Alarm prediction method for base station, electronic device, and computer readable medium

Abstract

The present application provides an alarm prediction method for a base station, an electronic device, and a computer readable medium. The alarm prediction method for a base station comprises: acquiring station information of a base station, wherein the station information comprises at least one of station basic information and device information of a device in the base station; and inputting the station information into a pre-trained alarm prediction model to obtain predicted alarm information, wherein the predicted alarm information comprises at least one of the following: the number of alarms, alarm type information, alarm start time information, alarm end time information, alarm level information and alarm duration information.

Description

Base station alarm prediction method, electronic device, and computer-readable medium

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This patent application claims priority to Chinese patent application 202311291254.9 filed with the State Intellectual Property Office of China on September 28, 2023, and the disclosure of this Chinese patent application is incorporated herein by reference in its entirety. Technical Field

[0003] The present application relates to the field of communication technology, and in particular to a base station alarm prediction method, an electronic device, and a computer-readable medium. Background Art

[0004] With the rapid development of mobile internet and the increasing popularity of mobile communication technology, more and more people are using mobile devices for communication, and they have higher and higher requirements for the quality of communication services, requiring wider and more accurate coverage and signal strength. Therefore, the number and quality of base stations need to be more guaranteed.

[0005] Summary of the Invention

[0006] Embodiments of the present application provide a base station alarm prediction method, an electronic device, and a computer-readable medium.

[0007] In a first aspect, an embodiment of the present application provides a base station alarm prediction method, comprising: obtaining site information of a base station, wherein the site information comprises: at least one of basic site information and device information of a device in the base station; and inputting the site information into a pre-trained alarm prediction model to obtain predicted alarm information, wherein the predicted alarm information comprises at least one of the following: number of alarms, alarm type information, alarm start time information, alarm end time information, alarm level information, and alarm duration information.

[0008] In a second aspect, an embodiment of the present application provides an electronic device, comprising: at least one processor; a memory, wherein at least one program is stored in the memory, and when the at least one program is executed by the at least one processor, the above-mentioned base station alarm prediction method is implemented.

[0009] In a third aspect, an embodiment of the present application provides a computer-readable medium having a computer program stored thereon, and when the computer program is executed by a processor, the above-mentioned base station alarm prediction method is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] FIG1 is a flow chart of a base station alarm prediction method according to an embodiment of the present application;

[0011] FIG2 is a flow chart of a base station alarm prediction method according to an example of the present application;

[0012] FIG3 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0013] In order to enable those skilled in the art to better understand the technical solution of the present application, the alarm prediction method, electronic device, and computer-readable medium provided by the present application are described in detail below with reference to the accompanying drawings.

[0014] Example embodiments will be described more fully hereinafter with reference to the accompanying drawings, but the example embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application will be thorough and complete and will fully convey the scope of this application to those skilled in the art.

[0015] In the absence of conflict, the various embodiments of the present application and the various features therein may be combined with each other.

[0016] As used herein, the term "and / or" includes any and all combinations of at least one of the associated listed items.

[0017] The terms used herein are used only to describe specific embodiments and are not intended to limit this application. As used herein, the singular forms "a," "an," and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, the presence of the features, wholes, steps, operations, elements, and / or components is specified, but the presence or addition of at least one other feature, whole, step, operation, element, component, and / or group thereof is not excluded.

[0018] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this application, and will not be interpreted as having an idealized or overly formal meaning unless expressly defined as such herein.

[0019] Base station maintenance often has the following difficulties:

[0020] 1. Base stations require long-term operation and regular maintenance and monitoring;

[0021] 2. Base stations are usually distributed in different geographical locations, making maintenance difficult;

[0022] 3. Base stations usually have many sites and devices, resulting in a large number of alarms with complex types, and low processing efficiency.

[0023] FIG1 is a flowchart of a base station alarm prediction method according to an embodiment of the present application.

[0024] 1 , a base station alarm prediction method according to an embodiment of the present application includes steps 100 and 101 .

[0025] In step 100, the site information of the base station is obtained. The site information includes at least one of basic site information and device information of a device in the base station.

[0026] In some exemplary embodiments, the basic site information includes at least one of the following: site location information, site environment information, site operation status information, site opening time information, etc.

[0027] In some exemplary embodiments, the site environment information includes at least one of the following: site temperature information, site humidity information, etc.

[0028] In some exemplary embodiments, the device information includes at least one of the following: device operating status information, device type information, device manufacturer information, device brand information, device deployment time information, device factory time information, etc.

[0029] In some exemplary embodiments, the acquired site information may be site information within a unit of time, for example, acquiring site information corresponding to a day in units of days.

[0030] In step 101, site information is input into a pre-trained alarm prediction model to obtain predicted alarm information. The predicted alarm information includes at least one of the following: alarm quantity, alarm type, alarm start time, alarm end time, alarm level, and alarm duration.

[0031] In this application, site information and corresponding predicted alarm information vary for different application scenarios. For example, in a lightning protection site inspection scenario, site information includes at least one of: site environment information, site operating status information, equipment operating status information, equipment type information, equipment manufacturer information, equipment brand information, equipment deployment time information, and equipment factory time information. The corresponding predicted alarm information includes: number of alarms, alarm start time information, and alarm end time information.

[0032] For example, in the site opening equipment matching scenario, the site information includes at least one of: site location information, site environment information, site operation status information, site opening time information, equipment type information, equipment manufacturer information, equipment brand information, equipment deployment time information, equipment factory time information, etc. The corresponding predicted alarm information includes: number of alarms, alarm type information, alarm start time information, and alarm end time information.

[0033] In some exemplary embodiments, the alarm prediction model may be any one of a long short-term memory network (LSTM), a random forest algorithm, a support vector machine (SVM) algorithm, and the like.

[0034] In some exemplary embodiments, before obtaining the site information of the base station, the method also includes: obtaining first historical site information and corresponding first historical alarm information; using the first historical site information as the input of the alarm prediction model and using the first historical alarm information as the output of the alarm prediction model to perform model training on the alarm prediction model to obtain a pre-trained alarm prediction model.

[0035] In some exemplary embodiments, the first historical site information includes at least one of first historical site basic information and first historical device information of a device in the first base station.

[0036] In some exemplary embodiments, the first historical site basic information includes at least one of the following: site historical location information, site historical environment information, site historical operation status information, site opening time information, etc.

[0037] In some exemplary embodiments, the site historical environmental information includes at least one of the following: site historical temperature information, site historical humidity information, etc.

[0038] In some exemplary embodiments, the first historical device information includes at least one of the following: device historical operating status information, device type information, device manufacturer information, device brand information, device deployment time information, device factory time information, etc.

[0039] In some exemplary embodiments, the first historical alarm information includes at least one of the following: historical alarm quantity, historical alarm type information, historical alarm start time information, historical alarm end time information, historical alarm level information, historical alarm duration information, etc.

[0040] In some exemplary embodiments, obtaining the first historical site information and the corresponding first historical alarm information includes: obtaining historical alarm record information; and determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information.

[0041] In some exemplary embodiments, as shown in Table 1, the historical alarm record information includes: second historical alarm information, second historical site basic information, and second historical device information of a device in the second base station.

[0042] In some exemplary embodiments, the second historical alarm information includes at least one of the following: a unique identifier of the historical alarm, the number of historical alarms, the start time information of the historical alarm, the end time information of the historical alarm, the duration information of the historical alarm, the storage time information of the historical alarm, information representing the cause of the historical alarm, historical alarm level information, historical alarm type information, etc.

[0043] In some exemplary embodiments, the second historical site basic information includes at least one of the following: site historical location information, site historical environment information (including site historical temperature information and site historical humidity information), site historical operating status information, site opening time information, etc.

[0044] In some exemplary embodiments, the second historical device information includes at least one of the following: device historical operating status information, device type information, device manufacturer information, device brand information, device deployment time information, device factory time information, etc.

[0045] Table 1

[0046] In some exemplary embodiments, the historical alarm record information refers to formatted data based on collected measurement point data.

[0047] In some exemplary embodiments, after acquiring the historical alarm record information and before determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information, the method further includes: performing data cleaning processing on the historical alarm record information.

[0048] Determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information includes: determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information after data cleaning processing.

[0049] In some exemplary embodiments, data cleaning processing of historical alarm record information includes at least one of the following: processing missing values ​​of historical alarm record information; eliminating historical alarm record information with abnormal values; text conversion of historical alarm record information; and scene filtering of historical alarm record information.

[0050] In the embodiment of the present application, after data cleaning processing, many historical alarm record information containing missing values ​​and abnormal values ​​are filtered out, and the presentation method of text data is optimized.

[0051] In some exemplary embodiments, processing missing values ​​in historical alarm records involves deleting historical alarm records with missing key field values ​​when the number of historical alarm records is greater than or equal to a preset threshold; and backfilling missing key field values ​​when the number of historical alarm records is less than the preset threshold. Different application scenarios correspond to different key fields. For example, in a lightning protection site inspection scenario, key fields include: historical alarm type information, site historical location information, historical alarm start time information, and historical alarm end time information.

[0052] In some exemplary embodiments, removing the historical alarm record information containing abnormal values ​​may refer to removing the entire historical alarm record information containing abnormal values, or removing the fields containing abnormal values ​​in the historical alarm record information.

[0053] In some exemplary embodiments, performing text conversion on the historical alarm record information refers to converting field values ​​of the historical alarm record information whose data type is text into numerical values.

[0054] In some exemplary embodiments, one-hot encoding can be used to convert text into numerical values. For example, if the alarm level information includes four types: warning, minor, major, and severe, one-hot encoding can be used to convert these four alarm levels into discrete numerical values, such as 00 for warning, 01 for minor, 10 for major, and 11 for severe.

[0055] In some exemplary embodiments, scenario filtering of historical alarm records refers to filtering out historical alarm records that do not pertain to the current prediction scenario. For example, in a lightning protection site inspection scenario, historical alarm records with non-AC lightning protection alarm types do not need to be included in model training and can be deleted.

[0056] In some exemplary embodiments, after acquiring the historical alarm record information and before determining the first historical site information and the corresponding first historical alarm information based on the historical alarm record information, the method further includes: performing dimensionality reduction processing on the historical alarm record information.

[0057] Determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information includes: determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information after dimensionality reduction processing.

[0058] In some exemplary embodiments, after performing data cleaning processing on the historical alarm record information, and before determining the first historical site information and the corresponding first historical alarm information based on the historical alarm record information after the data cleaning processing, the method also includes: performing dimensionality reduction processing on the historical alarm record information after the data cleaning processing.

[0059] Determining the first historical site information and the corresponding first historical alarm information based on the historical alarm record information after data cleaning processing includes: determining the first historical site information and the corresponding first historical alarm information based on the historical alarm record information after dimensionality reduction processing.

[0060] In some exemplary embodiments, a principal component analysis (PCA) method or a K-means clustering method may be used to perform dimensionality reduction processing on the historical alarm record information after data cleaning processing.

[0061] In some exemplary embodiments, the principal component analysis method is an unsupervised linear dimensionality reduction algorithm that achieves data dimensionality reduction by mapping high-dimensional data into a low-dimensional space and retaining the eigenvector with the maximum variance in the historical alarm record information as the new main feature. The underlying principle is to use the eigenvalues ​​and eigenvectors of the covariance matrix to perform principal component analysis.

[0062] The principal component analysis method mainly includes: calculating the covariance between any two features, retaining the two features if the covariance between the two features is greater than or equal to a preset threshold; deleting the two features if the covariance between the two features is less than the preset threshold.

[0063] In some exemplary embodiments, assuming feature x and feature y, the mean of feature x is The mean of feature y is Assuming the number of samples is m, the covariance between feature x and feature y is:

[0064] In some exemplary embodiments, the K-means clustering method divides data points into k clusters, maximizing the similarity between data points within the same cluster and minimizing the similarity between different clusters. The underlying principle of the K-means clustering method is to calculate the Euclidean distance or other distance between data points, assign the closest data points to the same cluster, and iteratively optimize until a specified convergence condition is met. The K-means clustering method can find similarities between data and classify the data.

[0065] In some exemplary embodiments, different dimensionality reduction methods can be used in different application scenarios. For example, when predicting the time period distribution of large amounts of data for future alerts, principal component analysis can be used for dimensionality reduction, retaining the main component features of the data for prediction. When single-point trend prediction of a feature of an alert is required, K-means clustering is more effective because it can aggregate the features with the highest similarity for model training.

[0066] In some exemplary embodiments, after model training is completed, the model's predictive performance can be evaluated based on a test dataset or a validation dataset, and evaluation metrics can be used to assess the model's performance. If the evaluation results do not meet expectations, the model is readjusted and optimized, and the model training is repeated until the expected results are achieved.

[0067] In some exemplary embodiments, the evaluation indicator may be at least one of accuracy, precision, recall, etc.

[0068] In some exemplary embodiments, trained models support both local and remote deployment. Local deployment uses a site's historical alarm record information to train an alarm prediction model for local application, while remote deployment uses a site's historical alarm record information to load the trained alarm prediction model onto other sites. For sites with limited historical alarm record information at launch, remote deployment is recommended. Predictions can be performed using an alarm prediction model trained at another site. Once the site's local data volume reaches a certain level, the model can be trained using its own local data.

[0069] In some exemplary embodiments, after inputting the site information into a pre-trained alarm prediction model to obtain predicted alarm information, the method further includes: determining a change trend of the alarm over time based on the predicted alarm information.

[0070] In some exemplary embodiments, site information corresponding to multiple different time points can be obtained, and the site information corresponding to each time point can be input into the alarm prediction model to obtain the predicted alarm information corresponding to the time point. The predicted alarm information corresponding to multiple time points is the changing trend of the alarm over time.

[0071] In some exemplary embodiments, after inputting the site information into a pre-trained alarm prediction model to obtain predicted alarm information, the method further includes: determining the geographical location distribution of the alarms based on the predicted alarm information.

[0072] In some exemplary embodiments, the site information of base stations with multiple different geographic location information can be obtained, and the site information corresponding to each geographic location information can be input into the alarm prediction model to obtain the predicted alarm information corresponding to the geographic location information. The predicted alarm information corresponding to multiple geographic location information is the geographic location distribution of the alarm.

[0073] In some exemplary embodiments, the site information of the base stations where devices with multiple different geographic location information are located can be obtained, and the site information corresponding to each geographic location information can be input into the alarm prediction model to obtain the predicted alarm information corresponding to the geographic location information. The predicted alarm information corresponding to multiple geographic location information is the geographic location distribution of the alarm.

[0074] In some exemplary embodiments, after inputting the site information into a pre-trained alarm prediction model to obtain predicted alarm information, the method further includes: determining the distribution of alarm equipment types based on the predicted alarm information.

[0075] In some exemplary embodiments, site information can be obtained when using multiple different types of equipment, and the site information corresponding to each type of equipment can be input into the alarm prediction model to obtain the predicted alarm information corresponding to the equipment type. The predicted alarm information corresponding to multiple device terminals is the distribution of equipment types of alarms.

[0076] In some exemplary embodiments, a statistical line graph may be used to display the changing trend of alarms over time.

[0077] In some exemplary embodiments, a geographical point map of location distribution may be used to display the geographical location distribution of alarms.

[0078] In some exemplary embodiments, at least one of the temporal trend of alarms, the geographical distribution of alarms, and the distribution of equipment types of alarms can be used to allow operation and maintenance personnel to make corresponding maintenance preparations, conduct problem troubleshooting for sites where alarms may occur instead of conducting full site inspections every time, increase investment in time periods and regions where alarms occur more frequently, and improve operation and maintenance efficiency. For example, sites are often prone to lightning strikes, and thunderstorms in each region are often difficult to predict. The alarm prediction method of the embodiment of the present application is used to train a model based on historical alarm record information of AC lightning protection alarms in the past, and then predict and warn sites where lightning strikes may occur in the future, reminding operation and maintenance personnel to consolidate and strengthen the lightning protection system of the site.

[0079] For example, core components like power supplies, rectifiers, and transformers at a site often age over time. If repairs are delayed until a problem occurs, base station service quality may degrade or even be interrupted. Equipment problems are often preceded by specific alarms. Manually analyzing these patterns can be time-consuming, labor-intensive, and error-prone. However, model training and model-based alarm prediction can guide operations and maintenance personnel in troubleshooting equipment, significantly reducing their scope and improving efficiency.

[0080] Another example is site equipment, such as batteries, rectifiers, transformers, and communication modules. Typically, sites feature a mix of various manufacturers and brands. When deploying a site, equipment selection is a complex task. Choosing the optimal equipment combination from the vast array of available equipment is a challenging task. Different equipment has varying qualities, and one indicator of quality is the number of alarms that occur. Using model-based alarm prediction, we can assess the site quality of the equipment being deployed. We can then select deployment equipment based on metrics such as the predicted alarm count and frequency, resulting in a strategically optimized deployment strategy to ensure future site operation quality.

[0081] In some exemplary embodiments, after training the alarm prediction model using the first historical site information as the input of the alarm prediction model and using the first historical alarm information as the output of the alarm prediction model to obtain a pre-trained alarm prediction model, the method further includes: adding the site information and the corresponding actual alarm information to the training data set to retrain the alarm prediction model.

[0082] In some exemplary embodiments, adding site information and corresponding actual alarm information to a training data set to retrain the alarm prediction model means using the site information as input to the alarm prediction model and using the corresponding actual alarm information as output to retrain the model.

[0083] In some exemplary embodiments, the actual alarm information includes at least one of the following: an actual alarm unique identifier, the actual number of alarms, actual alarm start time information, actual alarm end time information, actual alarm duration information, actual alarm storage time information, information representing the cause of the actual alarm, actual alarm level information, actual alarm type information, etc.

[0084] In some exemplary embodiments, the training data set may further include first historical site information and corresponding first historical alarm information.

[0085] In some exemplary embodiments, a feedback alignment method can be used to retrain the model. This method uses a random weight matrix to implement a backpropagation algorithm and updates the model parameters using the backpropagated error signal. This reduces the amount of computation and improves model training efficiency.

[0086] The alarm prediction method provided in the embodiment of the present application performs alarm prediction based on base station site information. By predicting the alarm information, the site monitoring system can be guided to take preventive measures and reduce the failure rate of the base station. The base stations for regular inspections can be preliminarily screened to reduce operation and maintenance costs. The matching of site construction equipment can be guided to ensure high-quality service of the base station.

[0087] In order to more completely present the alarm prediction method of the embodiment of the present application, an example is listed below for illustration. The listed example is not used to limit the protection scope of the embodiment of the present application.

[0088] FIG2 is a flow chart of a base station alarm prediction method according to an example of the present application. As shown in FIG2 , the base station alarm prediction method includes steps 200 to 208 .

[0089] In step 200, historical alarm record information is obtained. The historical alarm record information includes: second historical alarm information, second historical site basic information, and second historical device information of devices in the second base station.

[0090] In some exemplary embodiments, the second historical alarm information includes: a unique identifier of the historical alarm, the number of historical alarms, the start time information of the historical alarm, the end time information of the historical alarm, the duration information of the historical alarm, the storage time information of the historical alarm, information representing the cause of the historical alarm, historical alarm level information, historical alarm type information, etc.

[0091] In some exemplary embodiments, the second historical site basic information includes: site historical location information, site historical environment information, site historical operation status information, site opening time information, etc.

[0092] In some exemplary embodiments, the second historical device information includes: device historical operating status information, device type information, device manufacturer information, device brand information, device deployment time information, device factory delivery time information, etc.

[0093] In step 201, data cleaning is performed on historical alarm record information.

[0094] In some exemplary embodiments, data cleaning processing of historical alarm record information includes at least one of the following: processing missing values ​​of historical alarm record information; eliminating historical alarm record information with abnormal values; text conversion of historical alarm record information; and scene filtering of historical alarm record information.

[0095] In step 202, dimensionality reduction processing is performed on the historical alarm record information after data cleaning processing.

[0096] In step 203, first historical site information and corresponding first historical alarm information are determined based on the historical alarm record information after dimensionality reduction processing. The first historical site information includes: first historical site basic information and first historical device information of devices in the first base station. The first historical site basic information includes: historical site location information, historical site environment information, historical site operating status information, and site start time information. The first historical device information includes: historical device operating status information, device type information, device manufacturer information, device brand information, device deployment time information, and device factory release time information. The first historical alarm information includes: historical alarm quantity, historical alarm type information, historical alarm start time information, historical alarm end time information, historical alarm level information, and historical alarm duration information.

[0097] In step 204, the first historical site information is used as input of the alarm prediction model, and the first historical alarm information is used as output of the alarm prediction model to perform model training on the alarm prediction model, so as to obtain a pre-trained alarm prediction model.

[0098] In step 205, the site information of the base station is obtained. The site information includes basic site information and device information of the devices in the base station. The basic site information includes: site location information, site environment information, site operation status information, and site commissioning time. The device information includes: device operation status information, device type information, device manufacturer information, device brand information, device deployment time information, and device factory time information.

[0099] In step 206, the site information is input into the pre-trained alarm prediction model to obtain predicted alarm information. The predicted alarm information includes: alarm quantity, alarm type information, alarm start time, alarm end time, alarm level information, and alarm duration information.

[0100] In step 207, the change trend of the alarm over time is determined based on the predicted alarm information, and the geographical location distribution of the alarm is determined based on the predicted alarm information.

[0101] In step 208, the site information and the corresponding actual alarm information are added to the training data set and the alarm prediction model is retrained using a feedback alignment method.

[0102] FIG3 is a block diagram of an electronic device according to an embodiment of the present application.

[0103] 3 , an electronic device according to an embodiment of the present application includes: at least one processor 301 ; a memory 302 , wherein the memory 302 stores at least one program, and when the at least one program is executed by the at least one processor 301 , any one of the above-mentioned base station alarm prediction methods is implemented.

[0104] In some exemplary embodiments, the electronic device further includes: one or more I / O interfaces 303 connected between the processor 301 and the memory 302 , and configured to implement information interaction between the processor 301 and the memory 302 .

[0105] The processor 301 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 302 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read-write interface) 303 is connected between the processor 301 and the memory 302, and can realize information exchange between the processor 301 and the memory 302, including but not limited to a data bus (Bus), etc.

[0106] In some embodiments, the processor 301 , the memory 302 , and the I / O interface 303 are connected to each other via a bus 304 , and further connected to other components of the computing device.

[0107] The present application also provides a computer-readable medium having a computer program stored thereon. When the computer program is executed by a processor, the computer program implements the above-mentioned base station alarm prediction method.

[0108] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As is well known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable, and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, as is well known to those skilled in the art, communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0109] Example embodiments have been disclosed herein, and although specific terms are employed, they are used and should be interpreted only in a general illustrative sense and not for purposes of limitation. In some instances, it will be apparent to those skilled in the art that, unless otherwise expressly indicated, features, characteristics, and / or elements described in conjunction with a particular embodiment may be used alone or in combination with features, characteristics, and / or elements described in conjunction with other embodiments. Therefore, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the scope of the present application as set forth in the appended claims.

Claims

1. A base station alarm prediction method, comprising: Acquire site information of a base station, wherein the site information includes: at least one of basic site information and device information of a device in the base station; and The site information is input into a pre-trained alarm prediction model to obtain predicted alarm information, wherein the predicted alarm information includes at least one of the following: alarm quantity, alarm type information, alarm start time information, alarm end time information, alarm level information, and alarm duration information.

2. The base station alarm prediction method according to claim 1, after inputting the site information into the pre-trained alarm prediction model to obtain the predicted alarm information, further comprising at least one of the following: Determine the change trend of the alarm over time according to the predicted alarm information; Determine the address and location distribution of the alarm according to the predicted alarm information; and The distribution of equipment types of alarms is determined based on the predicted alarm information.

3. The base station alarm prediction method according to claim 1, before obtaining the site information of the base station, further comprising: Obtaining first historical site information and corresponding first historical alarm information; The first historical site information is used as input of an alarm prediction model, the first historical alarm information is used as output of the alarm prediction model, and model training is performed on the alarm prediction model to obtain the pre-trained alarm prediction model.

4. The base station alarm prediction method according to claim 3, wherein: Acquiring the first historical site information and the corresponding first historical alarm information includes: Get historical alarm record information; The first historical site information and the corresponding first historical alarm information are determined according to the historical alarm record information.

5. The base station alarm prediction method according to claim 4, after acquiring the historical alarm record information and before determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information, further comprising: Performing data cleaning on the historical alarm record information, Among them, determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information includes: determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information after data cleaning processing.

6. The base station alarm prediction method according to claim 5, after performing data cleaning processing on the historical alarm record information and before determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information after the data cleaning processing, further comprising: Performing dimensionality reduction processing on the historical alarm record information after the data cleaning processing, Among them, determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information after the data cleaning processing includes: determining the first historical site information and the corresponding first historical alarm information according to the historical alarm record information after the dimensionality reduction processing.

7. The base station alarm prediction method according to claim 3, after taking the first historical site information as the input of the alarm prediction model and taking the first historical alarm information as the output of the alarm prediction model to perform model training on the alarm prediction model to obtain the pre-trained alarm prediction model, further comprising: The site information and the corresponding actual alarm information are added to the training data set to retrain the alarm prediction model.

8. The base station alarm prediction method according to any one of claims 1 to 7, wherein: The basic site information includes at least one of the following: site location information, site environment information, site operation status information, and site opening time information; The device information includes at least one of the following: device operation status information, device type Model information, equipment manufacturer information, equipment brand information, equipment deployment time information, and equipment factory time information.

9. An electronic device, comprising: at least one processor; A memory, wherein at least one program is stored in the memory, and when the at least one program is executed by the at least one processor, the base station alarm prediction method according to any one of claims 1 to 8 is implemented.

10. A computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the base station alarm prediction method according to any one of claims 1 to 8.

Images