Multi-fault diagnosis method and system for electronic information system based on artificial intelligence

Abstract

The invention discloses an electronic information system multi-fault diagnosis method and system based on artificial intelligence, and belongs to the technical field of electronic information system fault diagnosis, and the method comprises the steps: data collection and preprocessing, feature extraction and selection, multi-fault diagnosis model construction, real-time fault diagnosis and analysis, and fault recovery and feedback. The system comprises a data acquisition module, a preprocessing module, a feature extraction and selection module, a multi-fault diagnosis model module and a fault analysis and repair module, and all the modules cooperate to achieve full-process automation from data acquisition to fault repair. The method has a self-learning optimization capability, can significantly improve the fault diagnosis efficiency and reliability of a complex electronic information system, and is suitable for real-time fault diagnosis scenes of various types of electronic equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic information system fault diagnosis, and in particular to an electronic information system multi-fault diagnosis method and system based on artificial intelligence. Background Art

[0002] With the rapid development of electronic information technology, the scale and complexity of electronic information systems are increasing. These systems contain a large number of electronic components, modules, and complex signal transmission links. In actual operation, electronic information systems inevitably experience various faults, and the occurrence of multiple faults simultaneously is becoming increasingly common. Traditional fault diagnosis methods rely primarily on manual experience and simple threshold judgments, which have the following obvious drawbacks: Low diagnostic efficiency: For complex electronic information systems, manually checking fault points one by one is time-consuming and labor-intensive, and it is difficult to determine the fault location and cause in a short period of time.

[0003] Poor accuracy: Simple threshold judgment cannot handle complex fault scenarios and is prone to misjudgment and missed judgment. Especially when multiple faults occur simultaneously, traditional methods often have difficulty accurately distinguishing the characteristics and interrelationships of each fault.

[0004] Weak adaptability: Electronic information systems are constantly updated and upgraded, and new failure modes and types continue to emerge. Traditional methods are unable to quickly adapt to these changes and lack the ability to self-learn and self-optimize.

[0005] In recent years, artificial intelligence (AI) technology has been widely used in fault diagnosis. However, existing AI-based fault diagnosis solutions mostly focus on single faults. They lack effective solutions for simultaneous multiple faults and are unable to meet the fault diagnosis needs of modern complex electronic information systems. Therefore, there is an urgent need for an AI-based multi-fault diagnosis method and system for electronic information systems to improve the efficiency, accuracy, and adaptability of fault diagnosis. Summary of the Invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an electronic information system multi-fault diagnosis method and system based on artificial intelligence.

[0007] In order to achieve the above object, the present invention adopts the following technical solution: an artificial intelligence-based multi-fault diagnosis method for electronic information systems, the method comprising the following steps: Step 1: Data collection and preprocessing: sensors are set up at the nodes of the electronic information system to collect various data in real time during the operation of the system. The collected data are preprocessed for subsequent processing and analysis, and missing values ​​are handled by using methods to fill in the missing data. Step 2: Feature extraction and selection: Use signal processing technology to extract features from the preprocessed data, convert the time domain signal into a frequency domain signal, and use machine learning algorithms to reduce the dimensionality of the extracted features. Step 3: Build a multi-fault diagnosis model using a deep learning algorithm. Based on the data characteristics, design a suitable network structure. Use historical fault data to train the constructed multi-fault diagnosis model. Use the backpropagation algorithm to adjust the model parameters so that the model can accurately identify different types of fault characteristics and output fault diagnosis results. Step 4: Real-time fault diagnosis and analysis: input the pre-processed and feature-extracted data into the multi-fault diagnosis model, output the fault diagnosis results of the electronic information system in real time, analyze the multiple faults diagnosed, and perform fault analysis; Step 5: Fault repair and feedback. Based on the fault diagnosis and analysis results, a plan is generated and sent to maintenance personnel to guide them in performing fault repair operations. After the repair is completed, data is collected and fed back to the multi-fault diagnosis model for updating and optimization.

[0008] As a further description of the above technical solution: In step 1, the data includes but is not limited to physical parameters such as voltage, current, temperature, signal strength, frequency, as well as system logs and error code information. The collected data is preprocessed, including data cleaning, removal of noise data and outliers, data normalization, and uniform mapping of data in different ranges to specific intervals. The method for processing missing data values ​​adopts interpolation and averaging methods.

[0009] As a further description of the above technical solution: In the step 2, signal processing techniques include Fourier transform and wavelet transform, machine learning algorithms include principal component analysis (PCA) and linear discriminant analysis (LDA), and the extracted signal features include frequency features and energy features.

[0010] As a further description of the above technical solution: In step three, the learning algorithm includes convolutional neural network (CNN), recurrent neural network (RNN) and its variant long short-term memory network (LSTM), and the network structure includes an input layer, a hidden layer and an output layer.

[0011] As a further description of the above technical solution: In step 4, the fault diagnosis results include the fault type, fault location and probability of the fault occurring. The fault analysis includes determining the relationship between the faults, determining the priority order and impact range of the faults and tracing the root cause of the faults.

[0012] As a further description of the above technical solution: In step five, the fault repair plan includes maintenance steps, required parts and tool information, and the actual data during the repair process includes repair time, repair method, and repair effect information.

[0013] As a further description of the above technical solution: The data acquisition module is used to set sensors at key nodes of the electronic information system to collect various data during the operation of the system in real time and transmit the collected data to the data preprocessing module; The data preprocessing module is used to process the collected data, improve the quality and availability of the data, and transmit the preprocessed data to the feature extraction and selection module; The feature extraction and selection module is used to extract and select features from preprocessed data using signal processing technology and machine learning algorithms, extract the most representative feature subset, and transmit the feature subset to the multi-fault diagnosis model module; The multi-fault diagnosis model module is used to build a multi-fault diagnosis model based on a deep learning algorithm, train and optimize the model using historical fault data, receive feature subsets transmitted by the feature extraction and selection module, output fault diagnosis results in real time, and transmit the diagnosis results to the fault analysis module; The fault analysis module is used to analyze the diagnosed faults, determine the relationship between faults, determine the priority and impact range of faults, trace the root cause of the faults, and transmit the analysis results to the fault repair module; The fault repair module generates a fault repair plan based on the results of the fault analysis module and sends the repair plan to the maintenance personnel. After the fault is repaired, the module collects data from the repair process and feeds the data back to the multi-fault diagnosis model module.

[0014] As a further description of the above technical solution: In the data preprocessing module, data processing includes cleaning, normalization and missing value processing preprocessing operations.

[0015] As a further description of the above technical solution: In the fault analysis module, fault analysis includes determining the relationship between faults, determining the priority order and impact range of faults, and tracing the root cause of the faults.

[0016] The present invention has the following beneficial effects: In the present invention, by adopting artificial intelligence technology, especially deep learning algorithms, it is possible to quickly process a large amount of fault data, accurately identify multiple faults in electronic information systems, and improve the efficiency and accuracy of fault diagnosis. Compared with traditional methods, the diagnosis time is greatly shortened and the diagnostic accuracy is significantly improved.

[0017] In the present invention, the multi-fault diagnosis model can self-update and optimize by continuously learning new fault data, adapt to the ever-changing fault modes and types of the electronic information system, and has strong adaptability and scalability.

[0018] The present invention can not only diagnose the fault type and location, but also analyze the relationship between multiple faults, trace the root cause of the fault, provide comprehensive and detailed guidance for fault repair, and reduce the difficulty and workload of maintenance personnel.

[0019] In the present invention, through the feedback mechanism after fault repair, actual repair data is used to update and optimize the model, so that the performance of the multi-fault diagnosis model is continuously improved, and the fault diagnosis capability and reliability of the system are further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is the overall structural block diagram of the electronic information system multi-fault diagnosis method based on artificial intelligence proposed by the present invention and the system. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making creative efforts are within the scope of protection of the present invention.

[0022] Reference Figure 1 The present invention provides an embodiment of an electronic information system multi-fault diagnosis method based on artificial intelligence, the method comprising the following steps: Step 1: Data collection and preprocessing: sensors are set up at the nodes of the electronic information system to collect various data in real time during the operation of the system. The collected data are preprocessed for subsequent processing and analysis, and missing values ​​are handled by using methods to fill in the missing data. Step 2: Feature extraction and selection: Use signal processing technology to extract features from the preprocessed data, convert the time domain signal into a frequency domain signal, and use machine learning algorithms to reduce the dimensionality of the extracted features. Step 3: Build a multi-fault diagnosis model using a deep learning algorithm. Based on the data characteristics, design a suitable network structure. Use historical fault data to train the constructed multi-fault diagnosis model. Use the backpropagation algorithm to adjust the model parameters so that the model can accurately identify different types of fault characteristics and output fault diagnosis results. Step 4: Real-time fault diagnosis and analysis: input the pre-processed and feature-extracted data into the multi-fault diagnosis model, output the fault diagnosis results of the electronic information system in real time, analyze the multiple faults diagnosed, and perform fault analysis; Step 5: Fault repair and feedback. Based on the fault diagnosis and analysis results, a plan is generated and sent to maintenance personnel to guide them in performing fault repair operations. After the repair is completed, data is collected and fed back to the multi-fault diagnosis model for updating and optimization.

[0023] In step one, various data collected during the operation of the system include but are not limited to voltage, current, temperature, signal strength, frequency physical parameters, as well as system logs and error code information. The collected data are preprocessed, including data cleaning, removal of noise data and outliers, data normalization, and uniform mapping of data in different ranges to specific intervals. The interpolation method and the mean method are used to handle missing data values. In step two, signal processing techniques include Fourier transform and wavelet transform, and machine learning algorithms include principal component analysis (PCA) and linear discriminant analysis (LDA). The features of the extracted signals include frequency features and energy features. In step three, learning algorithms include convolutional neural networks (CNN), Recurrent Neural Network (RNN) and its variant Long Short-Term Memory Network (LSTM), the network structure includes input layer, hidden layer and output layer. In step 4, the fault diagnosis results include fault type, fault location and probability of fault occurrence. Fault analysis includes judging the relationship between faults, determining the priority order and impact scope of faults and tracing the root cause of the fault. In step 5, the fault repair plan includes maintenance steps, required parts and tool information. The actual data in the repair process includes repair time, repair method, and repair effect information. The data acquisition module is used to set sensors at key nodes of the electronic information system to collect various types of data during the operation of the system in real time and transmit the collected data to the system. Input to the data preprocessing module; the data preprocessing module is used to process the collected data, improve the quality and availability of the data, and transmit the preprocessed data to the feature extraction and selection module; the feature extraction and selection module is used to use signal processing technology and machine learning algorithms to extract and select features from the preprocessed data, extract the most representative feature subset, and transmit the feature subset to the multi-fault diagnosis model module; the multi-fault diagnosis model module is used to build a multi-fault diagnosis model based on deep learning algorithm, and use historical fault data to train and optimize the model, receive the feature subset transmitted by the feature extraction and selection module, output the fault diagnosis results in real time, and transmit the diagnosis results to the fault analysis module. Module; Fault analysis module, used to analyze the diagnosed faults, judge the relationship between faults, determine the priority order and impact range of faults, trace the root cause of the fault, and transmit the analysis results to the fault repair module; Fault repair module, based on the results of the fault analysis module, generates a fault repair plan and sends the repair plan to the maintenance personnel. After the fault is repaired, the data of the repair process is collected and fed back to the multi-fault diagnosis model module. In the data preprocessing module, data processing includes cleaning, normalization and missing value processing preprocessing operations. In the fault analysis module, fault analysis includes judging the relationship between faults, determining the priority order and impact range of faults, and tracing the root cause of the fault.

[0024] Working Principle: In a large electronic information system, voltage sensors, current sensors, temperature sensors, and signal strength sensors are installed at key nodes such as the power module, signal processing module, and data transmission module. These sensors collect real-time system operating data at a frequency of 10 times per second, including the input and output voltages and currents of the power module, the operating temperature of the signal processing module, and the signal strength and frequency of the data transmission module. The collected data first enters the data preprocessing module. During data cleaning, a reasonable threshold range is set to remove noise data that significantly exceeds the normal range. Outliers are identified and processed using an outlier detection algorithm based on statistical methods. Then, all data are normalized to the interval [0, 1] using the minimum-maximum normalization method. Missing values ​​are filled in using linear interpolation based on the time series characteristics of the data to ensure data integrity and continuity. The preprocessed data is input into the feature extraction and selection module. First, the power module current data is processed using Fourier transform to extract its frequency characteristics and analyze the distribution of different frequency components in the current. For the signal processing module temperature data, wavelet transform is used to obtain detailed characteristics and trend characteristics of temperature changes. Next, the principal component analysis (PCA) algorithm was used to reduce the dimensionality of all extracted features. The covariance matrix between features was calculated, and the number of principal components was determined. The principal components that explained at least 95% of the data variance were retained. The original high-dimensional feature space was mapped to a low-dimensional feature space, resulting in the most representative feature subset. A convolutional neural network (CNN) was used to construct a multi-fault diagnosis model. The model's input layer was configured based on the dimensionality of the feature subset, and the hidden layers consisted of multiple convolutional, pooling, and fully connected layers. The convolutional layers extracted local features from the fault data, while the pooling layers reduced the data dimensionality and computational complexity. The fully connected layers integrated the extracted features and output the fault diagnosis results. The model was trained using 10,000 sets of historical fault data accumulated over the past year from the electronic information system, encompassing a variety of single fault types and multiple fault combinations. During training, stochastic gradient descent was used as the optimizer, with a cross-entropy loss function as the loss function. The model parameters were continuously adjusted using a backpropagation algorithm. Validation is performed every 100 training batches, and the model's hyperparameters, such as convolution kernel size and learning rate, are adjusted according to the accuracy of the validation set. After 50 training cycles, the model's accuracy on the test set reached 92%.

[0025] When an electronic information system experiences an anomaly during operation, real-time data is preprocessed and feature extracted before being input into a trained multi-fault diagnosis model. The model quickly outputs fault diagnosis results. For example, it diagnoses a power module voltage anomaly with a probability of 0.9, and a signal processing module overheating with a probability of 0.85. After receiving the diagnostic results, the fault analysis module analyzes these two faults using a fault tree analysis method. It finds that the power module voltage anomaly can cause unstable operation of the signal processing module, which in turn causes overheating. It identifies the power module voltage anomaly as the primary fault and the signal processing module overheating as a secondary fault. It also analyzes the impact of the fault, potentially affecting the entire data processing subsystem. Based on the fault analysis results, the fault repair module generates a detailed repair plan. For power module voltage anomaly, the plan recommends inspecting components such as the power input circuit and voltage regulator, and replacing any potentially damaged components. For signal processing module overheating, it recommends cleaning the cooling fan and heat sink, checking for dry thermal grease, and replacing it. The repair plan is then sent to maintenance personnel, who then perform the repair according to the plan. After a fault is repaired, maintenance personnel collect feedback on the repair time, specific repair methods, and the system's operating status after the repair. This data is then fed back to the multi-fault diagnosis model. The model fine-tunes its parameters based on the new data, further optimizing its performance and improving the accuracy and reliability of future fault diagnosis.

[0026] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments, or replace some of the technical features therein with equivalents. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An artificial intelligence-based multi-fault diagnosis method for electronic information systems, characterized in that: The method comprises the following steps: Step 1: Data collection and preprocessing: sensors are set up at the nodes of the electronic information system to collect various data in real time during the operation of the system. The collected data are preprocessed for subsequent processing and analysis, and missing values ​​are handled by using methods to fill in the missing data. Step 2: Feature extraction and selection: Use signal processing technology to extract features from the preprocessed data, convert the time domain signal into a frequency domain signal, and use machine learning algorithms to reduce the dimensionality of the extracted features. Step 3: Build a multi-fault diagnosis model using a deep learning algorithm. Based on the data characteristics, design a suitable network structure. Use historical fault data to train the constructed multi-fault diagnosis model. Use the backpropagation algorithm to adjust the model parameters so that the model can accurately identify different types of fault characteristics and output fault diagnosis results. Step 4: Real-time fault diagnosis and analysis: input the pre-processed and feature-extracted data into the multi-fault diagnosis model, output the fault diagnosis results of the electronic information system in real time, analyze the multiple faults diagnosed, and perform fault analysis; Step 5: Fault repair and feedback. Based on the fault diagnosis and analysis results, a plan is generated and sent to maintenance personnel to guide them in performing fault repair operations. After the repair is completed, data is collected and fed back to the multi-fault diagnosis model for updating and optimization.

2. The multi-fault diagnosis method for electronic information systems based on artificial intelligence according to claim 1 is characterized in that: In step 1, the data includes but is not limited to physical parameters such as voltage, current, temperature, signal strength, frequency, as well as system logs and error code information. The collected data is preprocessed, including data cleaning, removal of noise data and outliers, data normalization, and uniform mapping of data in different ranges to specific intervals. The method for processing missing data values ​​adopts interpolation and averaging methods.

3. The multi-fault diagnosis method for electronic information systems based on artificial intelligence according to claim 1 is characterized in that: In the step 2, signal processing techniques include Fourier transform and wavelet transform, machine learning algorithms include principal component analysis and linear discriminant analysis, and the features of the extracted signal include frequency features and energy features.

4. The multi-fault diagnosis method for electronic information systems based on artificial intelligence according to claim 1 is characterized in that: In step three, the learning algorithm includes convolutional neural networks, recurrent neural networks and their variants, long short-term memory networks, and the network structure includes an input layer, a hidden layer and an output layer.

5. The multi-fault diagnosis method for electronic information systems based on artificial intelligence according to claim 1 is characterized in that: In step 4, the fault diagnosis results include the fault type, fault location and probability of the fault occurring. The fault analysis includes determining the relationship between the faults, determining the priority order and impact range of the faults and tracing the root cause of the faults.

6. The multi-fault diagnosis method for electronic information systems based on artificial intelligence according to claim 1 is characterized in that: In step five, the fault repair plan includes maintenance steps, required parts and tool information, and the actual data during the repair process includes repair time, repair method, and repair effect information.

7. The multi-fault diagnosis system of electronic information system based on artificial intelligence is characterized by: include, The data acquisition module is used to set sensors at key nodes of the electronic information system to collect various data during the operation of the system in real time and transmit the collected data to the data preprocessing module; The data preprocessing module is used to process the collected data, improve the quality and availability of the data, and transmit the preprocessed data to the feature extraction and selection module; The feature extraction and selection module is used to extract and select features from preprocessed data using signal processing technology and machine learning algorithms, extract the most representative feature subset, and transmit the feature subset to the multi-fault diagnosis model module; The multi-fault diagnosis model module is used to build a multi-fault diagnosis model based on a deep learning algorithm, train and optimize the model using historical fault data, receive feature subsets transmitted by the feature extraction and selection module, output fault diagnosis results in real time, and transmit the diagnosis results to the fault analysis module; The fault analysis module is used to analyze the diagnosed faults, determine the relationship between faults, determine the priority and impact range of faults, trace the root cause of the faults, and transmit the analysis results to the fault repair module; The fault repair module generates a fault repair plan based on the results of the fault analysis module and sends the repair plan to the maintenance personnel. After the fault is repaired, the module collects data from the repair process and feeds the data back to the multi-fault diagnosis model module.

8. The artificial intelligence-based multi-fault diagnosis system for electronic information systems according to claim 7, characterized in that: In the data preprocessing module, data processing includes cleaning, normalization and missing value processing preprocessing operations.

9. The artificial intelligence-based multi-fault diagnosis system for electronic information systems according to claim 7, characterized in that: In the fault analysis module, fault analysis includes determining the relationship between faults, determining the priority order and impact range of faults, and tracing the root cause of the faults.

Images