Display control drive circuit fault diagnosis system based on deep learning

Abstract

The invention provides a display control drive circuit fault diagnosis system based on deep learning, and relates to the field of circuit fault diagnosis, and the system comprises a data collection module which is used for obtaining and storing the operation data of a display control drive circuit; the data processing module is used for preprocessing the operation data of the display control driving circuit; the feature extraction module is used for performing feature extraction and mutual complementation on the preprocessed operation data by using a capsule network and a graph neural network to obtain a feature set; the model training module is used for constructing a deep forest model according to the feature set and performing iterative training to obtain an optimized deep forest model; the fault diagnosis module is used for performing fault diagnosis on the display control driving circuit to obtain a diagnosis result; and the early warning module is used for obtaining an early warning signal according to the diagnosis result and a threshold value. According to the invention, the accuracy and efficiency of fault diagnosis can be improved, and sample data dependence is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of material handling and sorting, and in particular to a display control drive circuit fault diagnosis system based on deep learning. Background Art

[0002] In the field of fault diagnosis of display control drive circuits, existing diagnostic methods mainly rely on expert experience and rule systems, which are often unable to cope with complex and changeable fault modes because it is difficult to automatically learn and extract fault features, and it is also impossible to make full use of a large amount of operating data. It is often difficult and time-consuming to obtain and annotate fault data of display control drive circuits. This application uses a deep learning model to automatically learn and extract fault features to achieve accurate diagnosis of display control drive circuit faults, improve the accuracy and efficiency of fault diagnosis, reduce dependence on expert experience, and adopt technical means such as transfer learning and data enhancement to reduce the model's dependence on large amounts of labeled data. Summary of the invention

[0003] The present invention provides a display control drive circuit fault diagnosis system based on deep learning, which can improve the accuracy and efficiency of fault diagnosis and reduce the dependence on sample data.

[0004] The present invention provides a display control drive circuit fault diagnosis system based on deep learning, comprising: A data acquisition module, used to acquire and store the operating data of the display control drive circuit, wherein the operating data includes voltage, current, temperature and circuit working status information; A data processing module, used for preprocessing the operation data of the display control drive circuit to obtain preprocessed operation data; A capsule network feature extraction module is used to extract a first feature of the data according to the preprocessed running data, and convert the first feature into a capsule form to obtain a feature of the capsule network; A graph neural network feature extraction module, which is used to convert the preprocessed running data into a graph structure to extract the features of the graph structure; The feature fusion module is used to fuse the features of the capsule network and the features of the graph structure to obtain a feature set; The model training module is used to construct a deep forest model based on the feature set and perform iterative training to obtain an optimized deep forest model; A fault diagnosis module, used for performing fault diagnosis on a display control driving circuit using an optimized deep forest model to obtain a diagnosis result; The early warning module is used to obtain early warning signals according to the diagnosis results and thresholds.

[0005] Furthermore, the data acquisition module includes: A data sensing module is used to sense and collect the operation data of the display control drive circuit; A data conversion module, used for converting operation data into electrical signals; The data storage module is used to store data converted into electrical signals.

[0006] Furthermore, the data processing module includes: A denoising module is used to remove high-frequency noise and random fluctuations of the operating data using a moving average filter to obtain denoised operating data; A normalization module, used for normalizing the denoised operation data to obtain normalized operation data; The filtering module is used to remove abnormal data from the normalized operating data to obtain preprocessed operating data.

[0007] Further, according to the preprocessed running data, the first feature of the data is extracted, and the first feature is converted into a capsule form to obtain the features of the capsule network, including: According to the preprocessed running data, the convolution layer is used to extract the first feature of the data, and the main capsule layer is used to convert the first feature into capsule form. , Adjust the connection weights between capsules to obtain the characteristics of the capsule network, where and For two capsules, It's a capsule The input vector is is the connection weight, which is iteratively updated through the dynamic routing algorithm. It's a capsule The output vector of .

[0008] Furthermore, the preprocessed running data is converted into a graph structure to extract the features of the graph structure, including: Use the graph convolution layer to extract the features of the graph structure and use Perform nonlinear transformation to obtain the characteristics of graph neural network, where is the characteristic of graph neural network, It's a capsule The set of neighbor nodes of is the connection weight, is a learnable weight matrix, is a non-linear activation function.

[0009] Furthermore, the model training module includes: A model building module, used to determine the architecture of the deep forest model according to the feature set, wherein the architecture of the model includes the number of layers of the cascade forest and the number and type of random forests in each layer; The iterative training module is used to input the feature set into the deep forest model for iterative training. In each round of training, the deep forest model learns the feature set and updates the parameters of the deep forest model to obtain an optimized deep forest model.

[0010] Furthermore, the fault diagnosis module includes: A diagnosis module, used to diagnose the operation data of the display control drive circuit using the optimized deep forest model to obtain a prediction result, and convert the prediction result into fault diagnosis information to obtain a diagnosis result; The monitoring module is used to diagnose the operating data of the display control drive circuit in real time using the optimized deep forest model.

[0011] Furthermore, the early warning module includes: A comparison module, used for comparing the acquired diagnosis result with a threshold value to obtain a comparison result; The judgment module is used to determine whether to trigger an early warning based on the comparison result. If the diagnosis result exceeds or reaches the set threshold, the early warning is triggered; if the diagnosis result does not reach the set threshold, the early warning is not triggered.

[0012] The display control drive circuit fault diagnosis system based on deep learning provided by the present invention combines capsule networks and graph neural networks for feature extraction, fully utilizes the advantages of the two networks, and improves the accuracy of fault diagnosis; through the construction and iterative training of the deep forest model, the system can learn the deep-level characteristics of the display control drive circuit operation data, and enhance the generalization ability of the model; the system realizes an automated and intelligent fault diagnosis process, from data collection, preprocessing, feature extraction to model diagnosis, the whole process is efficient and fast, shortens the fault diagnosis time, improves maintenance efficiency, and reduces production downtime; the early warning module can determine in real time whether it is necessary to trigger an early warning based on the diagnosis results and the set threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 is a schematic diagram of a display control drive circuit fault diagnosis system based on deep learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0016] like Figure 1 As shown, the display control drive circuit fault diagnosis system based on deep learning provided by the embodiment of the present invention mainly includes: A data acquisition module, used to acquire and store the operating data of the display control drive circuit, wherein the operating data includes voltage, current, temperature and circuit working status information; A data processing module, used for preprocessing the operation data of the display control drive circuit to obtain preprocessed operation data; A feature extraction module is used to extract features from the preprocessed running data using a capsule network and a graph neural network and complement each other to obtain a feature set; The model training module is used to construct a deep forest model based on the feature set and perform iterative training to obtain an optimized deep forest model; A fault diagnosis module, used for performing fault diagnosis on a display control driving circuit using an optimized deep forest model to obtain a diagnosis result; The early warning module is used to obtain early warning signals according to the diagnosis results and thresholds.

[0017] In the embodiment of the present invention, the data acquisition module is responsible for acquiring the operating data of the display control drive circuit and storing it; the data includes voltage, current, temperature and circuit working status information, etc., which are all important bases for subsequent fault diagnosis; by real-time and accurate acquisition of circuit operation data, comprehensive and reliable data support is provided for system fault diagnosis; the data processing module preprocesses the collected operation data, such as cleaning, labeling and conversion, to ensure that the data meets the requirements of the deep learning algorithm; through preprocessing, the quality of the data can be improved, the influence of noise and outliers can be reduced, and a good foundation can be laid for subsequent feature extraction and model training; the feature extraction module uses capsule networks and graph neural networks to extract features from the preprocessed operation data, and complement each other to obtain a more comprehensive feature set; through the combination of capsule networks and graph neural networks, it can automatically extract The deep features in the circuit operation data play a key role in the subsequent fault diagnosis; the model training module constructs a deep forest model according to the feature set and performs iterative training to obtain an optimized deep forest model; through iterative training, the deep forest model can gradually learn the characteristics and laws of circuit faults, and improve the accuracy and efficiency of its fault diagnosis; the fault diagnosis module uses the optimized deep forest model to perform fault diagnosis on the display control drive circuit; through the powerful analysis ability of the deep forest model, accurate diagnosis of circuit faults can be achieved, potential problems can be discovered in time, and the expansion of faults can be avoided; the early warning module determines whether it is necessary to issue an early warning signal based on the diagnosis results and the preset threshold; through the early warning module, an early warning signal can be issued in time before or at the early stage of a circuit fault, reminding relevant personnel to take measures to deal with it, thereby avoiding the impact of the fault on production.

[0018] like Figure 1 As shown, the data acquisition module includes: A data sensing module is used to sense and collect the operation data of the display control drive circuit; A data conversion module, used for converting operation data into electrical signals; The data storage module is used to store data converted into electrical signals.

[0019] In the embodiment of the present invention, the data perception module is responsible for real-time perception and collection of operating data of the display control drive circuit; the data may include key parameters such as voltage, current, temperature, operating frequency, etc., which are important indicators reflecting the circuit operation status; through high-precision sensors, the data perception module can capture small changes in circuit operation, and provide an accurate and reliable data source for subsequent data processing and analysis; ensure the real-time and accuracy of the data, and provide strong data support for fault diagnosis; improve the response speed of the system, so that faults can be discovered and handled in time; the data conversion module is responsible for converting the perceived operating data into electrical signals for subsequent processing and transmission; the process may involve operations such as signal amplification, filtering, and format conversion; through data conversion, The original physical quantities (such as voltage and current) are converted into digital signals to facilitate computer processing and analysis; the readability and processability of the data are improved, so that the data can be more conveniently used by the system; the stability and reliability of the data are ensured, and errors and interference in the data transmission process are reduced; the data storage module is responsible for storing the converted electrical signal data for subsequent data analysis and fault diagnosis; the module needs to have a large enough storage capacity and efficient data access capabilities; through data storage, historical data can be traced back and analyzed, providing a strong basis for fault prediction and maintenance; the integrity and traceability of the data are guaranteed, so that fault analysis can be more comprehensive and in-depth; the system's data management capabilities are improved, so that data can be more conveniently queried and used.

[0020] like Figure 1 As shown, the data processing module includes: A denoising module is used to remove high-frequency noise and random fluctuations of the operating data using a moving average filter to obtain denoised operating data; A normalization module, used for normalizing the denoised operation data to obtain normalized operation data; The filtering module is used to remove abnormal data from the normalized operating data to obtain preprocessed operating data.

[0021] In an embodiment of the present invention, a noise removal module is used to remove high-frequency noise and random fluctuations in operating data. Noise and fluctuations are caused by various interference factors in the circuit working environment, such as electromagnetic interference, temperature changes, etc., which will conceal the actual situation of circuit operation and affect the accuracy of fault diagnosis. Moving average filtering is an effective noise removal method. It smoothes data by calculating the average value of data points, thereby eliminating high-frequency noise and random fluctuations. After being processed by the noise removal module, the operating data becomes smoother and more stable, and the noise level is significantly reduced, so that subsequent data processing and analysis can more accurately reflect the operating status of the circuit and improve the accuracy and reliability of fault diagnosis. The purpose of the normalization module is to normalize the operating data after noise removal, that is, to convert the data into a unified range (such as between 0 and 1). Normalization can eliminate the dimensional differences between different data, making the data comparable, and also helps to improve the training efficiency and accuracy of the deep learning model. Through normalization, the operating data is unified into the same range, and the distribution of the data becomes more uniform. It is not only conducive to the deep learning model's understanding and learning of data, but also improves the generalization ability of the model, so that the model can better adapt to different circuit operating states; the selection module is responsible for removing abnormal data from the normalized operating data; abnormal data may be caused by sensor failure, data acquisition error or abnormal state of the circuit itself; the existence of abnormal data will interfere with the training process of the deep learning model and reduce the accuracy of the model; therefore, the selection module identifies and removes these abnormal data by setting certain thresholds or rules; after being processed by the selection module, the abnormal data in the operating data is effectively removed, and the purity of the data is improved; the deep learning model can more accurately learn the normal characteristics and fault characteristics of the circuit operation, thereby improving the accuracy and efficiency of fault diagnosis; at the same time, it also avoids the interference of abnormal data on the model training process, ensuring the stability and reliability of the model; in summary, the data processing module effectively improves the accuracy and reliability of the operating data through the collaborative work of the three sub-modules of denoising, normalization and selection, and provides strong data support for subsequent deep learning model training and fault diagnosis.

[0022] like Figure 1 As shown, the feature extraction module includes: The capsule network feature extraction module is used to extract the first feature of the data using the convolution layer according to the preprocessed running data, and convert the first feature into capsule form using the main capsule layer. , Adjust the connection weights between capsules to obtain the characteristics of the capsule network, where and For two capsules, It's a capsule The input vector is is the connection weight, which is iteratively updated through the dynamic routing algorithm. It's a capsule The output vector of The graph neural network feature extraction module is used to convert the preprocessed running data into a graph structure, and use the graph convolution layer to extract the features of the graph structure. Perform nonlinear transformation to obtain the characteristics of graph neural network, where is the characteristic of graph neural network, It's a capsule The set of neighbor nodes of is the connection weight, is a learnable weight matrix, is a nonlinear activation function; The feature fusion module is used to fuse the features of the capsule network and the features of the graph neural network to obtain a feature set.

[0023] In an embodiment of the present invention, the capsule network feature extraction module uses a convolution layer to perform preliminary feature extraction on the preprocessed running data to obtain the first feature of the data. The convolution layer can effectively capture local features in the data by means of sliding windows and convolution kernels; the main capsule layer is used to convert the first feature into capsule form; capsules are the basic units in the capsule network, which can represent various attributes of the data, such as direction, size, etc., and are more expressive than traditional neurons; through the dynamic routing formula, the module adjusts the connection weights between capsules, so that the capsule network can adaptively learn the hierarchical structure and spatial relationship in the data, thereby obtaining the characteristics of the capsule network; the capsule network feature extraction module can capture the deep-level features in the data; the introduction of capsules enhances the model's ability to understand the data and improves the accuracy and robustness of feature extraction; dynamic The stateful routing mechanism enables the capsule network to adaptively adjust its structure to better adapt to different types of fault data; the graph neural network feature extraction module first converts the preprocessed operating data into a graph structure; a graph structure is a data form that can represent complex relationships between data, and is suitable for describing connection relationships, dependencies, etc. in circuits; a graph convolution layer is used to extract features from the graph structure, which updates the features of the current node by aggregating the features of neighboring nodes, thereby capturing the global and local relationships between data; a nonlinear transformation is performed through the graph convolution formula to further extract deep features in the graph structure and obtain the features of the graph neural network; the graph neural network feature extraction module The block can capture the complex relationship between data, which is crucial for understanding the operating status of the circuit and diagnosing faults; the introduction of the graph convolution layer enables the model to process non-Euclidean data, expanding the application scope of the model; through nonlinear transformation, the module can extract nonlinear features in the data, improving the accuracy and richness of feature extraction; the feature fusion module is responsible for fusing the features of the capsule network with the features of the graph neural network to obtain a feature set; the fusion process can be selected by splicing, weighted summation, etc. according to specific task requirements; the fused feature set contains features obtained by two different feature extraction methods, and has stronger expression and generalization capabilities; The feature fusion module enables the model to comprehensively utilize the advantages of two different feature extraction methods to improve the accuracy of fault diagnosis; the fused feature set is richer and more comprehensive, and can capture a variety of feature information in the data, providing strong support for subsequent fault diagnosis; through feature fusion, the model can better adapt to different types of fault data, improve generalization ability and robustness; in summary, the feature extraction module can effectively extract features useful for fault diagnosis from the preprocessed operating data through the collaborative work of the capsule network feature extraction module, the graph neural network feature extraction module and the feature fusion module, providing accurate and reliable feature information for subsequent fault diagnosis.

[0024] like Figure 1As shown, the model training module includes: A model building module, used to determine the architecture of the deep forest model according to the feature set, wherein the architecture of the model includes the number of layers of the cascade forest and the number and type of random forests in each layer; The iterative training module is used to input the feature set into the deep forest model for iterative training. In each round of training, the deep forest model learns the feature set and updates the parameters of the deep forest model to obtain an optimized deep forest model.

[0025] In an embodiment of the present invention, the model construction module is responsible for determining the architecture of the deep forest model according to the feature set; the module will comprehensively consider the scale of the feature set, the correlation between the features and the specific needs of fault diagnosis to determine the number of layers of the cascade forest, the number and type of random forests in each layer; through a carefully designed model architecture, it can be ensured that the model has sufficient expressive power to capture complex fault features and maintain high computational efficiency; the constructed deep forest model can be well adapted to fault diagnosis tasks and has excellent feature learning and generalization capabilities; a reasonable model architecture design helps to reduce the risks of overfitting and underfitting and improve the accuracy of fault diagnosis; the iterative training module is responsible for inputting the feature set into the deep forest model for iterative training; through continuous training and adjustment, the model can gradually learn the fault features and optimize its parameters; in each round of training, the deep forest model will learn the feature set, and The model parameters are updated by comparing the difference between the predicted results and the actual results; through multiple iterative training, the model can gradually converge to a better state and obtain the optimized deep forest model; the deep forest model after iterative training can more accurately identify fault features and improve the accuracy and efficiency of fault diagnosis; the iterative training process helps the model to explore the deep relationship between features and enhance the model's ability to diagnose complex faults; by continuously optimizing the model parameters, the model can maintain high accuracy while also having good robustness and stability; in summary, the model training module can construct a deep forest model suitable for fault diagnosis tasks through the collaborative work of the two sub-modules of model construction and iterative training, and continuously optimize the model parameters through iterative training to improve the accuracy and efficiency of fault diagnosis; the design and implementation of the module plays a vital role in improving the performance of the entire fault diagnosis system.

[0026] like Figure 1 As shown, the fault diagnosis module includes: A diagnosis module, used to diagnose the operation data of the display control drive circuit using the optimized deep forest model to obtain a prediction result, and convert the prediction result into fault diagnosis information to obtain a diagnosis result; The monitoring module is used to diagnose the operating data of the display control drive circuit in real time using the optimized deep forest model.

[0027] In the embodiment of the present invention, the fault diagnosis module is responsible for using the optimized deep forest model to perform fault diagnosis on the operation data of the display control drive circuit. By inputting the actual operation data, the model can output the corresponding prediction-result module; the prediction result is converted into easy-to-understand fault diagnosis information, such as fault type, fault location, fault severity, etc., so as to obtain a clear diagnosis result; by comparing with the actual fault records or expert experience, the accuracy and reliability of the diagnosis result can be verified; the diagnosis module can efficiently and accurately identify the fault in the display control drive circuit, and improve the efficiency and accuracy of fault diagnosis; the converted fault diagnosis information provides maintenance personnel with direct fault indications, which helps to quickly locate and solve problems and reduce maintenance time and costs; by comparing and verifying with the actual situation, the model can be continuously optimized to improve the diagnostic capability in different fault scenarios; the monitoring module is responsible for using the optimized deep forest model to diagnose the operation data of the display control drive circuit in real time; It is of great significance for timely discovery of potential faults and prevention of fault expansion; by continuously monitoring the operating status of the circuit, the module can issue an early warning signal at the early stage of the fault, providing maintenance personnel with the opportunity to intervene in advance; real-time monitoring also helps to collect more operating data, providing rich sample resources for further optimization and training of the model; the monitoring module can realize real-time fault monitoring of the display control drive circuit and improve the reliability and safety of the system; timely early warning signals help maintenance personnel to respond and handle faults quickly to avoid greater impact of faults on production or use; through continuous data collection and model optimization, the performance and accuracy of the fault diagnosis system can be continuously improved, providing strong support for future fault prediction and intelligent maintenance; in summary, the fault diagnosis module realizes efficient and accurate diagnosis and real-time monitoring of display control drive circuit faults through the collaborative work of the diagnosis module and the monitoring module, improves the efficiency and accuracy of fault diagnosis, and provides a strong guarantee for the reliability and safety of the system.

[0028] like Figure 1 As shown, the early warning module includes: A comparison module, used for comparing the acquired diagnosis result with a threshold value to obtain a comparison result; The judgment module is used to determine whether to trigger an early warning based on the comparison result. If the diagnosis result exceeds or reaches the set threshold, the early warning is triggered; if the diagnosis result does not reach the set threshold, the early warning is not triggered.

[0029] In an embodiment of the present invention, the comparison module is responsible for comparing the acquired diagnostic result with a preset threshold value; the threshold value is usually set based on experience, historical data or expert advice, and represents the reasonable range of certain key parameters when the circuit is operating normally; through comparison, the module can determine whether the diagnostic result exceeds the normal range, that is, whether it has reached a level that may cause a fault or require attention; the comparison module provides a basic judgment basis for the early warning system, making the issuance of early warning signals more accurate and targeted; by comparing with the threshold value, abnormal conditions in circuit operation can be discovered in time, winning valuable time for subsequent fault handling; the judgment module decides whether to trigger an early warning based on the output result of the comparison module, that is, the comparison result. If the diagnostic result exceeds or reaches the set threshold, it means that the circuit may have a fault or is about to fail, and the judgment module will trigger a warning signal; if the diagnostic result does not reach the set threshold, it means that the circuit is currently operating normally, and the judgment module will not trigger a warning to avoid unnecessary interference and false alarms; the judgment module ensures the accuracy and reliability of the warning signal, avoiding unnecessary maintenance costs and production interruptions caused by false alarms or missed alarms; by triggering the warning signal in time, the judgment module provides maintenance personnel with the opportunity to respond and handle faults quickly, which helps to reduce the impact of faults on production or use; at the same time, the judgment module can also flexibly adjust the warning strategy according to actual conditions, such as adjusting the threshold or adding other judgment conditions, to meet the needs of different circuits and operating environments; in summary, the warning module realizes effective warning of display control drive circuit faults through the collaborative work of the comparison module and the judgment module, which not only improves the practicality and reliability of the fault diagnosis system, but also provides a strong guarantee for the safe and stable operation of the circuit.

[0030] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A display control drive circuit fault diagnosis system based on deep learning, characterized in that: include: A data acquisition module, used to acquire and store the operating data of the display control drive circuit, wherein the operating data includes voltage, current, temperature and circuit working status information; A data processing module, used for preprocessing the operation data of the display control drive circuit to obtain preprocessed operation data; A capsule network feature extraction module is used to extract a first feature of the data according to the preprocessed running data, and convert the first feature into a capsule form to obtain a feature of the capsule network; A graph neural network feature extraction module, which is used to convert the preprocessed running data into a graph structure to extract the features of the graph structure; The feature fusion module is used to fuse the features of the capsule network and the features of the graph structure to obtain a feature set; The model training module is used to construct a deep forest model based on the feature set and perform iterative training to obtain an optimized deep forest model; A fault diagnosis module, used for performing fault diagnosis on a display control driving circuit using an optimized deep forest model to obtain a diagnosis result; The early warning module is used to obtain early warning signals according to the diagnosis results and thresholds.

2. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: The data acquisition module comprises: A data sensing module is used to sense and collect the operation data of the display control drive circuit; A data conversion module, used for converting operation data into electrical signals; The data storage module is used to store data converted into electrical signals.

3. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: The data processing module comprises: A denoising module is used to remove high-frequency noise and random fluctuations of the operating data using a moving average filter to obtain denoised operating data; A normalization module, used for normalizing the denoised operation data to obtain normalized operation data; The filtering module is used to remove abnormal data from the normalized operating data to obtain preprocessed operating data.

4. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: According to the preprocessed running data, the first feature of the data is extracted and converted into a capsule form to obtain the features of the capsule network, including: According to the preprocessed running data, the convolution layer is used to extract the first feature of the data, and the main capsule layer is used to convert the first feature into capsule form. , Adjust the connection weights between capsules to obtain the characteristics of the capsule network, where and For two capsules, It's a capsule The input vector is is the connection weight, which is iteratively updated through the dynamic routing algorithm. It's a capsule The output vector of .

5. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: Convert the preprocessed running data into a graph structure to extract the features of the graph structure, including: Use the graph convolution layer to extract the features of the graph structure and use Perform nonlinear transformation to obtain the characteristics of graph neural network, where is the characteristic of graph neural network, It's a capsule The set of neighbor nodes of is the connection weight, is a learnable weight matrix, is a non-linear activation function.

6. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: The model training module includes: A model building module, used to determine the architecture of the deep forest model according to the feature set, wherein the architecture of the model includes the number of layers of the cascade forest and the number and type of random forests in each layer; The iterative training module is used to input the feature set into the deep forest model for iterative training. In each round of training, the deep forest model learns the feature set and updates the parameters of the deep forest model to obtain an optimized deep forest model.

7. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: The fault diagnosis module comprises: A diagnosis module, used to diagnose the operation data of the display control drive circuit using the optimized deep forest model to obtain a prediction result, and convert the prediction result into fault diagnosis information to obtain a diagnosis result; The monitoring module is used to diagnose the operating data of the display control drive circuit in real time using the optimized deep forest model.

8. The display control drive circuit fault diagnosis system based on deep learning according to claim 1, characterized in that: The early warning module comprises: A comparison module, used for comparing the acquired diagnosis result with a threshold value to obtain a comparison result; The judgment module is used to determine whether to trigger an early warning based on the comparison result. If the diagnosis result exceeds or reaches the set threshold, the early warning is triggered; if the diagnosis result does not reach the set threshold, the early warning is not triggered.

Images