CN113132125B - Fault cause positioning method and device, electronic equipment and storage medium - Google Patents
Fault cause positioning method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113132125B CN113132125B CN201911392356.3A CN201911392356A CN113132125B CN 113132125 B CN113132125 B CN 113132125B CN 201911392356 A CN201911392356 A CN 201911392356A CN 113132125 B CN113132125 B CN 113132125B
- Authority
- CN
- China
- Prior art keywords
- fault
- alarm data
- training model
- model
- fault cause
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0677—Localisation of faults
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Debugging And Monitoring (AREA)
- Telephonic Communication Services (AREA)
Abstract
The application discloses a fault cause positioning method and device, electronic equipment and a storage medium, and relates to the technical field of communication. The fault cause positioning method comprises the following steps: preprocessing historical alarm data reported by a fault management system; optimizing an initial weight threshold of the training model through a PSO algorithm until a target function algorithm corresponding to the training model converges; based on the optimized initial weight threshold, training by taking the preprocessed historical alarm data as the input of a training model and taking the fault reason corresponding to the historical alarm data as the output of the training model to obtain a fault reason positioning model; and carrying out fault cause positioning based on the fault cause positioning model. The fault cause positioning method, the fault cause positioning device, the electronic equipment and the storage medium can quickly and accurately obtain the fault cause of the large-area fault.
Description
Technical Field
The present application relates to the field of communications technologies, and in particular, to a method and an apparatus for locating a fault cause, an electronic device, and a storage medium.
Background
In the process of providing communication service for users, if a communication service has a large-area fault, the user service in a corresponding coverage area is affected, complaints of the service of the user may be increased suddenly, and even the enterprise image is damaged and the service is slid down.
At present, the manual analysis mode is mostly adopted for positioning the cause of the large-area fault, when the cause of the large-area fault is positioned and analyzed, the analyst is dependent on the skilled grasp of business logic and accurate resource topology, and the investigation of the resource topology cannot be completed at once no matter the grasp capability of the business logic is not successful in one day, so the efficiency and the accuracy for positioning the cause of the large-area fault cannot meet the expected requirements by adopting the manual analysis mode.
Therefore, when a large-area fault occurs, how to quickly and accurately locate the fault cause to support the maintenance process and reduce the loss caused by the large-area fault has become a problem to be solved in the prior art.
Disclosure of Invention
The embodiment of the application provides a fault cause positioning method and device, electronic equipment and a storage medium, so as to at least solve the problem that the efficiency and the accuracy of large-area fault cause positioning in the related technology are low.
The technical scheme of the application is as follows:
a fault cause positioning method comprises the following steps:
preprocessing historical alarm data reported by a fault management system;
optimizing an initial weight threshold value of a training model through a PSO algorithm until a target function algorithm corresponding to the training model converges;
and based on the optimized initial weight threshold, training by taking the preprocessed historical alarm data as the input of the training model and taking the fault reason corresponding to the historical alarm data as the output of the training model to obtain a fault reason positioning model.
A fault cause locating device comprising:
the system comprises a preprocessing module, a fault management module and a fault detection module, wherein the preprocessing module is configured to preprocess historical alarm data reported by a fault management system;
the optimization module is configured to optimize an initial weight threshold of a training model through a PSO algorithm until an objective function algorithm corresponding to the training model converges;
and the training module is configured to train the preprocessed historical alarm data as the input of the training model and the fault reason corresponding to the historical alarm data as the output of the training model based on the optimized initial weight threshold value to obtain a fault reason positioning model.
An electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, the computer program realizing the above-mentioned method steps when executed by the processor.
A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method steps of the above.
The embodiment of the application adopts at least one technical scheme which can achieve the following beneficial effects:
because the fault cause positioning model is obtained by training by taking the historical alarm data reported by the fault management system as the input of the training model and taking the fault cause corresponding to the historical alarm data as the output of the training model under the condition that the initial weight threshold of the training model is optimized through the PSO algorithm until the algorithm converges, the fault cause positioning model can quickly and accurately obtain the fault cause of a large-area fault so as to support the maintenance processing process and reduce the loss caused by the large-area fault.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a schematic diagram of an application environment provided in the embodiment of the present application.
Fig. 2 is a flowchart of a fault cause positioning method provided in an embodiment of the present application.
Fig. 3 is a flowchart of optimizing the initial weight threshold of the training model by using the PSO algorithm according to the embodiment of the present application.
Fig. 4 is a flowchart of fault cause location based on a fault cause location model according to an embodiment of the present application.
Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Fig. 6 is a schematic structural diagram of a fault cause positioning device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be described in detail and completely with reference to the following specific embodiments of the present application and the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
In order to quickly and accurately locate a fault cause when a communication service has a large-area fault, embodiments of the present application provide a fault cause locating method, apparatus, electronic device, and storage medium, where the fault cause locating method, apparatus, electronic device, and storage medium can quickly and accurately locate a fault cause when a communication service provided by a communication operator for a user has a large-area fault.
First, in order to more intuitively understand the solution provided by the embodiment of the present application, a system architecture of the fault cause location model training solution provided by the embodiment of the present application is described below with reference to fig. 1.
Fig. 1 is a schematic view of an application environment of a fault cause location method, an apparatus, an electronic device, and a storage medium according to one or more embodiments of the present application. As shown in fig. 1, electronic device 100 is communicatively coupled to fault management system 200 via network 300 for data communication or interaction. The electronic device 100 may be a server or a user terminal device (e.g., a personal computer), etc. The fault management system can be a server used for collecting transmission service alarm, dynamic loop monitoring alarm, wireless service alarm, home broadband service alarm and the like in the process of providing communication service for users. The network 300 may be a wired or wireless network.
The method for locating the cause of the fault provided in the embodiment of the present application will be described in detail below.
The method for locating a fault cause provided in the embodiments of the present application may be applied to the electronic device 100, and for convenience of description, except for special description, the embodiments of the present application all use the electronic device 100 as an execution subject for description.
It is to be understood that the subject of the present application is not to be construed as limiting the embodiments of the present application.
Specifically, the method for locating the fault cause is shown in fig. 2, and may include the following steps:
and step S21, preprocessing the historical alarm data reported by the fault management system.
In this embodiment, the historical alarm data may be the number of transmission service alarms, dynamic loop monitoring alarms, wireless service alarms, home broadband service alarms, and other alarms that are reported by the fault management system in a certain time period before.
The wireless service alarm may include a base station service quit alarm and a base station side point cloud alarm, the transmission service alarm may include a Packet Transport Network (PTN)/Synchronous Digital Hierarchy (SDH) side on-off class alarm, a PTN/SDH temperature class alarm, a PTN/SDH device class alarm, a PTN/SDH performance class alarm, a wavelength division line side on-off class alarm, a wavelength division temperature class alarm, a wavelength division device class alarm, a wavelength division performance class alarm, and the like, the moving ring monitoring alarm may include a moving ring monitoring power down class alarm and a moving ring monitoring voltage class alarm, and the home broadband service alarm may include an Optical Line Terminal (OLT) service quit class alarm.
The certain time period may be set according to actual conditions, and may be set to 15 minutes, 30 minutes, or the like, for example. The time interval for reporting the dynamic ring monitoring alarm can be set relatively longer.
The historical alarm data may have certain noise data, so that when the historical alarm data is preprocessed, the historical alarm data can be firstly dried, and invalid data and error data in the historical alarm data are removed. The invalid data and the error data mainly refer to some data deviating from a normal range, such as that the alarm quantity is a non-integer or a negative number, or the alarm quantity is far beyond a normal value.
For example, a historical alarm data is (5, 7, 4, 2.8, 6), wherein 5 numbers represent the number of 5 types of alarms respectively. Since the historical alarm data is the number of the transmission service alarm, the dynamic loop monitoring alarm, the wireless service alarm, the home broadband service alarm and other alarms in a certain time period reported by the fault management system, and is inevitably 0 or a positive integer, the detail of the data of 2.8 in the historical alarm data is wrong, and the abnormal data needs to be removed at this time.
After invalid data and error data in the historical alarm data are removed, missing data in the historical alarm data can be complemented. The data completion may adopt, but is not limited to, a mean value substitution method, a regression substitution method, and the like, and is not specifically limited in this embodiment.
It can be understood that if there is no data deviating from the normal range in the historical alarm data, the data may not be removed and supplemented.
After the data completion, the historical alarm data of the data completion is normalized to obtain a multidimensional vector. In the embodiment of the application, the historical alarm data includes alarm numbers of 13 types of alarms such as base station service class quitting alarm number, base station side point cloud class alarm number, PTN/SDH side on-off class alarm number, PTN/SDH temperature class alarm number, PTN/SDH device class alarm number, PTN/SDH performance class alarm number, wavelength division line side on-off class alarm number, wavelength division temperature class alarm number, wavelength division device class alarm number, wavelength division performance class alarm number, dynamic ring monitoring power failure class alarm number, dynamic ring monitoring voltage class alarm number, OLT service class quitting alarm number and the like within a certain time period, so that after normalization is performed, a multidimensional vector containing 13 pieces of dimensional data can be obtained.
And step S22, optimizing the initial weight threshold value of the training model through the PSO algorithm until the objective function algorithm corresponding to the training model converges.
Referring to fig. 3, step S22 includes the following steps:
and S31, calculating the variation probability of the global optimal particles in the population after population initialization, fitness calculation, individual optimal particle update and global optimal particle update of the PSO algorithm.
The standard PSO algorithm has four steps of population initialization, fitness calculation, individual optimal particle updating and global optimal particle updating, and a specific description is not given in the embodiment of the application.
In the standard PSO algorithm, the speed and position of the particle at the next time are both determined by the current position and the current speed, and the current position is determined by the speed, the individual extremum and the global optimum extremum at the previous time. If the algorithm converges early, the globally optimal extremum is not necessarily globally optimal. Therefore, after population initialization, fitness calculation, individual optimal particle updating and global optimal particle updating of the PSO algorithm, global optimal self-variation operation is added to expand the search space of the particles.
Specifically, the variation probability of the globally optimal particles in the population is firstly calculated, and the calculation formula of the variation probability of the globally optimal particles is
Wherein k is the number of iterations, p is the expansion constant that changes the rate of change of the exponential curve,
is the population fitness variance, p, of the k-th generation max As maximum value of the mutation probability, p min Is the minimum of the mutation probability.
And S32, performing mutation on the global optimal extreme value based on the mutation probability of the global optimal particle.
From the above formula, the closer the particles are, the less the population diversity is, the smaller the population fitness variance is, and the larger the variation probability of the corresponding global optimum extreme value is. Conversely, the more significant the population diversity, the smaller the variation probability of the global optimum extremum. When the variance of the population fitness is larger, the smaller the variation probability of the global extreme value is, the better the variation probability is, the introduced exponential function has a faster descending trend which is just in line with the descending trend, so that the PSO algorithm has a better variation speed.
Therefore, a random number lambda is introduced, and the global optimum extreme value is mutated by the random number lambda and the mutation probability of the global optimum particle. The calculation formula for performing the mutation on the global optimum extreme value is
Where λ is a random number, η is a random number satisfying a standard normal distribution, and Gaussian (σ) is a random number following a Gaussian distribution with a standard deviation of σ.
And S33, updating the speed and the position of the particles in the population to find a global optimal extreme value.
Wherein the calculation formula for updating the speed and the position of the particles in the population is
And X k+1 ＝X k +V k+1 Where ω is the inertial weight, r 1 And r 2 Is a random number distributed over the interval 0 to 1, k is the current iteration number,
for the individual optimal particle position at the kth iteration,
for the global optimal particle position at the kth iteration (global optimal extremum), c 1 And c 2 Is a constant number, V k Is the speed at the k-th iteration, V k+1 Speed at the k +1 th iteration, X k Is the particle position at the kth iteration, X k+1 The particle position at the (k + 1) th iteration.
After the speed and the position of the particles in the population are updated, the updated global optimum extreme value can be obtained according to the updated speed and the updated position of the particles.
S34, judging whether the convergence condition is satisfied, if not, executing the step S35; if so, step S36 is performed.
Whether the convergence condition is met or not can be that the global optimal extreme value is used as an initial weight threshold value of the training model and the target function algorithm corresponding to the training model converges or the iteration frequency of the PSO algorithm reaches a preset frequency.
For example, the iteration number of the PSO algorithm is set to 50, and the current iteration number is exactly the 50 th iteration number, so that the global optimal extreme value obtained through calculation after the current iteration can be used as the initial weight threshold of the training model.
In the embodiment of the application, the global optimal extreme value obtained by each updating can be used as the initial weight threshold of the training model for operation, the error between the output value and the actual value of the training model is calculated, and if the error between the output value and the actual value of the training model is smaller than the preset threshold, the current global optimal extreme value meets the convergence condition.
In the embodiment of the present application, the training model is a BP neural network model.
And step S35, iteratively updating the global optimal extreme value.
And if the convergence condition is not met, carrying out the updating of the individual optimal particles, the updating of the global optimal particles and the self-variation operation of the global optimal again, and then updating the speed and the position of the particles in the population to update the global optimal extreme value until the convergence condition is met.
And step 36, taking the global optimal extreme value as an initial weight threshold of the training model.
If the current global optimal extreme value meets the convergence condition, the current global optimal extreme value can be used as an initial weight threshold value of the training model.
And step S23, based on the optimized initial weight threshold, training by taking the preprocessed historical alarm data as the input of the training model and taking the fault reason corresponding to the historical alarm data as the output of the training model to obtain a fault reason positioning model.
After the initial weight threshold of the training model is determined, the global optimal extreme value can be used as the initial weight threshold of the training model, the preprocessed historical alarm data is used as the input of the training model, and the fault reason corresponding to the historical alarm data is used as the output of the training model for training, so that the fault reason positioning model is obtained.
The fault reason corresponding to the historical alarm data can be derived from the real reason returned by the maintenance personnel.
The training process of the model is the prior art and is not described herein again.
And step S24, positioning the fault reason based on the fault reason positioning model.
Referring to fig. 4, the fault cause location based on the fault cause location model may include the following steps:
step S241, receiving the current alarm data reported by the fault management system.
Step S242, pre-process the current alarm data.
The preprocessing of the current alarm data includes denoising, data completion and normalization processing of the current alarm data, and the specific process is not described herein again.
Step S243, inputting the preprocessed current alarm data into a pre-trained fault cause positioning model to obtain a fault cause corresponding to the current alarm data.
According to the fault cause positioning method provided by the embodiment of the application, because the fault cause positioning model is obtained by training by taking historical alarm data reported by a fault management system as input of the training model and taking fault causes corresponding to the historical alarm data as output of the training model under the condition that the initial weight threshold of the training model is optimized through the PSO algorithm until the algorithm converges, the fault causes of large-area faults can be quickly positioned through the fault cause positioning model, so that the maintenance processing process is supported, the fault handling efficiency is improved, the user experience is improved, and the loss caused by the large-area faults is reduced. Meanwhile, since global optimal self-mutation operation is additionally performed after the standard PSO algorithm is performed, and mutation is performed based on Gaussian distribution random numbers and normal distribution random numbers, a more complex mutation situation is achieved, the diversity of mutation directions is enriched, the PSO can converge to a better global optimal extreme value, the fault cause can be accurately positioned by the trained fault cause positioning model, and the accuracy of fault cause positioning is improved. In addition, the fault cause can be accurately positioned without skillful mastering of full professional business logic, the method is convenient to use, is beneficial to popularization of the technology, does not need long-lasting large-scale resource clearing and hardness-attacking war, eliminates the dependence on resource barriers, saves the resource improvement cost and improves the economic benefit.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. Referring to fig. 5, at a hardware level, the electronic device includes a processor, and optionally further includes an internal bus, a network interface, and a memory. The Memory may include a Memory, such as a Random-Access Memory (RAM), and may further include a non-volatile Memory, such as at least 1 disk Memory. Of course, the electronic device may also include hardware required for other services.
The processor, the network interface, and the memory may be connected to each other via an internal bus, which may be an ISA (Industry standard architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry standard architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 5, but this does not indicate only one bus or one type of bus.
And the memory is used for storing programs. In particular, the program may include program code comprising computer operating instructions. The memory may include both memory and non-volatile storage and provides instructions and data to the processor.
The processor reads the corresponding computer program from the nonvolatile memory into the memory and then runs the computer program to form the fault cause positioning device on the logic level. The processor is used for executing the program stored in the memory and is specifically used for executing the following operations:
preprocessing historical alarm data reported by a fault management system;
optimizing an initial weight threshold value of a training model through a PSO algorithm until a target function algorithm corresponding to the training model converges;
based on the optimized initial weight threshold, training by taking the preprocessed historical alarm data as the input of the training model and the fault reason corresponding to the historical alarm data as the output of the training model to obtain a fault reason positioning model;
and carrying out fault reason positioning based on the fault reason positioning model.
The method performed by the fault cause locating device disclosed in the embodiments shown in fig. 2 to 4 of the present application may be applied to or implemented by a processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software. The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor.
The electronic device may also execute the method shown in fig. 2-4, and implement the functions of the fault cause positioning apparatus in the embodiments shown in fig. 2-4, which are not described herein again in this application.
Of course, besides the software implementation, the electronic device of the present application does not exclude other implementations, such as a logic device or a combination of software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or a logic device.
The electronic equipment provided by the embodiment of the application can be used for optimizing the initial weight threshold of the training model through the PSO algorithm to the condition of algorithm convergence, the historical alarm data reported by the fault management system is used as the input of the training model, and the fault reason corresponding to the historical alarm data is used as the output of the training model to be trained to obtain the fault reason positioning model, so that the fault reason of a large-area fault can be quickly positioned through the fault reason positioning model, the maintenance processing process is supported, the fault handling efficiency is improved, the user experience is improved, and the loss caused by the large-area fault is reduced. Meanwhile, global optimal self-mutation operation can be added after a standard PSO algorithm is carried out, mutation is carried out based on Gaussian distribution random numbers and normal distribution random numbers, a complex mutation situation is achieved, diversity of mutation directions is enriched, PSO can converge to a better global optimal extreme value, fault causes can be accurately positioned by a trained fault cause positioning model, and accuracy of fault cause positioning is improved. In addition, the fault cause can be accurately positioned without skillful mastering of full professional business logic, the method is convenient to use, is beneficial to popularization of the technology, does not need long-lasting large-scale resource clearing and hardness-attacking war, eliminates the dependence on resource barriers, saves the resource improvement cost and improves the economic benefit.
Embodiments of the present application also provide a computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a portable electronic device comprising a plurality of application programs, enable the portable electronic device to perform the method of the embodiments shown in fig. 2-4, and in particular to perform the following operations:
preprocessing historical alarm data reported by a fault management system;
optimizing an initial weight threshold value of a training model through a PSO algorithm until a target function algorithm corresponding to the training model converges;
based on the optimized initial weight threshold, training by taking the preprocessed historical alarm data as the input of the training model and taking the fault reason corresponding to the historical alarm data as the output of the training model to obtain a fault reason positioning model;
and carrying out fault reason positioning based on the fault reason positioning model.
Fig. 6 is a schematic structural diagram of a fault cause locating device 600 according to an embodiment of the present application. Referring to fig. 6, in a software implementation, the fault cause locating device 600 may include:
the preprocessing module 601 is configured to preprocess historical alarm data reported by the fault management system;
an optimizing module 602 configured to optimize an initial weight threshold of a training model through a PSO algorithm until an objective function algorithm corresponding to the training model converges;
a training module 603 configured to train, based on the optimized initial weight threshold, the preprocessed historical alarm data as an input of the training model and a fault cause corresponding to the historical alarm data as an output of the training model, so as to obtain a fault cause positioning model;
a positioning module 604 configured to perform fault cause positioning based on the fault cause positioning model.
The fault cause positioning device 600 provided by the embodiment of the application can optimize the initial weight threshold of the training model through the PSO algorithm to the condition of algorithm convergence, train the historical alarm data reported by the fault management system as the input of the training model, and train the fault cause corresponding to the historical alarm data as the output of the training model to obtain the fault cause positioning model, so that the fault cause of the large-area fault can be quickly positioned through the fault cause positioning model, so as to support the maintenance processing process, improve the fault handling efficiency, improve the user experience, and reduce the loss caused by the large-area fault. Meanwhile, as the global optimal self-variation operation is additionally arranged after the standard PSO algorithm is carried out, and the variation is carried out based on the Gaussian distribution random number and the normal distribution random number, the method has a more complex variation situation, enriches the diversity of variation directions, enables the PSO to converge to a better global optimal extreme value, enables the trained fault cause positioning model to accurately position the fault cause, and improves the accuracy of fault cause positioning. In addition, the fault cause can be accurately positioned without skillful mastering of full professional business logic, the method is convenient to use, is beneficial to popularization of the technology, does not need long-lasting large-scale resource clearing and hardness-attacking war, eliminates the dependence on resource barriers, saves the resource improvement cost and improves the economic benefit.
In short, the above description is only a preferred embodiment of the present application, and is not intended to limit the scope of the present application. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
Claims (8)
1. A fault cause positioning method is characterized by comprising the following steps:
preprocessing historical alarm data reported by a fault management system;
optimizing an initial weight threshold value of a training model through a PSO algorithm until a target function algorithm corresponding to the training model converges;
based on the optimized initial weight threshold, training by taking the preprocessed historical alarm data as the input of the training model and taking the fault reason corresponding to the historical alarm data as the output of the training model to obtain a fault reason positioning model;
carrying out fault cause positioning based on the fault cause positioning model;
optimizing an initial weight threshold of a training model through a PSO algorithm until an objective function algorithm corresponding to the training model converges comprises:
after population initialization, fitness calculation, individual optimal particle updating and global optimal particle updating of a PSO algorithm, calculating the variation probability of global optimal particles in a population;
performing mutation on the global optimal extreme value based on the mutation probability of the global optimal particle;
updating the speed and position of the particles in the population to find the global optimal extremum;
when the global optimal extreme value is used as an initial weight threshold value of the training model and an objective function algorithm corresponding to the training model converges or the iteration number of the PSO algorithm reaches a preset number, the global optimal extreme value is used as the initial weight threshold value of the training model;
the calculation formula for calculating the variation probability of the globally optimal particles in the population is
Wherein k is the number of iterations, p is the expansion constant that changes the rate of change of the exponential curve,
is the population fitness variance, p, of the kth generation max Is the maximum value of the probability of variation, p min Is the minimum of the variation probability;
the calculation formula for performing the variation on the global optimum extreme value is
Wherein, λ is a random number, η is a random number satisfying a standard normal distribution, and Gaussian (σ) is a random number obeying the Gaussian distribution and having a standard deviation of σ;
the calculation formula for updating the speed and the position of the particles in the population is
And X k+1 ＝X k +V k+1 Where ω is the inertial weight, r 1 And r 2 Is a random number distributed over the interval 0 to 1, k is the current iteration number,
for the individual optimal particle position at the kth iteration,
for the global optimal particle position at the kth iteration, c 1 And c 2 Is a constant number, V k Is the speed at the k-th iteration, V k+1 Speed at the k +1 th iteration, X k Is the particle position at the kth iteration, X k +1 The particle position at the (k + 1) th iteration.
2. The method of claim 1, wherein the locating the fault cause based on the fault cause location model comprises:
receiving current alarm data reported by the fault management system;
preprocessing the current alarm data;
inputting the preprocessed current alarm data into the pre-trained fault cause positioning model to obtain a fault cause corresponding to the current alarm data.
3. The method of claim 1, wherein the preprocessing the historical alarm data reported by the fault management system comprises:
and denoising, data completion and normalization processing are carried out on the historical alarm data to obtain a multi-dimensional vector corresponding to the historical alarm data.
4. The method of claim 1, wherein the historical alarm data comprises at least one of transmission alarm data, dynamic loop monitoring alarm data, wireless alarm data, and home broadband alarm data.
5. The method of claim 1, wherein the training model is a BP neural network model.
6. A fault cause locating device, comprising:
the system comprises a preprocessing module, a fault management module and a fault detection module, wherein the preprocessing module is configured to preprocess historical alarm data reported by a fault management system;
the optimization module is configured to optimize an initial weight threshold of a training model through a PSO algorithm until an objective function algorithm corresponding to the training model converges;
the training module is configured to train the preprocessed historical alarm data as the input of the training model and the fault reason corresponding to the historical alarm data as the output of the training model based on the optimized initial weight threshold value to obtain a fault reason positioning model;
a positioning module configured to perform fault cause positioning based on the fault cause positioning model;
the optimization module is further configured to calculate the variation probability of the globally optimal particles in the population after population initialization, fitness calculation, individual optimal particle update and globally optimal particle update of a PSO algorithm; is further configured to perform a mutation on the globally optimal extremum based on the mutation probability of the globally optimal particle; further configured to update the velocity and position of the particles in the population to find the global optimum extremum; is further configured to perform a training process when the global optimum extreme value is used as the training modelWhen the initial weight threshold value is reached and the objective function algorithm corresponding to the training model is converged or the iteration times of the PSO algorithm reach the preset times, taking the global optimal extreme value as the initial weight threshold value of the training model; the calculation formula for calculating the variation probability of the globally optimal particles in the population is
Wherein k is the number of iterations, p is the expansion constant that changes the rate of change of the exponential curve,
is the population fitness variance, p, of the k-th generation max Is the maximum value of the probability of variation, p min The minimum value of the variation probability is calculated as the formula for performing variation on the global optimum extreme value
Wherein λ is a random number, η is a random number satisfying a standard normal distribution, Gaussian (σ) is a random number obeying a Gaussian distribution and having a standard deviation of σ, and the calculation formula for updating the velocity and the position of the particles in the population is
And X k+1 ＝X k +V k+1 Where ω is the inertial weight, r 1 And r 2 Is a random number distributed over the interval 0 to 1, k is the current iteration number,
for the individual optimal particle position at the kth iteration,
for the global optimal particle position at the kth iteration, c 1 And c 2 Is a constant number, V k Is the speed at the k-th iteration, V k+1 Speed at the k +1 th iteration, X k As grains at the k-th iterationSub-position, X k+1 The particle position at the (k + 1) th iteration.
7. An electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, the computer program, when executed by the processor, performing the method steps of any of claims 1-5.
8. A computer-readable storage medium, characterized in that a computer program is stored in the storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911392356.3A CN113132125B (en) | 2019-12-30 | 2019-12-30 | Fault cause positioning method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911392356.3A CN113132125B (en) | 2019-12-30 | 2019-12-30 | Fault cause positioning method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113132125A CN113132125A (en) | 2021-07-16 |
| CN113132125B true CN113132125B (en) | 2022-08-23 |
Family
ID=76768944
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911392356.3A Active CN113132125B (en) | 2019-12-30 | 2019-12-30 | Fault cause positioning method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113132125B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116298686A (en) * | 2023-03-16 | 2023-06-23 | 广东电网有限责任公司广州供电局 | Fault positioning method, device, equipment and medium applied to power distribution network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104764980A (en) * | 2015-04-22 | 2015-07-08 | 福州大学 | Positioning method for distribution circuit fault section based on BPSO and GA |
| CN106229964A (en) * | 2016-07-22 | 2016-12-14 | 南京工程学院 | A kind of based on the electrical power distribution network fault location method improving binary particle swarm algorithm |
| CN107843812A (en) * | 2017-11-02 | 2018-03-27 | 海南电网有限责任公司琼海供电局 | A kind of electrical power distribution network fault location method and device |
| CN108334059A (en) * | 2018-02-26 | 2018-07-27 | 上海理工大学 | Based on particle cluster algorithm Optimized BP Neural Network Symbolic fault diagnosis method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104239694B (en) * | 2014-08-28 | 2016-11-23 | 北京交通大学 | The failure predication of a kind of municipal rail train bogie and condition maintenarnce method |
-
2019
- 2019-12-30 CN CN201911392356.3A patent/CN113132125B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104764980A (en) * | 2015-04-22 | 2015-07-08 | 福州大学 | Positioning method for distribution circuit fault section based on BPSO and GA |
| CN106229964A (en) * | 2016-07-22 | 2016-12-14 | 南京工程学院 | A kind of based on the electrical power distribution network fault location method improving binary particle swarm algorithm |
| CN107843812A (en) * | 2017-11-02 | 2018-03-27 | 海南电网有限责任公司琼海供电局 | A kind of electrical power distribution network fault location method and device |
| CN108334059A (en) * | 2018-02-26 | 2018-07-27 | 上海理工大学 | Based on particle cluster algorithm Optimized BP Neural Network Symbolic fault diagnosis method |
Non-Patent Citations (2)
| Title |
|---|
| "一种改进的自适应变异的粒子群优化算法";贾松浩;《实验室研究与探索》;20140415;第15-16页 * |
| Xin Wang ; He-nan Sun ; Dan-lu Wang."Fault diagnosis of cascaded inverter based on PSO-BP neural networks".《 Proceedings of the 33rd Chinese Control Conference》.2014, * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113132125A (en) | 2021-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10489363B2 (en) | Distributed FP-growth with node table for large-scale association rule mining | |
| EP3654610A1 (en) | Graphical structure model-based method for prevention and control of abnormal accounts, and device and equipment | |
| US20120304288A1 (en) | Modeling and Outlier Detection in Threat Management System Data | |
| CN111600746B (en) | Network fault positioning method, device and equipment | |
| CN110555172B (en) | User relationship mining method and device, electronic equipment and storage medium | |
| CN104995870A (en) | Multi-objective server placement determination | |
| CN104486129B (en) | The method and system of application service quality are ensured under distributed environment | |
| US20180285371A1 (en) | Method and apparatus for automatically discovering gas station poi, storage medium and device | |
| CN110022227A (en) | Method for processing business, device, equipment and storage medium | |
| CN113132125B (en) | Fault cause positioning method and device, electronic equipment and storage medium | |
| CN112491900B (en) | Abnormal node identification method, device, equipment and medium | |
| US20210160142A1 (en) | Generalized correlation of network resources and associated data records in dynamic network environments | |
| CN111193627B (en) | Information processing method, device, equipment and storage medium | |
| WO2023146690A2 (en) | Quantum computing monitoring system | |
| CN115373888A (en) | Fault positioning method and device, electronic equipment and storage medium | |
| CN112953768B (en) | Three-layer network change scheme verification method and system | |
| CN113591999B (en) | End-edge cloud federal learning model training system and method | |
| CN116028832A (en) | Sample clustering processing method and device, storage medium and electronic equipment | |
| CN115390113A (en) | BD3 high-precision positioning method and system for power system | |
| CN111582456B (en) | Method, apparatus, device and medium for generating network model information | |
| CN113497721B (en) | Network fault positioning method and device | |
| CN111782435B (en) | Method and system for recovering and processing cascade exception of video monitoring management platform | |
| CN113076451B (en) | Abnormal behavior identification and risk model library establishment method and device and electronic equipment | |
| CN107544248A (en) | A kind of task optimization method and apparatus in mobile robot | |
| CN113132136B (en) | Satisfaction degree prediction model establishment method, satisfaction degree prediction device and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |