A large model-based carbon black production equipment health management method, system, device and medium

Abstract

The application provides a kind of big model-based carbon black production equipment health management method, system, equipment and medium, belongs to industrial internet of things technical field, the method: through internet of things sensor, the equipment key parameters of carbon black production equipment, process parameter and environmental parameter of process pipeline are collected, feature is extracted, and structured feature vector is generated;Based on TFT model, construct and train carbon black production equipment health management big model;Real-time structured feature vector is input into carbon black production equipment health management big model, and equipment health evaluation index is calculated based on output;The extracted features are synchronized to the twin model by the edge computing node at regular intervals, and a digital twin is constructed;Based on digital twin, a health status board is generated, and the equipment health evaluation index, the output of the carbon black production equipment health management big model and the extracted features are visually displayed.The application realizes intelligent monitoring and prediction of carbon black equipment, accurately evaluates the health status, and reduces the risk of unplanned shutdown.

Description

Technical Field

[0001] This application belongs to the field of industrial Internet of Things technology, specifically relating to a method, system, equipment and medium for health management of carbon black production equipment based on a large model. Background Technology

[0002] In the carbon black production sector, the production process is complex, and the Industrial Internet of Things (IIoT) uses sensor networks to monitor the operation of key equipment such as reactors, crushers, and coolers. However, traditional health management methods have the following drawbacks:

[0003] First, traditional carbon black production equipment maintenance relies heavily on manual experience. Manual diagnosis of equipment faults is not only highly subjective and prone to errors, but also demands extremely high levels of professional expertise. For example, relying solely on equipment operating sounds and temperatures can easily overlook minor issues, often leading to equipment damage and disruptions to production schedules by the time they are discovered. Second, existing equipment fault warning systems are mostly based on simple threshold judgments, failing to capture the complex time-delay correlations between equipment status and process parameters. For instance, if the reactor temperature is abnormal, current systems only issue over-temperature alarms, unable to predict faults in advance based on changes in process parameters, resulting in significant equipment damage and downtime losses. Finally, equipment operating data in carbon black production is complex and multi-sourced, and current data is only processed superficially. For example, vibration signals are only used for vibration amplitude monitoring, lacking sufficient correlation with process parameters, resulting in wasted data value and failing to provide strong support for equipment health assessment and fault diagnosis.

[0004] In summary, traditional carbon black production equipment management mainly relies on manual experience for equipment maintenance, lacking accurate real-time monitoring of equipment status and fault early warning, resulting in low efficiency in equipment health management and failing to meet the needs of modern intelligent industrial production. Summary of the Invention

[0005] In a first aspect, embodiments of this application provide a method for health management of carbon black production equipment based on a large model, comprising the following steps:

[0006] S1. Collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors;

[0007] S2. Construct and train a large-scale health management model for carbon black production equipment based on the TFT model:

[0008] The input is a structured feature vector;

[0009] The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module.

[0010] Outputs failure probability, process parameter deviation, and health prediction score;

[0011] S3. Input the real-time structured feature vector into the large-scale health management model for carbon black production equipment, and calculate the equipment health assessment indicators based on the output: failure probability, current health score and health change rate;

[0012] S4. The extracted features are periodically synchronized to the twin model through edge computing nodes to construct a digital twin;

[0013] S5. Based on the digital twin, a health status dashboard is generated, which visualizes the output of equipment health assessment indicators, the large-scale health management model of carbon black production equipment, and the extracted features.

[0014] Furthermore, the specific steps of step S1 are as follows:

[0015] S11. Deploy an infrared thermal imager on the outer wall of the reactor to collect the temperature field matrix. ;

[0016] Where H is the number of rows in the matrix and W is the number of columns in the matrix;

[0017] S12. Vibration sensors are embedded in the bearing housings of the crusher and cooler to collect vibration acceleration signals. ;

[0018] Where N is the number of samples;

[0019] S13. Deploy flow meters, oxygen analyzers, and temperature sensors in the process pipeline to collect process parameter vectors. ;

[0020] S14. Deploy dust sensors in the carbon black production workshop to collect environmental data:

[0021] ;

[0022] S15. Deploy edge computing nodes in the carbon black production workshop to perform feature extraction on the data collected by various IoT sensors:

[0023] Time-domain characteristics of vibration acceleration signals from crushers and coolers: mean and kurtosis ;

[0024] After applying a Hanning window to the vibration acceleration signals of the crusher and cooler, an FFT was performed to calculate the frequency domain characteristics: FFT peak value. ;

[0025] Define the overheated region for the temperature field of the reactor. Calculate the proportion of overheated regions and construct statistical characteristics of the temperature field. ;

[0026] in, It is the reference temperature. It is the standard deviation of the temperature field;

[0027] S16. Generate structured feature vectors ;

[0028] Where P is the process parameter, It is the mean of the vibration acceleration signal. Indicates the kurtosis of the vibration acceleration signal. Indicates the peak value of the FFT. This represents the statistical characteristics of the temperature field. This represents environmental data.

[0029] Furthermore, the specific steps of step S2 are as follows:

[0030] S21. Construct a large-scale health management model for carbon black production equipment based on the TFT model;

[0031] S22. Set structured feature vectors For input;

[0032] S23. The delayed attention module is implemented by adding a time delay unit to the encoding layer of the TFT model and by gating the fusion of delayed features and real-time features;

[0033] S24. Set the classification of faults and the probability of each type of fault as the first output, calculate the deviation of each parameter from the parameter standard value as the second output, and use the health prediction score for the future set period as the third output;

[0034] S25. Train the large-scale health management model for carbon black production equipment and distribute it to edge computing nodes.

[0035] Furthermore, the specific steps of step S25 are as follows:

[0036] S251. Collect structural feature vectors of carbon black production equipment under various operating conditions such as cold start, hot start, normal shutdown and emergency shutdown, as well as various load changes, to construct a dataset;

[0037] S252. Label the dataset with normal samples and fault samples, and set the fault samples to include the process parameters within the time range [-t, t] before and after the change of process parameters;

[0038] S253. Construct a multi-task loss function;

[0039]

[0040] in, , , These are preset weight parameters. It is the loss due to fault classification. It is the loss due to deviation of process parameters. It is the loss in predicting health scores;

[0041] S254. Use the dataset to train a large model for the health management of carbon black production equipment, and use the Adam optimizer to optimize it during the training process until the loss function converges or the maximum number of iterations is met.

[0042] Furthermore, the specific steps of step S3 are as follows:

[0043] S31. Input the structured feature vector generated from the real-time collected data into the trained carbon black production equipment health management model to obtain the failure probability, process parameter deviation and health prediction score of each type of fault.

[0044] S32. Filter out the fault type with the highest failure probability and calculate the current health score:

[0045]

[0046] in, That is the highest probability of failure. It is the comprehensive deviation of industrial parameters, calculated from the deviation of process parameters output by the model; , These are preset weight parameters;

[0047] S33. Calculate the current health score based on the health prediction score:

[0048]

[0049] Where k is the time window.

[0050] Furthermore, the specific steps of step S4 are as follows:

[0051] S41. Edge computing nodes select features from the extracted features whose mutual information is higher than a preset mutual information threshold;

[0052]

[0053] in, Features With health score mutual information, It is a preset mutual information threshold;

[0054] S42. After compressing the selected features according to a preset ratio, transmit them to the cloud using the MQTT protocol;

[0055] S43. Building digital twins in the cloud:

[0056] Statistical characteristics of temperature field Mapped to a surface thermal map of a 3D model of the reactor;

[0057] Environmental data Convert to visual elements;

[0058] S44. Set current health score Current health score The probability of failure is bound to the twin's metadata to complete the injection of health indicators.

[0059] Furthermore, the specific steps of step S5 are as follows:

[0060] S51. Based on digital twins, construct equipment health status dashboards to visualize equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features.

[0061] S52. The equipment health status dashboard integrates process parameters, vibration acceleration signals, and synchronously displays equipment status and environmental data.

[0062] S53. Map the current health score to the color of the device twin model to display the health status of the device, and overlay the probability distribution of fault types onto the surface of the device twin model.

[0063] Secondly, embodiments of this application also provide a carbon black production equipment health management system based on a large model, including:

[0064] The data acquisition and feature extraction module is used to collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and to perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors.

[0065] The model building and training module is used to build and train a large-scale health management model for carbon black production equipment based on the TFT model.

[0066] The input is a structured feature vector;

[0067] The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module.

[0068] Outputs failure probability, process parameter deviation, and health prediction score;

[0069] The health assessment module is used to input real-time structured feature vectors into the large-scale health management model of carbon black production equipment, and calculate equipment health assessment indicators based on the output: failure probability, current health score and health change rate.

[0070] The digital twin construction module is used to periodically synchronize extracted features to the twin model via edge computing nodes to construct a digital twin.

[0071] The visualization module is used to generate a health status dashboard based on the digital twin, and to visualize the equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features.

[0072] Thirdly, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the carbon black production equipment health management method based on a large model as described in the first aspect.

[0073] Fourthly, embodiments of this application also provide a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the carbon black production equipment health management method based on a large model as described in the first aspect.

[0074] As can be seen from the above technical solutions, this application has the following advantages:

[0075] The health management method, system, equipment, and media for carbon black production equipment based on a large model provided in this application integrate the Internet of Things, large models, and digital twins to transform real-time data of carbon black production equipment into visualized indicators, enabling rapid location of potential faults and presenting predictions and health trends in a dynamic and visualized form to assist managers in anticipating risks. Ultimately, it achieves prediction of remaining equipment lifespan and intelligent optimization of maintenance strategies, reducing the risk of unplanned downtime and improving carbon black production efficiency and reliability. Attached Figure Description

[0076] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0077] Figure 1 This is a schematic diagram of the process for the health management method of carbon black production equipment based on a large model according to the present invention.

[0078] Figure 2 This is a schematic diagram of the health management system for carbon black production equipment based on a large model according to the present invention. Detailed Implementation

[0079] The various embodiments of this disclosure will be described more fully in the detailed steps of the health management method for carbon black production equipment based on a large model described below. This disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of this disclosure to the specific embodiments disclosed herein, but rather this disclosure should be understood to cover all adjustments, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of this disclosure.

[0080] For example, in the field of carbon black production, although the Industrial Internet of Things (IIoT) has achieved a certain degree of operational monitoring of key equipment such as reactors, crushers, and coolers through the deployment of sensor networks, traditional methods for health management of carbon black production equipment still have many drawbacks, mainly in the following aspects:

[0081] First, traditional equipment maintenance methods rely excessively on human experience. Since manual judgment of equipment malfunctions is highly subjective and prone to misjudgment, it also requires extremely high levels of professional expertise, resulting in significant labor costs. For example, relying solely on the sound and temperature of the equipment during operation often leads to the overlooking of minor potential problems, resulting in equipment damage and impacting production schedules by the time they are discovered. Second, current equipment malfunction early warning systems are mostly based on simple threshold judgments, failing to capture the complex time-delay correlation between equipment status and process parameters. For instance, in the case of abnormal reactor temperature, existing systems only issue alarms when the temperature exceeds the limit, unable to predict malfunctions in advance based on changes in process parameters, leading to equipment damage and substantial downtime losses. Third, carbon black production equipment generates a large amount of complex, multi-source data during operation, but currently, the collected data undergoes only simple processing, failing to fully realize its value. For example, vibration signals are typically only used to monitor vibration amplitude without deep correlation with process parameters, resulting in a serious waste of data value and failing to provide strong support for equipment health assessment and fault diagnosis.

[0082] In summary, traditional carbon black production equipment management mainly relies on manual experience and lacks accurate monitoring of the real-time status of equipment and effective fault early warning methods. This makes equipment health management inefficient, fails to meet the requirements of modern intelligent industrial production, and seriously affects the production efficiency and economic benefits of carbon black production enterprises.

[0083] To address the aforementioned issues, this embodiment provides a health management method for carbon black production equipment based on a large model, enabling intelligent monitoring and prediction of carbon black equipment, accurately assessing its health status, and reducing the risk of unplanned downtime.

[0084] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0085] Please see Figure 1 The diagram shows a flowchart of a health management method for carbon black production equipment based on a large model in a specific embodiment. The method includes the following steps:

[0086] S1. Collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors;

[0087] It should be noted that real-time monitoring of equipment operating status and preliminary data processing were achieved, providing data support for subsequent model analysis;

[0088] S2. Construct and train a large-scale health management model for carbon black production equipment based on the TFT model:

[0089] The input is a structured feature vector;

[0090] The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module.

[0091] Outputs failure probability, process parameter deviation, and health prediction score;

[0092] It should be noted that this enables the model to learn the complex relationships between equipment status and process parameters, making accurate fault prediction and health assessment possible.

[0093] S3. Input the real-time structured feature vector into the large-scale health management model for carbon black production equipment, and calculate the equipment health assessment indicators based on the output: failure probability, current health score and health change rate;

[0094] It should be noted that real-time assessment of equipment health status has been achieved, helping managers to understand the equipment's operating status in a timely manner;

[0095] S4. The extracted features are periodically synchronized to the twin model through edge computing nodes to construct a digital twin;

[0096] It should be noted that, as a virtual mirror image of a device, a digital twin can reflect the actual operating status of the device in real time, providing an intuitive visualization tool for device management;

[0097] S5. Based on the digital twin, a health status dashboard is generated, which visualizes the output of equipment health assessment indicators, the large-scale health management model for carbon black production equipment, and the extracted features.

[0098] It should be noted that this allows the health status and related data of the equipment to be presented to the management personnel in an intuitive way.

[0099] This embodiment integrates the Internet of Things, large-scale models, and digital twins to achieve real-time monitoring, fault warning, and accurate health status assessment of carbon black production equipment, thereby reducing the risk of unplanned downtime, improving production efficiency and equipment reliability, and reducing reliance on manual labor.

[0100] Furthermore, as a refinement and extension of the specific implementation methods of the above embodiments, in order to fully illustrate the specific implementation process of this embodiment, another method for health management of carbon black production equipment based on a large model is provided. This method includes the following steps:

[0101] S1. Collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors;

[0102] The specific steps of step S1 are as follows:

[0103] S11. Deploy an infrared thermal imager on the outer wall of the reactor to collect the temperature field matrix. ;

[0104] Where H is the number of rows in the matrix and W is the number of columns in the matrix;

[0105] S12. Vibration sensors are embedded in the bearing housings of the crusher and cooler to collect vibration acceleration signals. ;

[0106] Where N is the number of samples;

[0107] S13. Deploy flow meters, oxygen analyzers, and temperature sensors in the process pipeline to collect process parameter vectors. ;

[0108] S14. Deploy dust sensors in the carbon black production workshop to collect environmental data:

[0109] ;

[0110] S15. Deploy edge computing nodes in the carbon black production workshop to perform feature extraction on the data collected by various IoT sensors:

[0111] Time-domain characteristics of vibration acceleration signals from crushers and coolers: mean and kurtosis ;

[0112]

[0113]

[0114] in, Indicates vibration acceleration signal, Indicates the number of samples. This represents the mean value of the vibration acceleration signal. Indicates the kurtosis of the vibration acceleration signal;

[0115] After applying a Hanning window to the vibration acceleration signals of the crusher and cooler, an FFT was performed to calculate the frequency domain characteristics: FFT peak value. ;

[0116]

[0117] in, For Hanning window, It is a Fourier transform;

[0118] Define the overheated region for the temperature field of the reactor. Calculate the proportion of overheated area Construct statistical features of the temperature field ;

[0119]

[0120] in, It is a maximum value function. It is a function of standard deviation; It is the reference temperature. It is the standard deviation of the temperature field;

[0121] S16. Generate structured feature vectors ;

[0122] Where P is the process parameter, It is the mean of the vibration acceleration signal. Indicates the kurtosis of the vibration acceleration signal. Indicates the peak value of the FFT. This represents the statistical characteristics of the temperature field. Represents environmental data;

[0123] S2. Construct and train a large-scale health management model for carbon black production equipment based on the TFT model:

[0124] The input is a structured feature vector;

[0125] The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module.

[0126] Outputs failure probability, process parameter deviation, and health prediction score;

[0127] The specific steps of step S2 are as follows:

[0128] S21. Construct a large-scale health management model for carbon black production equipment based on the TFT model;

[0129] S22. Set structured feature vectors For input;

[0130] S23. The delayed attention module is implemented by adding a time delay unit to the encoding layer of the TFT model and by gating the fusion of delayed features and real-time features;

[0131] Specifically as follows:

[0132]

[0133] in, These are process parameters used as query parameters. , These are key device parameter characteristics that serve as key-value parameters. ;

[0134] It is the attention head dimension. , It is a learnable weight matrix. It is a projectible learning matrix; It is the query vector dimension. It is the dimension of the key-value vector;

[0135] It is a hysteresis mask matrix;

[0136] It is a time delay parameter. It is the optimal time delay parameter;

[0137] Only allow attention Feature correlation within a time window;

[0138] Optimal time delay parameters Determined by maximizing the target parameter through mutual information:

[0139]

[0140] in, Mutual information function, which measures the statistical correlation between two variables; It is currently lagging behind. Process parameters; This represents the mean value of the current vibration acceleration signal. This indicates the kurtosis of the current vibration acceleration signal. Indicates the current FFT peak value;

[0141] Specifically, , It reflects the process conditions at a certain point in the past and is used to predict the current state of the equipment;

[0142] S24. Set the classification of faults and the probability of each type of fault as the first output, calculate the deviation of each parameter from the parameter standard value as the second output, and use the health prediction score for the future set period as the third output;

[0143] S25. Train the large-scale health management model for carbon black production equipment and distribute it to edge computing nodes;

[0144] The specific steps of step S25 are as follows:

[0145] S251. Collect structural feature vectors of carbon black production equipment under various operating conditions such as cold start, hot start, normal shutdown and emergency shutdown, as well as various load changes, to construct a dataset;

[0146] S252. Label the dataset with normal samples and fault samples, and set the fault samples to include the process parameters within the time range [-t, t] before and after the change of process parameters;

[0147] S253. Construct a multi-task loss function;

[0148]

[0149] in, , , These are preset weight parameters. It is the loss due to fault classification. It is the loss due to deviation of process parameters. It is the loss in predicting health scores;

[0150] S254. Use the dataset to train a large model for health management of carbon black production equipment, and use the Adam optimizer to optimize it during the training process until the loss function converges or the maximum number of iterations is met;

[0151] S3. Input the real-time structured feature vector into the large-scale health management model for carbon black production equipment, and calculate the equipment health assessment indicators based on the output: failure probability, current health score and health change rate;

[0152] The specific steps of step S3 are as follows:

[0153] S31. Input the structured feature vector generated from the real-time collected data into the trained carbon black production equipment health management model to obtain the failure probability, process parameter deviation and health prediction score of each type of fault.

[0154] S32. Filter out the fault type with the highest failure probability and calculate the current health score:

[0155]

[0156] in, That is the highest probability of failure. It is the comprehensive deviation of industrial parameters, calculated from the deviation of process parameters output by the model; , These are preset weight parameters;

[0157] For example, For a healthy state, Yellow alert triggered Orange alert triggered Red alert triggered;

[0158] S33. Calculate the current health score based on the health prediction score:

[0159]

[0160] Where k is the time window;

[0161] For example, k can be 12, if A red alert is triggered at the minute / hour;

[0162] S4. The extracted features are periodically synchronized to the twin model through edge computing nodes to construct a digital twin;

[0163] The specific steps of step S4 are as follows:

[0164] S41. Edge computing nodes select features from the extracted features whose mutual information is higher than a preset mutual information threshold;

[0165]

[0166] in, Features With health score mutual information, It is a preset mutual information threshold;

[0167] S42. After compressing the selected features according to a preset ratio, transmit them to the cloud using the MQTT protocol;

[0168] S43. Building digital twins in the cloud:

[0169] Statistical characteristics of temperature field Mapped to a surface thermal map of a 3D model of the reactor;

[0170] Environmental data Convert to visual elements;

[0171] Specifically, dust concentration is mapped to particle flow density, and cumulative runtime is mapped to equipment wear texture;

[0172] S44. Set current health score Current health score And the failure probability is bound to the twin's metadata to complete the injection of health indicators;

[0173] S5. Based on the digital twin, a health status dashboard is generated, which visualizes the output of equipment health assessment indicators, the large-scale health management model for carbon black production equipment, and the extracted features.

[0174] The specific steps of step S5 are as follows:

[0175] S51. Based on digital twins, construct equipment health status dashboards to visualize equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features.

[0176] S52. The equipment health status dashboard integrates process parameters, vibration acceleration signals, and synchronously displays equipment status and environmental data.

[0177] S53. Map the current health score to the color of the device twin model to display the health status of the device, and overlay the probability distribution of fault types onto the surface of the device twin model.

[0178] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0179] like Figure 2 As shown, the following are embodiments of the carbon black production equipment health management system based on a large model provided in this disclosure. This system and the carbon black production equipment health management method based on a large model in the above embodiments belong to the same inventive concept. For details not described in detail in the embodiments of the carbon black production equipment health management system based on a large model, please refer to the embodiments of the carbon black production equipment health management method based on a large model described above.

[0180] The system includes:

[0181] The data acquisition and feature extraction module is used to collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and to perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors.

[0182] The model building and training module is used to build and train a large-scale health management model for carbon black production equipment based on the TFT model.

[0183] The input is a structured feature vector;

[0184] The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module.

[0185] Outputs failure probability, process parameter deviation, and health prediction score;

[0186] The health assessment module is used to input real-time structured feature vectors into the large-scale health management model of carbon black production equipment, and calculate equipment health assessment indicators based on the output: failure probability, current health score and health change rate.

[0187] The digital twin construction module is used to periodically synchronize extracted features to the twin model via edge computing nodes to construct a digital twin.

[0188] The visualization module is used to generate a health status dashboard based on the digital twin, and to visualize the equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features.

[0189] This embodiment effectively solves the problems of heavy reliance on manual experience, delayed fault warning, and insufficient data utilization in the management of traditional carbon black production equipment through the interactive collaboration of data acquisition and feature extraction modules, model construction and training modules, health assessment modules, digital twin construction modules, and visualization display modules, thus promoting the intelligent development of carbon black production.

[0190] The large-model-based carbon black production equipment health management method provided in this application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiments of this invention does not constitute a limitation on the electronic device. An electronic device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, the electronic device includes, but is not limited to, laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0191] Electronic devices may include processors, external memory interfaces, internal memory, universal serial bus (USB) interfaces, charging management modules, power management modules, batteries, wireless communication modules, audio modules, speakers, microphones, sensor modules, buttons, cameras, displays, and SIM card interfaces, etc.

[0192] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the electronic device. In other embodiments of this application, the electronic device may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0193] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0194] The processor can serve as the nerve center and command center of an electronic device. The controller can generate operation control signals based on the instruction opcode and timing signals to control the fetching and execution of instructions.

[0195] The processor may also include memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or that are used repeatedly. If the processor needs to use the instruction or data again, it can retrieve it directly from this memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0196] The aforementioned electronic device implements the carbon black production equipment health management method based on a large model, which collects key equipment parameters, process pipeline parameters, and environmental parameters of the carbon black production equipment through IoT sensors, and performs feature extraction on the collected data through edge computing nodes to generate structured feature vectors; it constructs and trains a large-scale model for carbon black production equipment health management based on a TFT model: the input is the structured feature vector; the time-delay correlation between process parameters and key equipment parameter features is captured through a hysteresis attention module; the output is the failure probability, process parameter deviation, and health prediction score; and the real-time structured feature vector is input into the carbon black production equipment health management system. The large-scale model is based on the output of calculated equipment health assessment indicators: failure probability, current health score, and health change rate. Extracted features are periodically synchronized to the twin model via edge computing nodes to construct a digital twin. A health status dashboard is generated based on the digital twin, providing a visual representation of the equipment health assessment indicators, the output of the large-scale carbon black production equipment health management model, and the extracted features. This technical solution achieves the beneficial effects of real-time monitoring, fault warning, and accurate health status assessment of carbon black production equipment through the integration of IoT, large-scale models, and digital twins. This reduces the risk of unplanned downtime, improves production efficiency and equipment reliability, and reduces reliance on manual labor.

[0197] The storage medium provided in this application stores a program product capable of implementing a health management method for carbon black production equipment based on a large model.

[0198] The large-scale model-based health management method for carbon black production equipment includes: collecting key equipment parameters, process pipeline parameters, and environmental parameters of the carbon black production equipment through IoT sensors; performing feature extraction on the collected data through edge computing nodes to generate structured feature vectors; constructing and training a large-scale health management model for carbon black production equipment based on a TFT model: inputting the structured feature vectors; capturing the time-delay correlation between process parameters and key equipment parameter features through a lag attention module; outputting failure probability, process parameter deviation, and health prediction score; inputting the real-time structured feature vectors into the large-scale health management model for carbon black production equipment, and calculating equipment health assessment indicators based on the output: failure probability, current health score, and health change rate; periodically synchronizing the extracted features to a twin model through edge computing nodes to construct a digital twin; generating a health status dashboard based on the digital twin to visualize the equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features.

[0199] In some possible implementations, the large-model-based carbon black production equipment health management method of this disclosure can be implemented as a program product comprising program code that, when run on a terminal device, causes the terminal device to perform the steps described in the "Exemplary Methods" section above according to various exemplary embodiments of this disclosure.

[0200] The storage medium disclosed herein may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0201] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A health management method for carbon black production equipment based on a large model, characterized in that, Includes the following steps: S1. Collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors; S2. Construct and train a large-scale health management model for carbon black production equipment based on the TFT model: The input is a structured feature vector; The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module. Outputs failure probability, process parameter deviation, and health prediction score; S3. Input the real-time structured feature vector into the large-scale health management model for carbon black production equipment, and calculate the equipment health assessment indicators based on the output: failure probability, current health score and health change rate; The specific steps of step S3 are as follows: S31. Input the structured feature vector generated from the real-time collected data into the trained carbon black production equipment health management model to obtain the failure probability, process parameter deviation and health prediction score of each type of fault. S32. Filter out the fault type with the highest failure probability and calculate the current health score: in, That is the highest probability of failure. It is the comprehensive deviation of industrial parameters, calculated from the deviation of process parameters output by the model; , These are preset weight parameters; S33. Calculate the current health score based on the health prediction score: Where k is the time window; S4. The extracted features are periodically synchronized to the twin model through edge computing nodes to construct a digital twin; S5. Based on the digital twin, a health status dashboard is generated, which visualizes the output of equipment health assessment indicators, the large-scale health management model of carbon black production equipment, and the extracted features.

2. The method for health management of carbon black production equipment based on a large model according to claim 1, characterized in that, The specific steps of step S1 are as follows: S11. Deploy an infrared thermal imager on the outer wall of the reactor to collect the temperature field matrix. ; Where H is the number of rows in the matrix and W is the number of columns in the matrix; S12. Vibration sensors are embedded in the bearing housings of the crusher and cooler to collect vibration acceleration signals. ; Where N is the number of samples; S13. Deploy flow meters, oxygen analyzers, and temperature sensors in the process pipeline to collect process parameter vectors. ; S14. Deploy dust sensors in the carbon black production workshop to collect environmental data: ； S15. Deploy edge computing nodes in the carbon black production workshop to perform feature extraction on the data collected by various IoT sensors: Time-domain characteristics of vibration acceleration signals from crushers and coolers: mean and kurtosis ; After applying a Hanning window to the vibration acceleration signals of the crusher and cooler, an FFT was performed to calculate the frequency domain characteristics: FFT peak value. ; Define the overheated region for the temperature field of the reactor. Calculate the proportion of overheated regions and construct statistical characteristics of the temperature field. ; in, It is the reference temperature. It is the standard deviation of the temperature field; S16. Generate structured feature vectors ; Where P is the process parameter, It is the mean of the vibration acceleration signal. Indicates the kurtosis of the vibration acceleration signal. Indicates the peak value of the FFT. This represents the statistical characteristics of the temperature field. This represents environmental data.

3. The method for health management of carbon black production equipment based on a large model according to claim 2, characterized in that, The specific steps of step S2 are as follows: S21. Construct a large-scale health management model for carbon black production equipment based on the TFT model; S22. Set structured feature vectors For input; S23. The delayed attention module is implemented by adding a time delay unit to the encoding layer of the TFT model and by gating the fusion of delayed features and real-time features; S24. Set the classification of faults and the probability of each type of fault as the first output, calculate the deviation of each parameter from the parameter standard value as the second output, and use the health prediction score for the future set period as the third output; S25. Train the large-scale health management model for carbon black production equipment and distribute it to edge computing nodes.

4. The method for health management of carbon black production equipment based on a large model according to claim 3, characterized in that, The specific steps of step S25 are as follows: S251. Collect structural feature vectors of carbon black production equipment under various operating conditions such as cold start, hot start, normal shutdown and emergency shutdown, as well as various load changes, to construct a dataset; S252. Label the dataset with normal samples and fault samples, and set the fault samples to include the process parameters within the time range [-t, t] before and after the change of process parameters; S253. Construct a multi-task loss function; in, , , These are preset weight parameters. It is the loss due to fault classification. It is the loss due to deviation of process parameters. It is the loss in predicting health scores; S254. Use the dataset to train a large model for the health management of carbon black production equipment, and use the Adam optimizer to optimize it during the training process until the loss function converges or the maximum number of iterations is met.

5. The method for health management of carbon black production equipment based on a large model according to claim 3, characterized in that, The specific steps of step S4 are as follows: S41. Edge computing nodes select features from the extracted features whose mutual information is higher than a preset mutual information threshold; in, Features With health score mutual information, It is a preset mutual information threshold; S42. After compressing the selected features according to a preset ratio, transmit them to the cloud using the MQTT protocol; S43. Building digital twins in the cloud: Statistical characteristics of temperature field Mapped to a surface thermal map of a 3D model of the reactor; Environmental data Convert to visual elements; S44. Set current health score Current health score The probability of failure is bound to the twin's metadata to complete the injection of health indicators.

6. The method for health management of carbon black production equipment based on a large model according to claim 5, characterized in that, The specific steps of step S5 are as follows: S51. Based on digital twins, construct equipment health status dashboards to visualize equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features. S52. The equipment health status dashboard integrates process parameters, vibration acceleration signals, and synchronously displays equipment status and environmental data. S53. Map the current health score to the color of the device twin model to display the health status of the device, and overlay the probability distribution of fault types onto the surface of the device twin model.

7. A system for health management of carbon black production equipment based on a large model according to any one of claims 1-6, characterized in that, include: The data acquisition and feature extraction module is used to collect key equipment parameters, process parameters of process pipelines, and environmental parameters of carbon black production equipment through IoT sensors, and to perform feature extraction on the collected data through edge computing nodes to generate structured feature vectors. The model building and training module is used to build and train a large-scale health management model for carbon black production equipment based on the TFT model. The input is a structured feature vector; The time-delay correlation between process parameters and key equipment parameters is captured using a delayed attention module. Outputs failure probability, process parameter deviation, and health prediction score; The health assessment module is used to input real-time structured feature vectors into the large-scale health management model of carbon black production equipment, and calculate equipment health assessment indicators based on the output: failure probability, current health score and health change rate. The digital twin construction module is used to periodically synchronize extracted features to the twin model via edge computing nodes to construct a digital twin. The visualization module is used to generate a health status dashboard based on the digital twin, and to visualize the equipment health assessment indicators, the output of the large-scale health management model for carbon black production equipment, and the extracted features.

8. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the large-model-based carbon black production equipment health management method as described in any one of claims 1 to 6.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the large-model-based carbon black production equipment health management method as described in any one of claims 1 to 6.

Images