps are performed: S1: Obtain various production indicators, which include corresponding production parameters; S2: Use production parameters to obtain the corresponding production index scores; S3: Analyze the production index scores to obtain strongly correlated production indicators; S4: Deconstruct strongly correlated production indicators to obtain associated production parameters; S5: Generate an early warning data set based on the associated production parameters.

[0047] Among them: production indicators also include equipment indicators, production capacity indicators, quality indicators, energy consumption indicators and environmental indicators and are used to characterize industrial production in all dimensions; production parameters include equipment parameter sets, production capacity parameter sets, quality parameter sets, energy consumption parameter sets and environmental parameter sets and are used to comprehensively detect industrial production; strongly correlated production indicators are production indicators with strong correlation obtained through analysis; related production parameters are deconstructed to obtain related generation parameters; early warning data sets are used to provide administrators with analysis data.

[0048] It should be noted that the industrial intelligent production system based on the pre-trained large model of this embodiment corresponds to the industrial intelligent production method based on the pre-trained large model in the previous text. Therefore, the contents not specifically described in the industrial intelligent production system based on the pre-trained large model of this embodiment may have, but are not limited to, the same functional definition, working principle and technical effect as the industrial intelligent production method based on the pre-trained large model in the previous text, and this text will not go into details here.

[0049] In a third aspect, this embodiment discloses a terminal device, comprising: at least one processor and at least one memory, wherein the memory is communicatively connected to the processor; a computer program that can be executed by the processor is stored in the memory, and when the computer program is executed by the processor, the industrial intelligent production method based on the pre-trained large model as described above is implemented.

[0050] In summary, the industrial intelligent production method, system and terminal equipment based on the pre-trained large model of this embodiment simplify industrial production data by accurately acquiring production indicators and production parameters, making it easier for administrators to obtain relatively high-value production data, and realizing data-driven production analysis through pre-trained large models, thereby reducing the proportion of human factors. The strongly correlated production indicators and related production parameters obtained provide a data basis for tracing the source of industrial production problems, optimize problem tracing, realize the supplementation of key data volumes and optimization of data processing methods, improve the level of industrial intelligence, and realize industrial intelligent production based on pre-trained large models.

[0051] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiment can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that can be accessed by a computer. Computer-readable storage media can include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer.

[0052] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. An industrial intelligent production method based on a pre-trained large model, characterized in that: The method includes: Construct a pre-trained large model, preset a supervision cycle, and perform the following steps based on the pre-trained large model in the supervision cycle: S1: Obtaining various production indicators, wherein the production indicators include corresponding production parameters; S2: using the production parameters to obtain the corresponding production index scores; S3: Analyze the production index scores to obtain strongly correlated production indicators; S4: Deconstructing the strongly correlated production indicators to obtain associated production parameters; S5: generating an early warning data set based on the associated production parameters; Among them: the pre-trained large model is used to supervise industrial production in all dimensions based on the production indicators; the production indicators also include equipment indicators, production capacity indicators, quality indicators, energy consumption indicators and environmental indicators and are used to characterize industrial production in all dimensions; the production parameters include equipment parameter sets, production capacity parameter sets, quality parameter sets, energy consumption parameter sets and environmental parameter sets and are used to comprehensively detect industrial production; the strongly correlated production indicators are the production indicators with strong correlation obtained through analysis; the associated production parameters are deconstructed to obtain the associated generation parameters; the early warning data set is used to provide analyzed data to administrators.

2. The industrial intelligent production method based on the pre-trained large model according to claim 1 is characterized in that: The construction of the pre-trained large model includes training the pre-trained large model to obtain the production indicator score, training the pre-trained large model to obtain the strongly correlated production indicator, training the pre-trained large model to obtain the associated production parameters and training the pre-trained large model to generate the early warning data set.

3. The industrial intelligent production method based on the pre-trained large model according to claim 1 is characterized in that: The device parameter set includes device parameters for characterizing device status, operating time, processing speed, device temperature, vibration amplitude, and current and voltage; The capacity parameter set includes capacity parameters for characterizing output, production cycle, process time and material consumption; The quality parameter set includes quality parameters for characterizing each item of yield rate, dimensional accuracy, surface roughness, hardness and strength; The energy consumption parameter set includes energy consumption parameters for characterizing various items of electricity consumption, water consumption and gas consumption; The environmental parameter set includes environmental parameters for characterizing each of temperature, humidity, noise level and air quality; The equipment parameters, the production capacity parameters, the quality parameters, the energy consumption parameters and the environmental parameters are all normalized data.

4. The industrial intelligent production method based on the pre-trained large model according to claim 3 is characterized in that: The training of the pre-trained large model to obtain the production index score specifically includes: The device parameter set and the device indicator scoring formula are used to calculate the device indicator score, wherein the device indicator scoring formula is specifically: in: A first device parameter that characterizes a device state; A second device parameter characterizing the operating time; A third equipment parameter that characterizes the processing speed; A fourth device parameter characterizing device temperature; A fifth equipment parameter characterizing the vibration amplitude; A sixth device parameter characterizing current and voltage; Score equipment metrics; The capacity parameter set and the capacity index scoring formula are used to calculate the capacity index score, wherein the capacity index scoring formula is specifically: in: It is the first capacity parameter that characterizes the output; A second capacity parameter that characterizes the production cycle; is the sum of the third capacity data set representing the process time; The fourth capacity parameter characterizing material consumption; Score the capacity indicator; The quality index score is calculated using the quality parameter set and the quality index score formula, wherein the quality index score formula is specifically: in: It is the first quality parameter that characterizes the yield rate; It is the second quality parameter that characterizes dimensional accuracy; It is the third quality parameter that characterizes the surface roughness; It is the fourth quality parameter that characterizes hardness; It is the fifth quality parameter that characterizes strength; score quality indicators; The energy consumption index score is calculated using the energy consumption parameter set and the energy consumption index scoring formula, wherein the energy consumption index scoring formula is specifically: in: A first energy consumption parameter for characterizing electric energy consumption; A second energy consumption parameter to characterize water energy consumption; The third energy consumption parameter characterizing gas energy consumption; is the maximum value operator; Score energy consumption indicators; The environmental index score is calculated using the environmental parameter set and the environmental index score formula, wherein the environmental index score formula is specifically: in: It is the first environmental parameter that characterizes temperature; It is the second environmental parameter that characterizes humidity; A third environmental parameter that characterizes the noise level; It is the fourth environmental parameter that characterizes air quality; is the minimum value operator; Score environmental indicators.

5. The industrial intelligent production method based on the pre-trained large model according to claim 4 is characterized in that: The training of the pre-trained large model to obtain the strongly correlated production index specifically includes: The strongly correlated production index is captured by using the production index of each item and the strongly correlated capture formula, wherein the strongly correlated capture formula is specifically: in: is the first constraint of the strong correlation capture formula, is the judgment result of the first constraint condition, Indicates that when the production index score is less than the preset production index threshold ,otherwise ; The second constraint for the strong correlation capture formula is, To capture the strongly correlated production indicators, Indicates when Write the corresponding production index score into the corresponding production index , otherwise it is not written and the for The placeholder in is an index symbol, indicating that The traversal index satisfies The corresponding production index score corresponds to the production index, Score production indicators, corner marks Indicates the corresponding threshold.

6. The industrial intelligent production method based on the pre-trained large model according to claim 5 is characterized in that: The training of the pre-trained large model to obtain the associated production parameters specifically includes: Get And analyze to obtain the corresponding strongly correlated index, deconstruct the strongly correlated index to obtain the corresponding production index, deconstruct the production index to obtain the corresponding production parameter, use the preset production parameter threshold table to verify the production parameter, obtain the production parameter that fails the verification, and define all the obtained production parameters that fail the verification as the associated production parameters and then output them.

7. The industrial intelligent production method based on the pre-trained large model according to claim 2 is characterized in that: The training of the pre-trained large model to generate the early warning data set specifically includes: Obtain the associated production parameters and obtain the previous The historical production parameters corresponding to the supervision cycles are packaged, and the packaged results are defined as the early warning data set.

8. The industrial intelligent production method based on the pre-trained large model according to claim 1 is characterized in that: When the pre-trained large model is used to supervise industrial production, the following steps are specifically included: A1: Obtaining various real-time production indicators, wherein the real-time production indicators include corresponding real-time production parameters; A2: Using the real-time production parameters to obtain corresponding real-time production index scores; A3: Analyze the real-time production index scores to obtain real-time strongly correlated production indicators; A4: Deconstructing the real-time strongly correlated production indicators to obtain real-time correlated production parameters; A5: Generate a real-time warning data set based on the real-time associated production parameters.

9. An industrial intelligent production system based on a pre-trained large model, the system is applicable to the industrial intelligent production method based on a pre-trained large model as described in any one of claims 1 to 8, characterized in that: The system includes: A pre-trained large model construction module, wherein the pre-trained large model construction module is configured to construct the pre-trained large model for full-dimensional supervision of industrial production; A pre-trained large model operation module, wherein the pre-trained large model operation module is configured to: preset a supervision cycle, and based on the supervision cycle, the pre-trained large model runs and performs the following steps: S1: Obtaining various production indicators, wherein the production indicators include corresponding production parameters; S2: using the production parameters to obtain the corresponding production index scores; S3: Analyze the production index scores to obtain strongly correlated production indicators; S4: Deconstructing the strongly correlated production indicators to obtain associated production parameters; S5: generating an early warning data set based on the associated production parameters; Among them: the production indicators also include equipment indicators, production capacity indicators, quality indicators, energy consumption indicators and environmental indicators and are used to characterize industrial production in all dimensions; the production parameters include equipment parameter sets, production capacity parameter sets, quality parameter sets, energy consumption parameter sets and environmental parameter sets and are used to comprehensively detect industrial production; the strongly correlated production indicators are the production indicators with strong correlation obtained through analysis; the associated production parameters are deconstructed to obtain the associated generation parameters; the early warning data set is used to provide the administrator with analysis data.

10. A terminal device, characterized in that: include: at least one processor and at least one memory, the memory being communicatively connected to the processor; The memory stores a computer program that can be executed by the processor. When the computer program is executed by the processor, the industrial intelligent production method based on the pre-trained large model as described in any one of claims 1 to 8 is implemented.

Images