Intelligent network equipment hardware alarm processing method and system

Abstract

The invention particularly relates to an intelligent network equipment hardware alarm processing method and system. According to the intelligent network equipment hardware alarm processing method, hardware state data are collected regularly according to a self-defined period, and key feature information is extracted from the hardware state data after preprocessing; self-defining a set threshold value, analyzing the feature information by using a trained alarm identification model, and identifying whether a hardware fault exists or not; processing measures corresponding to the fault types are set in a self-defined mode, and the corresponding processing measures are triggered in a self-defined mode according to the recognized fault types; and the alarm information is notified to a user or a maintainer through various self-defined modes, so that measures can be taken in time. According to the intelligent network equipment hardware alarm processing method and system, through comprehensive utilization of modern information technology means, automatic identification and processing of network equipment hardware faults are realized, reliability and maintenance efficiency of equipment are improved, and cost and false alarm rate of manual monitoring are reduced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent device computing technology, and in particular to an intelligent network device hardware alarm processing method and system. Background Art

[0002] With the development of information technology, network equipment (such as routers, switches, servers, etc.) plays an increasingly important role in business and personal life. These devices need to operate stably for a long time and be able to be discovered and handled in time when problems occur.

[0003] Traditional network equipment fault detection methods mainly rely on manual monitoring or simple hardware alarm systems, which have problems such as slow response speed, high false alarm rate, and high maintenance cost. Therefore, developing a method that can automatically identify and process network equipment hardware alarms is of great significance to improving equipment reliability and maintenance efficiency.

[0004] In order to automatically detect hardware failures of equipment and take corresponding processing measures according to alarm information, the present invention proposes a method and system for processing hardware alarms of intelligent network equipment. Summary of the invention

[0005] In order to overcome the defects of the prior art, the present invention provides a simple and efficient intelligent network equipment hardware alarm processing method and system.

[0006] The present invention is achieved through the following technical solutions:

[0007] A method for processing hardware alarms of intelligent network equipment comprises the following steps:

[0008] Step S1: Data collection

[0009] Collect hardware status data generated by network equipment during operation according to custom cycles;

[0010] Step S2: Data processing

[0011] Preprocess the collected hardware status data, including filtering, noise reduction and smoothing, to remove erroneous values; Step S3, feature extraction

[0012] Customize key features based on device information and extract key feature information from preprocessed data;

[0013] Step S4: Alarm identification

[0014] Customize thresholds and use trained alarm recognition models to analyze feature information and identify whether there is a hardware failure.

[0015] The alarm recognition model uses a support vector machine algorithm, a random forest algorithm or a convolutional neural network to train and classify feature information.

[0016] Step S5: Decision making

[0017] Customize the corresponding handling measures for the fault type;

[0018] After the trained machine learning model is used to identify hardware faults, the corresponding processing measures are triggered according to the identified fault type, such as sending alarm information, starting emergency procedures, etc.

[0019] Step S6: User interaction

[0020] The alarm information is notified to users or maintenance personnel through various customized methods (such as SMS, email, APP push, etc.) so that timely measures can be taken.

[0021] In step S4, the alarm recognition model is used to identify whether there is a hardware fault, and the steps are as follows:

[0022] Step S4.1, customizing and selecting historical alarm data of network devices to build an alarm label library, and using the alarm label library to evaluate the alarm recognition model;

[0023] Step S4.2: Build an alarm rule library and establish a data synchronization mechanism with the alarm tag library;

[0024] Step S4.3: After the current hardware status data is obtained, first determine whether the current hardware status data exists in the alarm rule library:

[0025] If it does not exist, the alarm recognition model is input for recognition and judgment, and the recognition result is output;

[0026] If it exists, the corresponding fault category will be output directly.

[0027] In step S4.3, manual troubleshooting is performed and feedback is provided on the recognition results output by the alarm recognition model:

[0028] If the fault category identification result is correct, the corresponding hardware status data and fault category are written into the alarm rule library;

[0029] If the fault category identification result is wrong, the corresponding hardware status data will be input into the alarm recognition model again, and the alarm recognition model will be adjusted and optimized.

[0030] An intelligent network equipment hardware alarm processing system, comprising:

[0031] The data collection module is responsible for collecting the hardware status data generated by the network equipment during operation according to the custom cycle;

[0032] The data processing module is responsible for preprocessing the collected hardware status data, including filtering, noise reduction and smoothing, and removing erroneous values;

[0033] The feature extraction module is responsible for defining key features based on device information and extracting key feature information from preprocessed data;

[0034] The alarm recognition module is responsible for customizing the threshold and using the trained alarm recognition model to analyze the feature information and identify whether there is a hardware failure.

[0035] The alarm recognition model uses a support vector machine algorithm, a random forest algorithm or a convolutional neural network to train and classify feature information.

[0036] The decision-making processing module is responsible for customizing the processing measures corresponding to the fault type. After identifying the hardware fault using the trained machine learning model, it triggers the corresponding processing measures according to the identified fault type, such as sending an alarm message, starting the emergency procedure, etc.

[0037] The user interaction module is responsible for notifying users or maintenance personnel of alarm information through various customized methods (such as SMS, email, APP push, etc.) so that timely measures can be taken.

[0038] The alarm identification module uses the alarm identification model to identify whether there is a hardware fault, the steps are as follows:

[0039] Step S4.1, customizing and selecting historical alarm data of network devices to build an alarm label library, and using the alarm label library to evaluate the alarm recognition model;

[0040] Step S4.2: Build an alarm rule library and establish a data synchronization mechanism with the alarm tag library;

[0041] Step S4.3: After the current hardware status data is obtained, first determine whether the current hardware status data exists in the alarm rule library:

[0042] If it does not exist, the alarm recognition model is input for recognition and judgment, and the recognition result is output;

[0043] If it exists, the corresponding fault category will be output directly.

[0044] In step S4.3, manual troubleshooting is performed and feedback is provided on the recognition results output by the alarm recognition model:

[0045] If the fault category identification result is correct, the corresponding hardware status data and fault category are written into the alarm rule library;

[0046] If the fault category identification result is wrong, the corresponding hardware status data will be input into the alarm recognition model again, and the alarm recognition model will be adjusted and optimized.

[0047] An intelligent network equipment hardware alarm processing device is characterized by comprising a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above method steps when executing the computer program.

[0048] A readable storage medium, characterized in that: a computer program is stored on the readable storage medium, and the computer program implements the above method steps when executed by a processor.

[0049] The beneficial effects of the present invention are as follows: the intelligent network equipment hardware alarm processing method and system, by comprehensively utilizing modern information technology means, realizes automatic identification and processing of network equipment hardware failures, which not only improves the reliability and maintenance efficiency of the equipment, but also reduces the cost and false alarm rate of manual monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are 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.

[0051] Attached Figure 1 It is a schematic diagram of the hardware alarm processing method of the intelligent network device of the present invention.

[0052] Attached Figure 2 Schematic diagram of the warning feature recognition training method of the present invention. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a 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 should fall within the scope of protection of the present invention.

[0054] The intelligent network device hardware alarm processing method comprises the following steps:

[0055] Step S1: Data collection

[0056] Collect hardware status data generated by network devices (such as temperature sensors, fan sensors, power modules, etc.) during operation according to custom cycles;

[0057] Step S2: Data processing

[0058] Preprocess the collected hardware status data, including filtering, noise reduction and smoothing, to remove erroneous values; Step S3, feature extraction

[0059] Determine key features based on device information and extract key feature information (such as temperature sensors, fan sensors, power modules, etc.) from preprocessed data;

[0060] Step S4: Alarm identification

[0061] Customize thresholds and use trained alarm recognition models to analyze feature information and identify whether there is a hardware failure.

[0062] The alarm recognition model uses a support vector machine algorithm, a random forest algorithm or a convolutional neural network to train and classify feature information.

[0063] Step S5: Decision making

[0064] Customize the corresponding handling measures for the fault type;

[0065] After the trained machine learning model is used to identify hardware faults, the corresponding processing measures are triggered according to the identified fault type, such as sending alarm information, starting emergency procedures, etc.

[0066] Step S6: User interaction

[0067] The alarm information is notified to users or maintenance personnel through various customized methods (such as SMS, email, APP push, etc.) so that timely measures can be taken.

[0068] In step S4, the alarm recognition model is used to identify whether there is a hardware fault, and the steps are as follows:

[0069] Step S4.1, customizing and selecting historical alarm data of network devices to build an alarm label library, and using the alarm label library to evaluate the alarm recognition model;

[0070] Step S4.2: Build an alarm rule library and establish a data synchronization mechanism with the alarm tag library;

[0071] Step S4.3: After the current hardware status data is obtained, first determine whether the current hardware status data exists in the alarm rule library:

[0072] If it does not exist, the alarm recognition model is input for recognition and judgment, and the recognition result is output;

[0073] If it exists, the corresponding fault category will be output directly.

[0074] In step S4.3, manual troubleshooting is performed and feedback is provided on the recognition results output by the alarm recognition model:

[0075] If the fault category identification result is correct, the corresponding hardware status data and fault category are written into the alarm rule library;

[0076] If the fault category identification result is wrong, the corresponding hardware status data will be input into the alarm recognition model again, and the alarm recognition model will be adjusted and optimized.

[0077] The intelligent network equipment hardware alarm processing system includes:

[0078] The data acquisition module is responsible for collecting the hardware status data generated by the operation of network devices (such as temperature sensors, fan sensors, power modules, etc.) according to a custom cycle;

[0079] The data processing module is responsible for preprocessing the collected hardware status data, including filtering, noise reduction and smoothing, and removing erroneous values;

[0080] The feature extraction module is responsible for defining key features based on device information and extracting key feature information (such as temperature sensors, fan sensors, power modules, etc.) from preprocessed data.

[0081] The alarm recognition module is responsible for customizing the threshold and using the trained alarm recognition model to analyze the feature information and identify whether there is a hardware failure.

[0082] The alarm recognition model uses a support vector machine algorithm, a random forest algorithm or a convolutional neural network to train and classify feature information.

[0083] The decision-making processing module is responsible for customizing the processing measures corresponding to the fault type. After identifying the hardware fault using the trained machine learning model, it triggers the corresponding processing measures according to the identified fault type, such as sending an alarm message, starting the emergency procedure, etc.

[0084] The user interaction module is responsible for notifying users or maintenance personnel of alarm information through various customized methods (such as SMS, email, APP push, etc.) so that timely measures can be taken.

[0085] The alarm identification module uses the alarm identification model to identify whether there is a hardware fault, the steps are as follows:

[0086] Step S4.1, customizing and selecting historical alarm data of network devices to build an alarm label library, and using the alarm label library to evaluate the alarm recognition model;

[0087] Step S4.2: Build an alarm rule library and establish a data synchronization mechanism with the alarm tag library;

[0088] Step S4.3: After the current hardware status data is obtained, first determine whether the current hardware status data exists in the alarm rule library:

[0089] If it does not exist, the alarm recognition model is input for recognition and judgment, and the recognition result is output;

[0090] If it exists, the corresponding fault category will be output directly.

[0091] In step S4.3, manual troubleshooting is performed and feedback is provided on the recognition results output by the alarm recognition model:

[0092] If the fault category identification result is correct, the corresponding hardware status data and fault category are written into the alarm rule library;

[0093] If the fault category identification result is wrong, the corresponding hardware status data will be input into the alarm recognition model again, and the alarm recognition model will be adjusted and optimized.

[0094] Example 1: Router overheat protection

[0095] Data collection: A temperature sensor is installed inside the router to monitor the temperature changes of the router in real time. The temperature sensor collects data once a minute and sends the data to the data collection module through the network.

[0096] Data preprocessing: After receiving the temperature data, the data processing module uses the sliding window method to smooth the data and remove outliers to ensure the continuity and stability of the data.

[0097] Feature extraction: Extract features such as maximum, minimum, and average values ​​from temperature data, and calculate the temperature fluctuation range and change trend.

[0098] Alarm identification: Use the random forest algorithm to classify temperature features and identify whether there is an overheating risk.

[0099] When training the model, use the normal temperature range in historical data as a reference to set reasonable alarm thresholds.

[0100] Decision-making process: If overheating is detected, the cooling fan speed is automatically adjusted and an alarm message is sent to maintenance personnel. Emergency procedures are initiated, such as reducing equipment load and reducing heat sources.

[0101] User notification: Alarm information is sent to maintenance personnel via SMS or email, including temperature data and recommended operating procedures.

[0102] Example 2: Identifying a switch fan failure

[0103] Data collection: Install a fan status sensor inside the switch to monitor the fan's working status in real time. The fan status sensor collects data every five minutes and sends the data to the data collection module through the network.

[0104] Data preprocessing: After receiving the fan status data, the data processing module uses Fourier transform to perform frequency domain analysis on the data, filter out high-frequency noise, and remove background noise.

[0105] Feature extraction: Extract features such as speed and vibration frequency from fan status data, and analyze the changing trend of fan speed and vibration mode.

[0106] Alarm identification: Use the support vector machine model to analyze the fan status characteristics and determine whether there is a fault.

[0107] When training the model, use the normal fan status in historical data as a reference to set reasonable alarm thresholds.

[0108] Decision-making process: Once a fan failure is identified, the emergency procedures are immediately initiated, such as sending an alarm message, recommending fan replacement, and notifying maintenance personnel to conduct on-site inspection and repair.

[0109] User notification: The switch management system notifies the system administrator of the alarm information, which includes fan status data and recommended operation steps.

[0110] Example 3: Server power module fault detection

[0111] Data collection: Install a power module status sensor inside the server to monitor the working status of the power module in real time. The power module status sensor collects data every ten minutes and sends the data to the data collection module through the network.

[0112] Data preprocessing: After receiving the power module status data, the data processing module uses the sliding window method to smooth the data and remove outliers to ensure the continuity and stability of the data.

[0113] Feature extraction: Extract voltage, current and other features from the power module status data to analyze the power supply status and current change trend of the power module.

[0114] Alarm identification: Use a convolutional neural network model to analyze the status characteristics of the power module and identify whether there is a fault.

[0115] When training the model, use the normal power module status in historical data as a reference to set reasonable alarm thresholds.

[0116] Decision-making process: If a power module failure is detected, the emergency procedure is immediately initiated, such as sending an alarm message, suggesting replacing the power module, and notifying maintenance personnel to conduct on-site inspection and repair.

[0117] User notification: The server management system notifies the system administrator of the alarm information, which includes the power module status data and recommended operation steps.

[0118] The intelligent network device hardware alarm processing device comprises a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above method steps when executing the computer program.

[0119] The readable storage medium stores a computer program, and when the computer program is executed by a processor, the above method steps are implemented.

[0120] The intelligent network equipment hardware alarm processing method and system realizes the automatic identification and processing of network equipment hardware failures by comprehensively utilizing modern information technology means; it not only improves the reliability and maintenance efficiency of the equipment, but also reduces the cost and false alarm rate of manual monitoring.

[0121] With the continuous advancement of technology, the application scope of the intelligent network equipment hardware alarm processing method and system will be further expanded to provide security protection for more intelligent network devices.

[0122] The embodiment described above is only one specific implementation of the present invention. Common changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for processing hardware alarms of intelligent network equipment, characterized in that: The following steps are involved: Step S1: Data collection Collect hardware status data generated by network equipment during operation according to custom cycles; Step S2: Data processing Preprocess the collected hardware status data, including filtering, noise reduction and smoothing, to remove erroneous values; Step S3: Feature extraction Customize key features based on device information and extract key feature information from preprocessed data; Step S4: Alarm identification Customize thresholds and use trained alarm recognition models to analyze feature information and identify whether there is a hardware failure. The alarm recognition model uses a support vector machine algorithm, a random forest algorithm or a convolutional neural network to train and classify feature information; Step S5: Decision making Customize the corresponding handling measures for the fault type; After the trained machine learning model is used to identify hardware faults, the corresponding processing measures are triggered according to the identified fault type; Step S6: User interaction The alarm information is notified to users or maintenance personnel through various customized methods so that timely measures can be taken.

2. The method for processing hardware alarms of intelligent network devices according to claim 1, characterized in that: In step S4, the alarm recognition model is used to identify whether there is a hardware fault, and the steps are as follows: Step S4.1, customizing and selecting historical alarm data of network devices to build an alarm label library, and using the alarm label library to evaluate the alarm recognition model; Step S4.2: Build an alarm rule library and establish a data synchronization mechanism with the alarm tag library; Step S4.3: After the current hardware status data is obtained, first determine whether the current hardware status data exists in the alarm rule library: If it does not exist, the alarm recognition model is input for recognition and judgment, and the recognition result is output; If it exists, the corresponding fault category will be output directly.

3. The method for processing hardware alarms of intelligent network devices according to claim 2, characterized in that: In step S4.3, manual troubleshooting is performed and feedback is provided on the recognition results output by the alarm recognition model: If the fault category identification result is correct, the corresponding hardware status data and fault category are written into the alarm rule library; If the fault category identification result is wrong, the corresponding hardware status data will be input into the alarm recognition model again, and the alarm recognition model will be adjusted and optimized.

4. An intelligent network equipment hardware alarm processing system, characterized in that: include: The data collection module is responsible for collecting the hardware status data generated by the network equipment during operation according to the custom cycle; The data processing module is responsible for preprocessing the collected hardware status data, including filtering, noise reduction and smoothing, and removing erroneous values; The feature extraction module is responsible for defining key features based on device information and extracting key feature information from preprocessed data; The alarm recognition module is responsible for customizing the threshold and using the trained alarm recognition model to analyze the feature information and identify whether there is a hardware failure. The alarm recognition model uses a support vector machine algorithm, a random forest algorithm or a convolutional neural network to train and classify feature information; The decision-making processing module is responsible for customizing the processing measures corresponding to the fault type. After identifying the hardware fault using the trained machine learning model, it triggers the corresponding processing measures according to the identified fault type. The user interaction module is responsible for notifying users or maintenance personnel of alarm information through various customized methods so that timely measures can be taken.

5. The intelligent network device hardware alarm processing system according to claim 4, characterized in that: The alarm identification module uses the alarm identification model to identify whether there is a hardware fault, the steps are as follows: Step S4.1, customizing and selecting historical alarm data of network devices to build an alarm label library, and using the alarm label library to evaluate the alarm recognition model; Step S4.2: Build an alarm rule library and establish a data synchronization mechanism with the alarm tag library; Step S4.3: After the current hardware status data is obtained, first determine whether the current hardware status data exists in the alarm rule library: If it does not exist, the alarm recognition model is input for recognition and judgment, and the recognition result is output; If it exists, the corresponding fault category will be output directly.

6. The intelligent network device hardware alarm processing system according to claim 5, characterized in that: In step S4.3, manual troubleshooting is performed and feedback is provided on the recognition results output by the alarm recognition model: If the fault category identification result is correct, the corresponding hardware status data and fault category are written into the alarm rule library; If the fault category identification result is wrong, the corresponding hardware status data will be input into the alarm recognition model again, and the alarm recognition model will be adjusted and optimized.

7. An intelligent network equipment hardware alarm processing device, characterized in that: The method comprises a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the method according to any one of claims 1 to 3 when executing the computer program.

8. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Images