Alarm information fault locating method and system based on markov inference

Abstract

The application relates to the technical field of network communication fault positioning, and provides an alarm information fault positioning method and system based on Markov inference, which comprises the following steps: evaluating device state parameters and network link state parameters collected in real time by using a support vector machine, and outputting a fault cause judgment result; summarizing historical data accumulated by a network management system into a state transition probability matrix used for describing the probability of a communication network system shifting from a certain working state to another working state, and calculating a fault cause positioning judgment result of a Markov model through the state transition probability matrix, wherein a working state sequence of the communication network system is calculated through the Markov model; and fusing the fault cause judgment result and the fault cause positioning judgment result through a DS evidence theory to obtain a final fault cause positioning result. The application can help operation and maintenance personnel to timely find key factors causing network faults, and improve the accuracy of network fault diagnosis and the overall operation and maintenance quality.

Description

Technical Field

[0001] This invention relates to the field of network communication fault location technology, and in particular to a fault location method and system for alarm information based on Markov inference. Background Technology

[0002] Traditional methods for fault location and diagnosis in communication networks have the following main problems: 1. Fault feature knowledge bases are usually built using historical data, but the logical relationships between state transitions are not considered, nor are comprehensive considerations made of state information such as power supply status, fan status, CPU utilization, and temperature.

[0003] 2. In network operation and maintenance, situations often arise where a single type of fault leads to multiple abnormal indicators, or a single device failure causes multiple node parameter errors. In real-world scenarios, simply listing and labeling anomalies in tables can only handle the ideal situation where abnormal parameters correspond one-to-one with the cause of the fault. For more common and complex real-world scenarios, machine learning methods are needed to map and analyze abnormal indicators against the causes of faults. Summary of the Invention

[0004] The purpose of this invention is to solve at least one technical problem in the background art and to provide a method and system for locating alarm information faults based on Markov inference.

[0005] To achieve the above objectives, this invention provides a fault location method for alarm information based on Markov inference, comprising: The device status parameters and network link status parameters collected in real time are evaluated by support vector machine, and the results of fault cause judgment are output. The historical data accumulated by the network management system is summarized into a state transition probability matrix to describe the probability of the communication network system transitioning from one working state to another. The fault cause location judgment result of the Markov model is calculated through the state transition probability matrix. The working state sequence of the communication network system is calculated through the Markov model. The fault cause determination result and the fault cause location determination result are fused by DS evidence theory to obtain the final fault cause location result.

[0006] According to one aspect of the present invention, the step of evaluating the real-time collected device status parameters and network link status parameters using a support vector machine and outputting a fault cause determination result includes: Support Vector Machines (SVMs) map the device state parameters and network link state parameters of the sample data to a Hilbert space using kernel functions, constructing a hyperplane. After discrimination by the hyperplane, the sample data is finally classified into multiple categories based on fault cause determination. The classification function of the hyperplane is: ; in, The result of the fault cause judgment when the sample data is x represents Represents support vectors; Represents the Lagrange multipliers; Represents the kernel function; , is the support vector in the specific category; b represents the threshold.

[0007] According to one aspect of the invention, it further includes: representing the fault cause determination result in probabilistic form, and using the Sigmoid function to obtain the probabilistic output result, expressed as: ; in, , The morphological parameters of the Sigmoid function are obtained by solving the maximum likelihood problem.

[0008] According to one aspect of the present invention, the operating state sequence of the communication network is calculated using a Markov model as follows: The operating state sequence of the communication network is considered as a Markov process, that is, the operating state sequence of the communication network in... The state at any given time is determined by and only by The state at any given moment is determined, and is represented as: ; Where q is the state transition probability. , for time, The actual working status of the communication network at all times includes both normal working status and abnormal working status.

[0009] According to one aspect of the present invention, the step of summarizing historical data accumulated by the network management system into a state transition probability matrix to describe the probability of a communication network transitioning from one operating state to another, and calculating the fault cause localization judgment result of the Markov model through the state transition probability matrix, includes: Historical data accumulated during the actual operation of the network management system are summarized into a state transition probability matrix to describe the probability of the communication network transitioning from one operating state to another. , represented as: ; in, express The working status at all times is hour, The working status at all times is The probability, 0≤ ≤1; Through the state transition probability matrix Calculate the fault cause localization judgment results of the Markov model , represented as: , ; Among them, X t-1 This refers to the work status at the previous moment. Represents the state transition probability matrix; The state transition probability matrix at any given time is assumed to be an equal probability matrix.

[0010] According to one aspect of the present invention, the fault cause determination result and the fault cause location determination result are fused using DS evidence theory to obtain the final fault cause location result. for: ; ; in, represent , represent .

[0011] To achieve the above objectives, the present invention also provides an alarm information fault location system based on Markov inference, comprising: The SVM evaluation module evaluates the real-time collected device status parameters and network link status parameters using a support vector machine, and outputs the fault cause judgment result. The Markov inference module summarizes the historical data accumulated by the network management system into a state transition probability matrix, which describes the probability of the communication network system transitioning from one working state to another. The fault cause localization judgment result of the Markov model is calculated through the state transition probability matrix. The working state sequence of the communication network system is calculated through the Markov model. The DS fusion evaluation module fuses the fault cause judgment result and the fault cause location judgment result through DS evidence theory to obtain the final fault cause location result.

[0012] To achieve the above objectives, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the alarm information fault location method based on Markov inference as described above.

[0013] To achieve the above objectives, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the alarm information fault location method based on Markov inference as described above.

[0014] According to the present invention, the present invention first uses SVM to obtain the fault location result at the current moment based on the indicator layer features, then uses the working state transition probability matrix to calculate the Markov prediction result, and finally uses DS evidence theory to fuse the two to obtain the final fault location result. The present invention establishes an evaluation indicator system based on the data extraction capabilities of the integrated network management system, uses the SVM machine learning method to determine the network fault location, and fully extracts multi-dimensional information, making the fault location reasonable and interpretable. It innovatively approximates the correlation between the current state of the network and the state at the previous moment as a Markov process, mining the historical data value that is easily overlooked in previous methods, further improving the accuracy of fault location. It uses the current indicator judgment result and the judgment result inferred from historical experience as independent evidence for DS fusion, making up for the deficiencies of incomplete processing and insufficient fusion judgment capability of uncertain alarm information in operation and maintenance scenarios, which is conducive to improving the overall network maintenance quality.

[0015] This invention helps operations and maintenance personnel better utilize historical system data to promptly identify key factors that lead to network failures, thereby improving the accuracy of network fault diagnosis and the overall quality of operations and maintenance. Attached Figure Description

[0016] Figure 1 The flowchart illustrates a method for locating alarm information faults based on Markov inference according to an embodiment of the present invention. Detailed Implementation

[0017] The invention will now be discussed with reference to exemplary embodiments. It should be understood that the described embodiments are merely intended to enable those skilled in the art to better understand and thus implement the invention, and are not intended to imply any limitation on the scope of the invention.

[0018] As used herein, the term "comprising" and its variations are to be interpreted as open-ended terms meaning "including but not limited to". The term "based on" is to be interpreted as "at least partially based on". The terms "one embodiment" and "an embodiment" are to be interpreted as "at least one embodiment".

[0019] Figure 1 The flowchart schematically illustrates a Markov inference-based alarm information fault location method according to an embodiment of the present invention. Figure 1As shown, in this embodiment, the alarm information fault location method based on Markov inference includes: The device status parameters and network link status parameters collected in real time are evaluated by support vector machine (SVM), and the results of fault cause judgment are output. The historical data accumulated by the network management system is summarized into a state transition probability matrix to describe the probability of the communication network system transitioning from one working state to another. The fault cause location judgment result of the Markov model is calculated through the state transition probability matrix. The working state sequence of the communication network system is calculated through the Markov model. By integrating the fault cause judgment results and fault cause location judgment results using the DS evidence theory, the final fault cause location result is obtained.

[0020] In this implementation, based on common protocol standards such as SNMP, Syslog, and Netflow of the integrated network management system, network device operating status data (including device status parameters and network link status parameters) is collected periodically and then characterized. This includes device status metrics such as NTP clock, power status, fan status, CPU utilization, temperature, and primary / standby engine status. For these metrics, an anomaly results in a score of 0, while the absence of anomalies results in a score of 1. Simultaneously, network link performance metrics such as link latency, bandwidth utilization, and packet loss rate are collected, with a score range of 0 to 1, where 1 indicates optimal performance.

[0021] Furthermore, according to one embodiment of the present invention, the real-time collected device status parameters and network link status parameters are evaluated using a support vector machine to output a fault cause judgment result, including: Support Vector Machines (SVMs) map the device state parameters and network link state parameters of the sample data to a Hilbert space using kernel functions, constructing a hyperplane. After discrimination by the hyperplane, the sample data is finally classified into multiple categories based on fault cause determination. The classification function of the hyperplane is: ; in, The result of the fault cause judgment when the sample data is x represents Represents support vectors; Represents the Lagrange multipliers; Represents the kernel function; , is the support vector in the specific category; b represents the threshold.

[0022] In this embodiment, SVM is used as the fault location classifier, and evaluation metrics (i.e., device status parameters and network link status parameters) are used as input features of the SVM. Optical device faults, network configuration errors, and line blockages are output fault causes. In this embodiment, the SVM is trained using a known knowledge base to obtain the mapping relationship between the evaluation metrics and the fault evaluation results of the criterion layer. Specifically, a kernel function maps the sample data to a Hilbert space, constructing a hyperplane to transform a non-linearly bisectable problem into a linearly separable one. This hyperplane satisfies the condition that the distance between each sample and the hyperplane is as large as possible, while minimizing the classification error. After discrimination by the hyperplane, the final sample data is classified into multiple fault cause judgment results by the aforementioned classification function.

[0023] Furthermore, according to one embodiment of the present invention, the present invention further includes: representing the fault cause determination result in probabilistic form, and using the Sigmoid function to obtain the probabilistic output result, expressed as: ; in, , The morphological parameters of the Sigmoid function, obtained by solving the maximum likelihood problem, are expressed as: ; ; in, , , This represents the number of positive and negative class samples in the SVM decision matrix.

[0024] In this embodiment, the SVM output is converted into a soft output to facilitate probabilistic fusion with the Markov inference results.

[0025] Furthermore, according to one embodiment of the present invention, the operating state sequence of the communication network is calculated using a Markov model as follows: The operating state sequence of the communication network is considered as a Markov process, that is, the operating state sequence of the communication network in... The state at any given time is determined by and only by The state at any given moment is determined, and is represented as: ; in, Let be the state transition probability. , for time, The actual working status of the communication network at all times includes both normal working status and abnormal working status caused by various faults.

[0026] In this embodiment, the communication network is considered. Constant working status and The network's operating state is highly correlated at any given time and exhibits Markov characteristics, so the sequence of network operating states is regarded as a Markov process and calculated using a Markov model.

[0027] Furthermore, according to one embodiment of the present invention, historical data accumulated by the network management system is summarized into a state transition probability matrix to describe the probability of the communication network transitioning from one working state to another. The fault cause localization judgment result of the Markov model is calculated using the state transition probability matrix, including: Historical data accumulated during the actual operation of the network management system are summarized into a state transition probability matrix to describe the probability of the communication network transitioning from one operating state to another. , represented as: ; in, express The working status at all times is hour, The working status at all times is The probability, 0≤ ≤1, =1; Through the state transition probability matrix Calculate the fault cause localization judgment results of the Markov model , represented as: , ; Among them, X t-1 This refers to the work status at the previous moment. Represents the state transition probability matrix; The state transition probability matrix at any given time is assumed to be an equal probability matrix.

[0028] Furthermore, according to one embodiment of the present invention, the fault cause judgment result and the fault cause location judgment result are fused using DS evidence theory to obtain the final fault cause location result. Specifically, fuzzy reasoning is employed using DS evidence theory to fuse the basic probability assignments of multiple pieces of evidence. This fusion process is part of the identification framework. To achieve the target, It consists of mutually independent and mutually exclusive target objects. Let... , They are the same identification frame The two pieces of evidence above represent the DS fusion of the two as follows: ; ; in, The "evidence 1" representing participation in DS fusion, i.e., the fault cause determined by the support vector machine method in this embodiment, is... ; Evidence 2, representing participation in DS fusion, refers to the fault cause determined using the Markov method in this embodiment. By integrating the two through the DS evidence theory, the final result of fault cause localization is obtained. .

[0029] According to the above-described scheme of this invention, the invention first uses SVM to obtain the fault location result at the current moment based on the indicator layer features, then uses the working state transition probability matrix to calculate the Markov prediction result, and finally uses DS evidence theory to fuse the two to obtain the final fault location result. This invention establishes an evaluation indicator system based on the data extraction capabilities of the integrated network management system, uses the SVM machine learning method to determine the network fault location, and fully extracts multi-dimensional information, making the fault location reasonable and interpretable. It innovatively approximates the correlation between the current state of the network and the state at the previous moment as a Markov process, mining the historical data value that is easily overlooked in previous methods, further improving the accuracy of fault location. It uses the current indicator judgment result and the judgment result inferred from historical experience as independent evidence for DS fusion, making up for the deficiencies of incomplete processing and insufficient fusion judgment capability of uncertain alarm information in operation and maintenance scenarios, which is conducive to improving the overall network maintenance quality.

[0030] This invention helps operations and maintenance personnel better utilize historical system data to promptly identify key factors that lead to network failures, thereby improving the accuracy of network fault diagnosis and the overall quality of operations and maintenance.

[0031] Furthermore, to achieve the above objectives, the present invention also provides an alarm information fault location system based on Markov inference, comprising: The SVM evaluation module evaluates the real-time collected device status parameters and network link status parameters using a support vector machine, and outputs the fault cause judgment result. The Markov inference module summarizes the historical data accumulated by the network management system into a state transition probability matrix, which describes the probability of the communication network system transitioning from one working state to another. The fault cause localization judgment result of the Markov model is calculated through the state transition probability matrix. The working state sequence of the communication network system is calculated through the Markov model. The DS fusion evaluation module fuses the fault cause judgment result and the fault cause location judgment result through DS evidence theory to obtain the final fault cause location result.

[0032] In this implementation, based on common protocol standards such as SNMP, Syslog, and Netflow of the integrated network management system, network device operating status data (including device status parameters and network link status parameters) is collected periodically and then characterized. This includes device status metrics such as NTP clock, power status, fan status, CPU utilization, temperature, and primary / standby engine status. For these metrics, an anomaly results in a score of 0, while the absence of anomalies results in a score of 1. Simultaneously, network link performance metrics such as link latency, bandwidth utilization, and packet loss rate are collected, with a score range of 0 to 1, where 1 indicates optimal performance.

[0033] Furthermore, according to one embodiment of the present invention, the real-time collected device status parameters and network link status parameters are evaluated using a support vector machine to output a fault cause judgment result, including: Support Vector Machines (SVMs) map the device state parameters and network link state parameters of the sample data to a Hilbert space using kernel functions, constructing a hyperplane. After discrimination by the hyperplane, the sample data is finally classified into multiple categories based on fault cause determination. The classification function of the hyperplane is: ; in, The result of the fault cause judgment when the sample data is x represents Represents support vectors; Represents the Lagrange multipliers; Represents the kernel function; , is the support vector in the specific category; b represents the threshold.

[0034] In this embodiment, SVM is used as the fault location classifier, and evaluation metrics (i.e., device status parameters and network link status parameters) are used as input features of the SVM. Optical device faults, network configuration errors, and line blockages are output fault causes. In this embodiment, the SVM is trained using a known knowledge base to obtain the mapping relationship between the evaluation metrics and the fault evaluation results of the criterion layer. Specifically, a kernel function maps the sample data to a Hilbert space, constructing a hyperplane to transform a non-linearly bisectable problem into a linearly separable one. This hyperplane satisfies the condition that the distance between each sample and the hyperplane is as large as possible, while minimizing the classification error. After discrimination by the hyperplane, the final sample data is classified into multiple fault cause judgment results by the aforementioned classification function.

[0035] Furthermore, according to one embodiment of the present invention, the present invention further includes: representing the fault cause determination result in probabilistic form, and using the Sigmoid function to obtain the probabilistic output result, expressed as: ; in, , The morphological parameters of the Sigmoid function, obtained by solving the maximum likelihood problem, are expressed as: ; ; in, , , This represents the number of positive and negative class samples in the SVM decision matrix.

[0036] In this embodiment, the SVM output is converted into a soft output to facilitate probabilistic fusion with the Markov inference results.

[0037] Furthermore, according to one embodiment of the present invention, the operating state sequence of the communication network is calculated using a Markov model as follows: The operating state sequence of the communication network is considered as a Markov process, that is, the operating state sequence of the communication network in... The state at any given time is determined by and only by The state at any given moment is determined, and is represented as: ; in, Let be the state transition probability. , for time, The actual working status of the communication network at all times includes both normal working status and abnormal working status caused by various faults.

[0038] In this embodiment, the communication network is considered. Constant working status and The network's operating state is highly correlated at any given time and exhibits Markov characteristics, so the sequence of network operating states is regarded as a Markov process and calculated using a Markov model.

[0039] Furthermore, according to one embodiment of the present invention, historical data accumulated by the network management system is summarized into a state transition probability matrix to describe the probability of the communication network transitioning from one working state to another. The fault cause localization judgment result of the Markov model is calculated using the state transition probability matrix, including: Historical data accumulated during the actual operation of the network management system are summarized into a state transition probability matrix to describe the probability of the communication network transitioning from one operating state to another. , represented as: ; in, express The working status at all times is hour, The working status at all times is The probability, 0≤ ≤1, =1; Through the state transition probability matrix Calculate the fault cause localization judgment results of the Markov model , represented as: , ; Among them, X t-1 This refers to the work status at the previous moment. Represents the state transition probability matrix; The state transition probability matrix at any given time is assumed to be an equal probability matrix.

[0040] Furthermore, according to one embodiment of the present invention, the fault cause judgment result and the fault cause location judgment result are fused using DS evidence theory to obtain the final fault cause location result. Specifically, fuzzy reasoning is employed using DS evidence theory to fuse the basic probability assignments of multiple pieces of evidence. This fusion process is part of the identification framework. To achieve the target, It consists of mutually independent and mutually exclusive target objects. Let... , They are the same identification frame The two pieces of evidence above represent the DS fusion of the two as follows: ; ; in, The "evidence 1" representing participation in DS fusion, i.e., the fault cause determined by the support vector machine method in this embodiment, is... ; Evidence 2, representing participation in DS fusion, refers to the fault cause determined using the Markov method in this embodiment. By integrating the two through the DS evidence theory, the final result of fault cause localization is obtained. .

[0041] According to the above-described scheme of this invention, the invention first uses SVM to obtain the fault location result at the current moment based on the indicator layer features, then uses the working state transition probability matrix to calculate the Markov prediction result, and finally uses DS evidence theory to fuse the two to obtain the final fault location result. This invention establishes an evaluation indicator system based on the data extraction capabilities of the integrated network management system, uses the SVM machine learning method to determine the network fault location, and fully extracts multi-dimensional information, making the fault location reasonable and interpretable. It innovatively approximates the correlation between the current state of the network and the state at the previous moment as a Markov process, mining the historical data value that is easily overlooked in previous methods, further improving the accuracy of fault location. It uses the current indicator judgment result and the judgment result inferred from historical experience as independent evidence for DS fusion, making up for the deficiencies of incomplete processing and insufficient fusion judgment capability of uncertain alarm information in operation and maintenance scenarios, which is conducive to improving the overall network maintenance quality.

[0042] This invention helps operations and maintenance personnel better utilize historical system data to promptly identify key factors that lead to network failures, thereby improving the accuracy of network fault diagnosis and the overall quality of operations and maintenance.

[0043] Furthermore, to achieve the above objectives, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the alarm information fault location method based on Markov inference as described above.

[0044] Furthermore, to achieve the above objectives, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the alarm information fault location method based on Markov inference as described above.

[0045] Those skilled in the art will recognize that the modules and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented using electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0046] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and equipment can be referred to the corresponding process in the foregoing method implementation, and will not be repeated here.

[0047] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0048] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the objectives of the embodiments of the present invention, depending on actual needs.

[0049] In addition, the functional modules in the embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0050] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the sending / receiving methods of various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0051] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in this application.

[0052] It should be understood that the sequence number of each step in the invention and its embodiments does not absolutely imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Claims

1. A fault location method for alarm information based on Markov inference, characterized in that, include: The device status parameters and network link status parameters collected in real time are evaluated by support vector machine, and the results of fault cause judgment are output. The historical data accumulated by the network management system is summarized into a state transition probability matrix to describe the probability of the communication network system transitioning from one working state to another. The fault cause location judgment result of the Markov model is calculated through the state transition probability matrix. The working state sequence of the communication network system is calculated through the Markov model. The fault cause judgment result and the fault cause location judgment result are fused by DS evidence theory to obtain the final fault cause location result; The process of evaluating real-time collected device status parameters and network link status parameters using a support vector machine and outputting fault cause judgment results includes: Support Vector Machines (SVMs) map the device state parameters and network link state parameters of the sample data to a Hilbert space using kernel functions, constructing a hyperplane. After discrimination by the hyperplane, the sample data is finally classified into multiple categories based on fault cause determination. The classification function of the hyperplane is: ； in, The result of fault cause determination when the sample data is x. Represents support vectors; Represents the Lagrange multipliers; Represents the kernel function; , 'b' represents the support vectors in a specific category; 'b' represents the threshold. The operating state sequence of the communication network is calculated using a Markov model as follows: The operating state sequence of the communication network is considered as a Markov process, that is, the operating state sequence of the communication network in... The state at any given time is determined by and only by The state at any given moment is determined, and is represented as: ； in, Let be the state transition probability. , for time, The actual working status of the communication network at all times, including both normal working status and abnormal working status; The process involves summarizing historical data accumulated by the network management system into a state transition probability matrix, which describes the probability of a communication network transitioning from one operating state to another. The fault cause localization result of the Markov model is then calculated using this state transition probability matrix, including: Historical data accumulated during the actual operation of the network management system are summarized into a state transition probability matrix to describe the probability of the communication network transitioning from one operating state to another. , represented as: ； in, express The working status at all times is hour, The working status at all times is The probability, 0≤ ≤1; Through the state transition probability matrix Calculate the fault cause localization judgment results of the Markov model , represented as: ， ； Among them, X t-1 This refers to the work status at the previous moment. Represents the state transition probability matrix; The state transition probability matrix at each time step is assumed to be an equal probability matrix. The fault cause determination result and the fault cause location determination result are fused using the DS evidence theory to obtain the final fault cause location result. for: ； ； in, represent ; represent .

2. The alarm information fault location method based on Markov inference according to claim 1, characterized in that, Also includes: The fault cause determination result is expressed in probabilistic form, and the probability output result is obtained using the Sigmoid function, as follows: ； in, , The morphological parameters of the Sigmoid function are obtained by solving the maximum likelihood problem.

3. A Markov reasoning-based alarm information fault location system that implements the Markov reasoning-based alarm information fault location method of claim 1 or 2, characterized in that, include: The SVM evaluation module evaluates the real-time collected device status parameters and network link status parameters using a support vector machine, and outputs the fault cause judgment result. The Markov inference module summarizes the historical data accumulated by the network management system into a state transition probability matrix, which describes the probability of the communication network system transitioning from one working state to another. The fault cause localization judgment result of the Markov model is calculated through the state transition probability matrix. The working state sequence of the communication network system is calculated through the Markov model. The DS fusion evaluation module fuses the fault cause judgment result and the fault cause location judgment result through DS evidence theory to obtain the final fault cause location result.

4. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the alarm information fault location method based on Markov inference as described in claim 1 or 2.

5. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the alarm information fault location method based on Markov inference as described in claim 1 or 2.

Images