An automatic fault prediction method and system for base station operation and maintenance

Abstract

This invention discloses an automatic fault prediction method and system for base station operation and maintenance, relating to the field of base station operation and maintenance technology. The method includes: collecting multiple signal features at multiple monitoring locations of the communication base station; analyzing the signal feature changes of the base station antenna to obtain multiple signal change parameters; performing base station antenna offset analysis based on the multiple signal change parameters to obtain multiple antenna offset parameters; performing offset reliability analysis to obtain multiple offset reliability levels; acquiring environmental parameters within the communication base station's environment; combining these with the multiple antenna offset parameters to perform base station antenna influence offset analysis to obtain multiple antenna influence offset parameters; and calculating the antenna fault scale based on the multiple offset reliability levels and the multiple antenna influence offset parameters, using this as the base station operation and maintenance fault prediction result. This invention solves the technical problem of poor fault prediction accuracy in existing technologies for base station operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of base station operation and maintenance technology, and specifically to an automatic fault prediction method and system for base station operation and maintenance. Background Technology

[0002] In wireless communication networks, communication base stations, as core infrastructure, are crucial for ensuring network coverage quality and user experience through the stable operation of their antennas. Current base station operation and maintenance fault monitoring technologies primarily rely on threshold alarm mechanisms for key performance indicators such as signal strength. However, these methods have significant limitations; they are reactive and only trigger alarms after performance has severely deteriorated or a fault has actually occurred, resulting in poor timeliness and susceptibility to misjudgments due to localized data anomalies or noise. Existing methods suffer from bottlenecks in the timeliness, accuracy, and robustness of base station fault prediction, making it difficult to meet the urgent needs of modern networks for high reliability and proactive operation and maintenance. Summary of the Invention

[0003] This application provides an automatic fault prediction method and system for base station operation and maintenance, which addresses the technical problems of poor accuracy and timeliness in fault prediction in existing base station operation and maintenance.

[0004] In view of the above problems, this application provides an automatic fault prediction method and system for base station operation and maintenance.

[0005] Firstly, this application provides an automatic fault prediction method for base station operation and maintenance, the method comprising:

[0006] Multiple signal features are collected at multiple monitoring locations of the communication base station, and the signal feature changes of the base station antenna are analyzed to obtain multiple signal change parameters.

[0007] Base station antenna offset analysis is performed based on multiple signal variation parameters to obtain multiple antenna offset parameters. Offset confidence analysis is then performed to obtain multiple offset confidence levels.

[0008] The environmental parameters within the environment where the communication base station is located are obtained, and combined with multiple antenna offset parameters, the base station antenna influence offset analysis is performed to obtain multiple antenna influence offset parameters.

[0009] Based on multiple offset confidence levels and combined with multiple antenna influence offset parameters, the antenna fault scale is calculated and used as the base station operation and maintenance fault prediction result.

[0010] Optionally, multiple signal features are collected at multiple monitoring locations of the communication base station to analyze the signal feature changes of the base station antenna and obtain multiple signal change parameters, including:

[0011] Multiple signal characteristics are collected at various monitoring locations of the communication base station, including signal quality.

[0012] Acquire multiple standard signal characteristics at multiple monitoring locations;

[0013] The variation amplitudes of multiple signal features and multiple standard signal features are calculated separately to obtain multiple signal variation parameters.

[0014] Optionally, base station antenna offset analysis is performed based on multiple signal variation parameters to obtain multiple antenna offset parameters, and offset confidence analysis is performed to obtain multiple offset confidence levels, including:

[0015] Each signal variation parameter is input into the antenna offset analyzer, and multiple antenna offset parameters are output. Each antenna offset parameter includes the offset direction and offset scale.

[0016] Calculate the similarity between each antenna offset parameter and the mean of multiple antenna offset parameters to obtain multiple first offset confidence scores;

[0017] Multiple monitoring distances between multiple monitoring locations and base station antennas are obtained, and multiple second offset confidence levels are calculated. The magnitude of the monitoring distance is negatively correlated with the magnitude of the second offset confidence level.

[0018] Multiple offset confidence levels are calculated based on multiple first offset confidence levels and multiple second offset confidence levels.

[0019] Optionally, the steps for obtaining the antenna offset analyzer include:

[0020] Based on the historical offset maintenance data of the base station antenna, multiple sets of sample signal change parameters are collected from multiple monitoring locations, and antenna offset parameters of the base station antenna under different signal change parameters are collected and labeled to obtain multiple sets of sample antenna offset parameters.

[0021] Based on machine learning, multiple antenna offset analysis paths with the same architecture are constructed.

[0022] Multiple antenna offset analysis paths are trained under supervision using multiple sets of sample signal variation parameters and multiple sets of sample antenna offset parameters. After convergence, they are integrated to obtain an antenna offset analyzer.

[0023] Optionally, environmental parameters within the environment where the communication base station is located are obtained, and combined with multiple antenna offset parameters, base station antenna influence offset analysis is performed to obtain multiple antenna influence offset parameters, including:

[0024] Obtain environmental parameters within the environment where the communication base station is located, including meteorological parameters;

[0025] The environmental parameters are combined with each antenna offset parameter and input into the antenna influence offset analyzer to obtain multiple antenna influence offset parameters.

[0026] Optionally, the construction steps of the antenna influence offset analyzer include:

[0027] Based on base station antenna operation and maintenance data over a historical period, a set of sample antenna offset parameters, a set of sample environmental parameters, and antenna impact offset parameters under the influence of different sample antenna offset parameters and sample environmental parameters are collected to obtain a set of sample antenna impact offset parameters.

[0028] Construct an antenna effect offset analyzer based on machine learning;

[0029] The antenna influence offset analyzer is trained under supervision using the sample antenna offset parameter set, sample environment parameter set, and sample antenna influence offset parameter set, and the analysis is completed after convergence.

[0030] Optionally, based on multiple offset confidence levels and combined with multiple antenna influence offset parameters, an antenna fault scale is calculated and used as the base station operation and maintenance fault prediction result, including:

[0031] The magnitude by which the influence offset parameters of multiple antennas exceed the fault offset parameters is calculated separately to obtain the fault scale of multiple monitoring antennas;

[0032] By assigning weights to multiple offset confidence levels, the fault scales of multiple monitoring antennas are weighted and calculated to obtain the antenna fault scale, which serves as the prediction result for base station operation and maintenance faults.

[0033] Secondly, this application provides an automatic fault prediction system for base station operation and maintenance, comprising:

[0034] The signal feature analysis module is used to collect multiple signal features at multiple monitoring locations of the communication base station, analyze the signal feature changes of the base station antenna, and obtain multiple signal change parameters.

[0035] The credibility analysis module is used to perform base station antenna offset analysis based on multiple signal change parameters, obtain multiple antenna offset parameters, perform offset credibility analysis, and obtain multiple offset credibility values.

[0036] The influence offset analysis module is used to obtain environmental parameters in the environment where the communication base station is located, and combine them with multiple antenna offset parameters to perform base station antenna influence offset analysis and obtain multiple antenna influence offset parameters.

[0037] The fault prediction module is used to calculate the antenna fault scale based on multiple offset confidence levels and multiple antenna influence offset parameters, and use it as the base station operation and maintenance fault prediction result.

[0038] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0039] This application proposes an automatic fault prediction method and system for base station operation and maintenance. By collecting signal characteristics and analyzing their changing parameters at multiple monitoring locations, and dynamically calculating the antenna fault scale by combining environmental factors and offset reliability, the method significantly improves the early prediction accuracy and reliability of potential base station antenna offset faults. Compared with traditional methods, the technical solution provided in this application overcomes the limitations of passive threshold alarms, realizes multi-dimensional collaborative analysis and quantitative evaluation of fault causes, and can stably output highly reliable fault warning information even in complex and ever-changing actual deployment environments. Ultimately, it achieves the technical effect of accurately predicting base station antenna offset faults and guiding timely operation and maintenance under environmental interference conditions. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating an automatic fault prediction method for base station operation and maintenance provided in an embodiment of this application.

[0042] Figure 2 This is a schematic diagram of an automatic fault prediction system for base station operation and maintenance provided in an embodiment of this application.

[0043] The components represented by each number in the attached diagram are explained below:

[0044] Signal feature analysis module 100, credibility analysis module 200, influence offset analysis module 300, and fault prediction module 400. Detailed Implementation

[0045] This application provides an automatic fault prediction method and system for base station operation and maintenance, which addresses the technical problems of poor accuracy and timeliness in fault prediction in existing base station operation and maintenance technologies.

[0046] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0047] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0048] Example 1, as Figure 1 As shown, this application provides an automatic fault prediction method for base station operation and maintenance, wherein the method includes:

[0049] S10: Collect multiple signal features at multiple monitoring locations of the communication base station, analyze the signal feature changes of the base station antenna, and obtain multiple signal change parameters.

[0050] In traditional base station operation and maintenance, antenna condition monitoring typically relies on signal feature acquisition from single or limited locations, resulting in insufficient coverage and sensitivity for signal change analysis. Limited signal feature information makes it difficult to comprehensively capture signal changes caused by physical antenna offset, especially regarding the ability to detect early, subtle signs of offset.

[0051] Step S10 in the method provided in this application embodiment includes:

[0052] Multiple signal characteristics are collected at various monitoring locations of the communication base station, including signal quality.

[0053] Acquire multiple standard signal characteristics at multiple monitoring locations;

[0054] The variation amplitudes of multiple signal features and multiple standard signal features are calculated separately to obtain multiple signal variation parameters.

[0055] In this embodiment of the application, at multiple locations of the communication base station, such as 2 meters east of the base station center and 2 meters west of the base station center, a general spectrum analyzer is used to collect multiple signal features, including signal quality, which is a parameter reflecting signal strength and is characterized by received signal strength, with the unit being dBm.

[0056] Multiple standard signal characteristics at multiple monitoring locations are obtained. Standard signal characteristics refer to the signal strength that the monitoring location should have under standard operating conditions according to the theoretical design value. These characteristics can be extracted from the production design data of the communication base station.

[0057] The variation amplitudes of multiple signal features and multiple standard signal features are calculated separately to obtain multiple signal variation parameters. Wherein, the signal variation parameter = |signal feature - standard signal feature|.

[0058] By simultaneously acquiring signal features at multiple monitoring locations and calculating their variation amplitudes compared to standard features, the spatial coverage and sensitivity of signal change analysis are significantly improved. Multi-location data acquisition can capture regional signal anomaly distribution patterns caused by antenna offset. The dynamic comparison mechanism based on standard features can transform the raw signal into quantifiable signal change parameters, eliminating inaccuracies caused by benchmark differences and providing highly consistent input data for subsequent analysis.

[0059] S20: Perform base station antenna offset analysis based on multiple signal variation parameters to obtain multiple antenna offset parameters, perform offset reliability analysis, and obtain multiple offset reliability parameters.

[0060] Existing technologies do not take into account the differences in the reliability of data from different monitoring locations. For example, signals that are far from the base station antenna have poor reliability, which leads to inaccurate offset analysis results.

[0061] Step S20 in the method provided in this application embodiment includes:

[0062] Each signal variation parameter is input into the antenna offset analyzer, and multiple antenna offset parameters are output. Each antenna offset parameter includes the offset direction and offset scale.

[0063] The steps for obtaining the antenna offset analyzer include:

[0064] Based on the historical offset maintenance data of the base station antenna, multiple sets of sample signal change parameters are collected from multiple monitoring locations, and antenna offset parameters of the base station antenna under different signal change parameters are collected and labeled to obtain multiple sets of sample antenna offset parameters.

[0065] Based on machine learning, multiple antenna offset analysis paths with the same architecture are constructed.

[0066] Multiple antenna offset analysis paths are trained under supervision using multiple sets of sample signal variation parameters and multiple sets of sample antenna offset parameters. After convergence, they are integrated to obtain an antenna offset analyzer.

[0067] Calculate the similarity between each antenna offset parameter and the mean of multiple antenna offset parameters to obtain multiple first offset confidence scores;

[0068] Multiple monitoring distances between multiple monitoring locations and base station antennas are obtained, and multiple second offset confidence levels are calculated. The magnitude of the monitoring distance is negatively correlated with the magnitude of the second offset confidence level.

[0069] Multiple offset confidence levels are calculated based on multiple first offset confidence levels and multiple second offset confidence levels.

[0070] In this embodiment, multiple sets of sample signal change parameters are collected from multiple monitoring locations based on the historical offset maintenance data of the base station antenna. Preferably, the historical event can be set to the past 180 days, and multiple signal change parameters from multiple monitoring locations within the past 180 days are collected. The signal change parameters of each monitoring location are integrated into a set to obtain multiple sets of sample signal change parameters. Antenna offset parameters of the base station antenna are also collected under different signal change parameters. During operation, the base station antenna may experience slight offsets, leading to signal changes. The offset of the base station antenna azimuth angle is used to characterize the base station antenna offset parameters, where each antenna offset parameter includes the offset direction and offset scale. For example, due east is denoted as 0 degrees, and the clockwise offset angle is the offset direction, in degrees. The offset scale is characterized by the straight-line distance between the offset antenna end position and the unoffset antenna end position, in centimeters. The antenna offset parameters are labeled and integrated, with the labeling content including the signal change parameters and monitoring location. The antenna offset parameters of each monitoring location are integrated into a set to obtain multiple sets of sample antenna offset parameters.

[0071] Based on machine learning, multiple antenna offset analysis paths with the same architecture are constructed. Specifically, a three-layer structure is built, where the input layer receives the sample signal variation parameters, the hidden layer uses 32 nodes and is activated using the ReLU function, and the output layer outputs the analyzed antenna offset path.

[0072] Multiple antenna offset analysis paths are trained under supervised supervision using multiple sets of sample signal variation parameters and multiple sets of sample antenna offset parameters until convergence. For example, if the accuracy of the output antenna offset parameters is above 90% when input signal variation parameters are used, the antenna offset analysis path training is considered complete. The multiple trained antenna offset paths are then integrated to obtain the antenna offset analyzer.

[0073] Each signal variation parameter is input into the antenna offset analyzer, which outputs multiple antenna offset parameters. Each antenna offset parameter includes the offset direction and offset scale.

[0074] Calculate the similarity between each antenna offset parameter and the mean of multiple antenna offset parameters to obtain multiple first offset confidence scores. The first offset confidence score is the mean of the offset direction similarity and the offset scale similarity. Offset direction confidence score = 1 - |Antenna offset direction - Mean antenna offset direction| ÷ [(Antenna offset direction + Mean antenna offset direction) ÷ 2], Offset scale confidence score = 1 - |Antenna offset scale - Mean antenna offset scale| ÷ [(Antenna offset scale + Mean antenna offset scale) ÷ 2], First offset confidence score = (Antenna offset direction confidence score + Offset scale confidence score) ÷ 2. For example, if the antenna offset direction is 3 degrees, the average offset direction is 5 degrees, the antenna offset scale is 30 cm, and the average offset scale is 40 cm, then the confidence level of the offset direction is 1 - |3 - 5| ÷ [(3 + 5) ÷ 2] = 0.5, the confidence level of the offset scale is 1 - |30 - 40| ÷ [(30 + 40) ÷ 2] = 0.71, and the confidence level of the first offset is (0.5 + 0.71) ÷ 2 = 0.605. The more similar the offset direction and scale, the closer the offset patterns are, and the greater the confidence level.

[0075] Multiple monitoring distances between multiple monitoring locations and base station antennas are obtained, and multiple second offset confidence levels are calculated. The magnitude of the monitoring distance is negatively correlated with the magnitude of the second offset confidence level. For example, based on map locations, multiple straight-line distances between the base station antennas at multiple monitoring locations are obtained as monitoring distances, in meters. Second offset confidence level = 1 - (monitoring distance - minimum monitoring distance) ÷ (maximum monitoring distance - minimum monitoring distance). For example, if a monitoring distance is 50 meters, the maximum monitoring distance is 100 meters, and the minimum monitoring distance is 1 meter, then the second offset confidence level = 1 - (50 - 1) ÷ (100 - 1) = 0.5. The farther the monitoring location, the more potential influencing factors there are, the greater the signal interference, and the lower the confidence level.

[0076] Multiple offset confidence levels are calculated based on multiple first offset confidence levels and multiple second offset confidence levels. Offset confidence level = (first offset confidence level + second offset confidence level) ÷ 2. For example, if the first offset confidence level is 0.605 and the second offset confidence level is 0.5, then the offset confidence level = (0.605 + 0.5) ÷ 2 = 0.552.

[0077] By employing a offset reliability analysis mechanism, after generating antenna offset parameters, the similarity between these parameters and the overall offset mean is calculated simultaneously. Furthermore, the negative impact of monitoring distance on data reliability is considered to dynamically generate a multi-dimensional offset reliability index. This mechanism effectively quantifies the reliability of data from different locations, significantly suppresses the interference of local anomalies on the overall offset judgment, and improves the noise resistance and reliability of the offset analysis results.

[0078] S30: Obtain environmental parameters within the environment where the communication base station is located, and combine them with multiple antenna offset parameters to perform base station antenna influence offset analysis and obtain multiple antenna influence offset parameters.

[0079] Traditional fault prediction methods often process environmental parameters and antenna offset data in isolation, without considering the impact of external dynamic disturbances such as weather changes. Environmental factors are treated as independent threshold alarm items or simply filtered, leading to distorted fault prediction.

[0080] Step S30 in the method provided in this application embodiment includes:

[0081] Obtain environmental parameters within the environment where the communication base station is located, including meteorological parameters;

[0082] The environmental parameters are combined with each antenna offset parameter and input into the antenna influence offset analyzer to obtain multiple antenna influence offset parameters.

[0083] The steps for constructing the antenna influence offset analyzer include:

[0084] Based on base station antenna operation and maintenance data over a historical period, a set of sample antenna offset parameters, a set of sample environmental parameters, and antenna impact offset parameters under the influence of different sample antenna offset parameters and sample environmental parameters are collected to obtain a set of sample antenna impact offset parameters.

[0085] Construct an antenna effect offset analyzer based on machine learning;

[0086] The antenna influence offset analyzer is trained under supervision using the sample antenna offset parameter set, sample environment parameter set, and sample antenna influence offset parameter set, and the analysis is completed after convergence.

[0087] Wind can cause antenna misalignment, affecting signal characteristics. Therefore, meteorological parameters must be considered to ensure more comprehensive and accurate base station operation and maintenance. In this embodiment, environmental parameters of the communication base station's location are obtained based on meteorological station data. These environmental parameters include meteorological parameters such as wind speed and direction. For example, the wind speed is level 3, and the wind direction is northeast.

[0088] Data is collected from base station antenna operation and maintenance data to obtain a set of sample antenna offset parameters, a set of sample environmental parameters, and antenna impact offset parameters under the influence of different sample antenna offset parameters and sample environmental parameters, thus obtaining a set of sample antenna impact offset parameters. The antenna impact offset parameters refer to the parameters that cause antenna offset during operation and maintenance due to changes in environmental parameters, such as strong winds. These parameters include the direction and scale of the impact offset. For example, due east is designated as 0 degrees, and the clockwise offset angle is the direction of impact offset, measured in degrees. The scale of impact offset is characterized by the straight-line distance between the offset antenna end position and the unoffset antenna end position, measured in centimeters.

[0089] A machine learning-based antenna influence offset analyzer is constructed, preferably using a four-layer structure. The input layer is used to receive sample antenna offset parameters and environmental parameters. The first hidden layer uses 32 nodes activated by the ReLU function, the second hidden layer uses 16 nodes activated by the ReLU function, and the output layer is used to output the analyzed antenna influence offset parameters.

[0090] The antenna influence offset analyzer is trained under supervision using a set of sample antenna offset parameters, a set of sample environmental parameters, and a set of sample antenna influence offset parameters until convergence. For example, if the input antenna offset parameters and environmental parameters have an error in the output antenna influence offset parameters within the range of 0.2 degrees and 3 centimeters, the training is considered converged and the antenna influence offset analyzer training is complete.

[0091] The environmental parameters are combined with each antenna offset parameter and input into the antenna influence offset analyzer, which outputs the antenna influence offset parameters.

[0092] By designing an antenna influence offset analyzer, environmental parameters and antenna offset parameters at various locations are analyzed collaboratively, and antenna influence offset parameters with fused environmental information are output. This achieves coupled analysis of environmental factors and physical offset, and significantly improves the accuracy of offset parameters in complex environments.

[0093] S40: Based on multiple offset confidence levels and combined with multiple antenna influence offset parameters, the antenna fault scale is calculated and used as the base station operation and maintenance fault prediction result.

[0094] In the final fault prediction stage, existing methods often directly aggregate offset data from multiple locations without distinguishing the contribution weights of data with different levels of credibility. This makes the prediction results susceptible to the influence of low-credibility data and difficult to support operation and maintenance decisions.

[0095] Step S40 in the method provided in this application embodiment includes:

[0096] The magnitude by which the influence offset parameters of multiple antennas exceed the fault offset parameters is calculated separately to obtain the fault scale of multiple monitoring antennas;

[0097] By assigning weights to multiple offset confidence levels, the fault scales of multiple monitoring antennas are weighted and calculated to obtain the antenna fault scale, which serves as the prediction result for base station operation and maintenance faults.

[0098] In this embodiment, the magnitude by which multiple antenna influence offset parameters exceed the fault offset parameter is calculated to obtain multiple monitoring antenna fault scales. The fault offset parameter refers to the parameter indicating a potential fault requiring maintenance if the offset exceeds this parameter. Monitoring antenna fault scale = |Antenna influence offset parameter - Fault offset parameter|. A tolerance range is set for the monitoring antenna fault scale; for example, the tolerance range for the offset direction is set to 0.3 degrees, and the tolerance range for the offset scale is set to 5 centimeters. Fault scales within this tolerance range are acceptable.

[0099] Based on weighted assignments to multiple offset confidence levels, the fault scales of multiple monitoring antennas are weighted and calculated to obtain the antenna fault scale, which serves as the prediction result for base station operation and maintenance faults. For example, one monitoring antenna fault scale has an offset direction of 1 degree, an offset scale of 10 cm, and a confidence level of 0.65; another monitoring antenna fault scale has an offset direction of 1.5 degrees, an offset scale of 8 cm, and a confidence level of 0.7. Then, in the antenna fault scale, the offset direction fault scale = ∑[offset direction × (confidence level ÷ sum of confidence levels)] = 1 × (0.65 ÷ 1.35) + 1.5 × (0.7 ÷ 1.35) = 1.25, and the offset scale fault scale = ∑[offset scale × (confidence level ÷ sum of confidence levels)] = 10 × (0.65 ÷ 1.35) + 8 × (0.7 ÷ 1.35) = 8.95. The fault scale in the offset direction and the fault scale in the offset scale are compared with the tolerance range of the offset scale. If the offset direction or the offset scale is greater than the tolerance range of the offset scale, it indicates that the probability of a fault is relatively high. A fault warning message is sent to the operation and maintenance personnel, such as sending a text message "The base station antenna may be faulty, please check it in time".

[0100] This application calculates fault scales at each location using dynamic weighting based on offset reliability and aggregates them to generate fault scales. By assigning weights based on reliability, it mitigates the negative impact of low-reliability data. Furthermore, by setting a tolerance range, it reduces false alarms that may be caused by minor changes, improving the accuracy of early warnings and providing a reliable basis for operational decisions.

[0101] Example 2, as Figure 2 As shown, based on the same inventive concept as the automatic fault prediction method for base station operation and maintenance provided in Embodiment 1, this embodiment of the invention also provides an automatic fault prediction system for base station operation and maintenance, comprising:

[0102] The signal feature analysis module 100 is used to collect multiple signal features at multiple monitoring locations of the communication base station, perform signal feature change analysis of the base station antenna, and obtain multiple signal change parameters.

[0103] The credibility analysis module 200 is used to perform base station antenna offset analysis based on multiple signal change parameters, obtain multiple antenna offset parameters, perform offset credibility analysis, and obtain multiple offset credibility values.

[0104] The influence offset analysis module 300 is used to obtain environmental parameters in the environment where the communication base station is located, and combine them with multiple antenna offset parameters to perform base station antenna influence offset analysis and obtain multiple antenna influence offset parameters.

[0105] The fault prediction module 400 is used to calculate the antenna fault scale based on multiple offset confidence levels and multiple antenna influence offset parameters, and use it as the base station operation and maintenance fault prediction result.

[0106] In one embodiment, the signal feature analysis module 100 is further configured to:

[0107] Multiple signal characteristics are collected at various monitoring locations of the communication base station, including signal quality.

[0108] Acquire multiple standard signal characteristics at multiple monitoring locations;

[0109] The variation amplitudes of multiple signal features and multiple standard signal features are calculated separately to obtain multiple signal variation parameters.

[0110] In one embodiment, the credibility analysis module 200 is further configured to:

[0111] Each signal variation parameter is input into the antenna offset analyzer, and multiple antenna offset parameters are output. Each antenna offset parameter includes the offset direction and offset scale.

[0112] The steps for obtaining the antenna offset analyzer include:

[0113] Based on the historical offset maintenance data of the base station antenna, multiple sets of sample signal change parameters are collected from multiple monitoring locations, and antenna offset parameters of the base station antenna under different signal change parameters are collected and labeled to obtain multiple sets of sample antenna offset parameters.

[0114] Based on machine learning, multiple antenna offset analysis paths with the same architecture are constructed.

[0115] Multiple antenna offset analysis paths are trained under supervision using multiple sets of sample signal variation parameters and multiple sets of sample antenna offset parameters. After convergence, they are integrated to obtain an antenna offset analyzer.

[0116] Calculate the similarity between each antenna offset parameter and the mean of multiple antenna offset parameters to obtain multiple first offset confidence scores;

[0117] Multiple monitoring distances between multiple monitoring locations and base station antennas are obtained, and multiple second offset confidence levels are calculated. The magnitude of the monitoring distance is negatively correlated with the magnitude of the second offset confidence level.

[0118] Multiple offset confidence levels are calculated based on multiple first offset confidence levels and multiple second offset confidence levels.

[0119] In one embodiment, the influence offset analysis module 300 is further configured to:

[0120] Obtain environmental parameters within the environment where the communication base station is located, including meteorological parameters;

[0121] The environmental parameters are combined with each antenna offset parameter and input into the antenna influence offset analyzer to obtain multiple antenna influence offset parameters.

[0122] The steps for constructing the antenna influence offset analyzer include:

[0123] Based on base station antenna operation and maintenance data over a historical period, a set of sample antenna offset parameters, a set of sample environmental parameters, and antenna impact offset parameters under the influence of different sample antenna offset parameters and sample environmental parameters are collected to obtain a set of sample antenna impact offset parameters.

[0124] Construct an antenna effect offset analyzer based on machine learning;

[0125] The antenna influence offset analyzer is trained under supervision using the sample antenna offset parameter set, sample environment parameter set, and sample antenna influence offset parameter set, and the analysis is completed after convergence.

[0126] In one embodiment, the fault prediction module 400 is further configured to:

[0127] The magnitude by which the influence offset parameters of multiple antennas exceed the fault offset parameters is calculated separately to obtain the fault scale of multiple monitoring antennas;

[0128] By assigning weights to multiple offset confidence levels, the fault scales of multiple monitoring antennas are weighted and calculated to obtain the antenna fault scale, which serves as the prediction result for base station operation and maintenance faults.

[0129] In summary, the embodiments of this application have at least the following technical effects:

[0130] This application proposes an automatic fault prediction method and system for base station operation and maintenance. By collecting signal characteristics and analyzing their changing parameters at multiple monitoring locations, and dynamically calculating the antenna fault scale by combining environmental factors and offset reliability, the early prediction accuracy and reliability of potential antenna offset faults are significantly improved. Specifically, by extracting and comparing the signal changing parameters at multiple locations, early and weak signs of antenna physical offset can be keenly captured. Furthermore, an offset reliability analysis mechanism is introduced to effectively quantify the reliability differences of data from different monitoring locations, significantly reducing the interference of local noise or abnormal data on the overall judgment. By coupling environmental parameters with antenna offset parameters, the actual impact of external dynamic disturbances on antenna stability is deeply analyzed, overcoming the shortcomings of traditional methods that view base station data in isolation and ignore the coupling effect of the environment. Finally, the antenna fault scale is calculated comprehensively based on the dynamic weighted offset reliability, so that the prediction results can reflect both the severity of potential offset and the reliability of the data source, greatly improving the accuracy of the prediction conclusion. Through a systematic multi-source information fusion and reliability quantification mechanism, the generalization ability and anti-interference ability of fault prediction are significantly optimized. Compared with traditional methods, the technical solution provided in this application breaks through the limitations of passive threshold alarms, realizes multi-dimensional collaborative analysis and quantitative evaluation of fault causes, and can stably output highly reliable fault warning information in complex and ever-changing actual deployment environments. Ultimately, it achieves the technical effect of accurately predicting base station antenna offset faults and guiding timely operation and maintenance under environmental interference conditions.

[0131] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0132] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0133] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An automatic fault prediction method for base station operation and maintenance, characterized in that, The method includes: Multiple signal features are collected at multiple monitoring locations of the communication base station, and the signal feature changes of the base station antenna are analyzed to obtain multiple signal change parameters. Base station antenna offset analysis is performed based on multiple signal variation parameters to obtain multiple antenna offset parameters. Offset confidence analysis is then performed to obtain multiple offset confidence levels. The environmental parameters within the environment where the communication base station is located are obtained, and combined with multiple antenna offset parameters, the base station antenna impact offset analysis is performed to obtain multiple antenna impact offset parameters. The antenna impact offset parameters refer to the parameters that cause antenna offset due to changes in environmental parameters during operation and maintenance. Based on multiple offset confidence levels and combined with multiple antenna influence offset parameters, the antenna fault scale is calculated and used as the base station operation and maintenance fault prediction result. Base station antenna offset analysis is performed based on multiple signal variation parameters to obtain multiple antenna offset parameters. Offset reliability analysis is then conducted to obtain multiple offset reliability values, including: Each signal variation parameter is input into the antenna offset analyzer, and multiple antenna offset parameters are output. Each antenna offset parameter includes the offset direction and offset scale. Calculate the similarity between each antenna offset parameter and the mean of multiple antenna offset parameters to obtain multiple first offset confidence scores; Multiple monitoring distances between multiple monitoring locations and base station antennas are obtained, and multiple second offset confidence levels are calculated. The magnitude of the monitoring distance is negatively correlated with the magnitude of the second offset confidence level. Multiple offset confidence levels are calculated based on multiple first offset confidence levels and multiple second offset confidence levels; Environmental parameters within the environment where the communication base station is located are obtained, and combined with multiple antenna offset parameters, base station antenna impact offset analysis is performed to obtain multiple antenna impact offset parameters, including: Obtain environmental parameters within the environment where the communication base station is located, including meteorological parameters; The environmental parameters are combined with each antenna offset parameter and input into the antenna influence offset analyzer to obtain multiple antenna influence offset parameters.

2. The automatic fault prediction method for base station operation and maintenance according to claim 1, characterized in that, Multiple signal characteristics were collected at various monitoring locations of the communication base station. Signal characteristic changes in the base station antenna were analyzed to obtain multiple signal change parameters, including: Multiple signal characteristics are collected at various monitoring locations of the communication base station, including signal quality. Acquire multiple standard signal characteristics at multiple monitoring locations; The variation amplitudes of multiple signal features and multiple standard signal features are calculated separately to obtain multiple signal variation parameters.

3. The automatic fault prediction method for base station operation and maintenance according to claim 1, characterized in that, The antenna offset analyzer includes: Based on the historical offset maintenance data of the base station antenna, multiple sets of sample signal change parameters are collected from multiple monitoring locations, and antenna offset parameters of the base station antenna under different signal change parameters are collected and labeled to obtain multiple sets of sample antenna offset parameters. Based on machine learning, multiple antenna offset analysis paths with the same architecture are constructed. Multiple antenna offset analysis paths are trained under supervision using multiple sets of sample signal variation parameters and multiple sets of sample antenna offset parameters. After convergence, they are integrated to obtain an antenna offset analyzer.

4. The automatic fault prediction method for base station operation and maintenance according to claim 1, characterized in that, The steps for constructing the antenna influence offset analyzer include: Based on base station antenna operation and maintenance data over a historical period, a set of sample antenna offset parameters, a set of sample environmental parameters, and antenna impact offset parameters under the influence of different sample antenna offset parameters and sample environmental parameters are collected to obtain a set of sample antenna impact offset parameters. Construct an antenna effect offset analyzer based on machine learning; The antenna influence offset analyzer is trained under supervision using the sample antenna offset parameter set, sample environment parameter set, and sample antenna influence offset parameter set, and the analysis is completed after convergence.

5. The automatic fault prediction method for base station operation and maintenance according to claim 1, characterized in that, Based on multiple offset confidence levels and combined with multiple antenna influence offset parameters, the antenna fault scale is calculated and used as the base station operation and maintenance fault prediction result, including: The magnitude by which the influence offset parameters of multiple antennas exceed the fault offset parameters is calculated separately to obtain the fault scale of multiple monitoring antennas; By assigning weights to multiple offset confidence levels, the fault scales of multiple monitoring antennas are weighted and calculated to obtain the antenna fault scale, which serves as the prediction result for base station operation and maintenance faults.

6. An automatic fault prediction system for base station operation and maintenance, characterized in that, The system is used to implement the automatic fault prediction method for base station operation and maintenance as described in any one of claims 1-5, the system comprising: The signal feature analysis module is used to collect multiple signal features at multiple monitoring locations of the communication base station, analyze the signal feature changes of the base station antenna, and obtain multiple signal change parameters. The credibility analysis module is used to perform base station antenna offset analysis based on multiple signal change parameters, obtain multiple antenna offset parameters, perform offset credibility analysis, and obtain multiple offset credibility values. The influence offset analysis module is used to obtain environmental parameters in the environment where the communication base station is located, and combine them with multiple antenna offset parameters to perform base station antenna influence offset analysis and obtain multiple antenna influence offset parameters. The fault prediction module is used to calculate the antenna fault scale based on multiple offset confidence levels and multiple antenna influence offset parameters, and use it as the base station operation and maintenance fault prediction result.

Images