Fault determination method, device and equipment and computer readable storage medium
By combining the Long Short-Term Memory model and the XGBoost classification model, the problem of accuracy in fault prediction for telecommunications operation systems was solved, achieving more accurate fault prediction and intelligent early warning, and improving system reliability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA MOBILE GROUP ZHEJIANG
- Filing Date
- 2022-02-16
- Publication Date
- 2026-07-10
AI Technical Summary
Existing predictive models based on bidirectional LSTM are not accurate enough in predicting faults in telecommunications operating systems.
A method combining a long short-term memory model and an XGBoost classification model is adopted. By acquiring the operational parameter data of the telecommunications operation system, the prediction model is used to predict the fault type at the target time, and the fault of the telecommunications operation system is determined when the classification result of the XGBoost classification model indicates that a fault has occurred.
It improves the accuracy of fault prediction in telecommunications operating systems and enables intelligent early warning and timely handling of faults.
Smart Images
Figure CN116663692B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, and computer-readable storage medium for determining faults. Background Technology
[0002] Telecommunications operating systems are large-scale and complex, with long business processes. From service acceptance to network construction and activation, they span numerous platforms, including the CRM front-end, business capability middleware (customer center, order center, billing and pricing center, accounting center, etc.), instruction orchestrator, network management resource center, and various network elements on the network side. Currently, when predicting whether a telecommunications operating system will experience a fault, the prediction results are inaccurate due to the influence of a single model. Summary of the Invention
[0003] The main objective of this invention is to provide a method, apparatus, device, and computer-readable storage medium for determining faults, with the aim of improving the accuracy of predicting faults occurring in telecommunications operating systems.
[0004] To achieve the above objectives, the present invention provides a method for determining a fault, the method comprising the following steps:
[0005] The system acquires operational parameter data of the telecommunications operation system during a preset time period, inputs the operational parameter data into the input layer of the prediction model, and obtains the fault type predicted by the prediction model at the target time.
[0006] When the fault type is the type corresponding to the occurrence of a fault, the operational parameter data is input into the input layer of the XGBoost classification model to obtain the classification result of the XGBoost classification model;
[0007] When the classification result indicates a fault has occurred, it is determined that the telecommunications operation system has experienced a fault at the target time.
[0008] In one embodiment, before the step of inputting the operational parameter data into the input layer of the prediction model to obtain the fault type predicted by the prediction model at the target time, the method includes:
[0009] Obtain the first sample data and clean it.
[0010] The target data for a preset time period is obtained from the first sample data after cleaning to predict the fault type at the preset time.
[0011] Update the preset time period and obtain the predicted fault type to construct target sample data;
[0012] The operational parameter data of the target sample data is input into the input layer of the preset long short-term memory model, and the network hierarchy of the long short-term memory model is adjusted.
[0013] The dimension of the fully connected layer of the long short-term memory model is determined based on the number of fault types, and the activation parameter of the prediction network model is set to softmax.
[0014] The prediction model is obtained by adjusting the parameters of the long short-term memory model based on the fault types in the first sample data.
[0015] In one embodiment, the step of cleaning the first sample data includes:
[0016] Remove abnormal data from the first sample data and fill in missing values to complete the cleaning of the first sample data.
[0017] In one embodiment, the step of inputting the operational parameter data into the input layer of an XGBoost classification model and obtaining the classification result of the XGBoost classification model includes:
[0018] The operational parameter data is input into the input layer of the XGBoost classification model;
[0019] The XGBoost classification model, which uses a decision tree as a base classifier, classifies the prediction results of the operational parameter data.
[0020] The output layer of the XGBoost classification model outputs the classification with the highest score.
[0021] In one embodiment, before the step of inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of the XGBoost classification model when the fault type is the type corresponding to the occurrence of a fault, the method includes:
[0022] Obtain the second sample data and label the second sample data;
[0023] The operational parameter data of the labeled second sample data is input into a preset XGBoost classification model for training, resulting in the XGBoost classification model.
[0024] In one embodiment, after the step of determining that the telecommunications operating system has failed at the target time when the classification result indicates a failure, the method further includes:
[0025] Retrieve the staff members corresponding to the fault type from the association table;
[0026] Obtain the contact information of the aforementioned staff member;
[0027] The notification information, carrying the fault type and the target time, will be sent to the staff through the aforementioned contact method to notify the staff to handle the situation.
[0028] In one embodiment, before the step of obtaining the staff member corresponding to the fault type from the association table, the following steps are included:
[0029] Obtain third sample data and analyze the third sample data using a statistical model;
[0030] The association table is constructed based on the analysis results.
[0031] To achieve the above objectives, the present invention also provides a fault determination apparatus, the fault determination apparatus comprising:
[0032] The first acquisition module is used to acquire the operation parameter data of the telecommunications operation system during a preset time period, input the operation parameter data into the input layer of the prediction model, and obtain the fault type predicted by the prediction model at the target time.
[0033] The second acquisition module is used to input the operation parameter data into the input layer of the XGBoost classification model when the fault type is the type corresponding to the occurrence of the fault, and to obtain the classification result of the XGBoost classification model.
[0034] The determination module is used to determine that the telecommunications operation system has failed at the target time when the classification result indicates that a failure has occurred.
[0035] To achieve the above objectives, the present invention also provides a fault determination device, the fault determination device comprising a memory, a processor, and a fault determination program stored in the memory and executable on the processor, wherein the fault determination program, when executed by the processor, implements the various steps of the fault determination method as described above.
[0036] To achieve the above objectives, the present invention also provides a computer-readable storage medium storing a fault determination program, which, when executed by a processor, implements the various steps of the fault determination method described above.
[0037] This invention provides a method, apparatus, device, and computer-readable storage medium for determining faults. It acquires operational parameter data of a telecommunications operating system over a preset time period, inputs this data into the input layer of a prediction model to obtain the fault type predicted by the model at a target time, and when the fault type matches the predicted fault type, inputs the operational parameter data into the input layer of an XGBoost classification model to obtain the classification result. If the classification result indicates a fault has occurred, it is determined that a fault has occurred in the telecommunications operating system at the target time. By determining the fault type of the telecommunications operating system at the target time through the prediction model and then using the classification result of the XGBoost classification model to determine when a fault has occurred, the accuracy of determining that a fault has occurred in the telecommunications operating system at the target time is improved. Attached Figure Description
[0038] Figure 1 This is a schematic diagram of the hardware structure of the fault determination device involved in an embodiment of the present invention;
[0039] Figure 2 This is a flowchart illustrating the first embodiment of the fault determination method of the present invention;
[0040] Figure 3 This is a detailed flowchart of step S30 in the fourth embodiment of the fault determination method of the present invention.
[0041] Figure 4 This is a schematic diagram of the module of the fault determination method of the present invention.
[0042] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0043] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0044] The main solution of this invention is as follows: Obtain operational parameter data of a telecommunications operating system during a preset time period; input the operational parameter data into the input layer of a prediction model to obtain the fault type predicted by the prediction model at the target time; when the fault type corresponds to the type of fault occurrence, input the operational parameter data into the input layer of an XGBoost classification model to obtain the classification result of the XGBoost classification model; when the classification result indicates a fault occurrence, determine that the telecommunications operating system has experienced a fault at the target time.
[0045] As one implementation scheme, the fault determination device can be as follows: Figure 1 As shown.
[0046] The present invention relates to a fault determination device, which includes: a processor 101, such as a CPU, a memory 102, and a communication bus 103. The communication bus 103 is used to establish communication between these components.
[0047] Memory 102 can be high-speed RAM or stable memory (non-volatile memory), such as disk storage. Figure 1 As shown, the memory 102, which is a computer-readable storage medium, may include a fault determination program; and the processor 101 may be used to call the fault determination program stored in the memory 102 and perform the following operations:
[0048] The system acquires operational parameter data of the telecommunications operation system during a preset time period, inputs the operational parameter data into the input layer of the prediction model, and obtains the fault type predicted by the prediction model at the target time.
[0049] When the fault type is the type corresponding to the occurrence of a fault, the operational parameter data is input into the input layer of the XGBoost classification model to obtain the classification result of the XGBoost classification model;
[0050] When the classification result indicates a fault has occurred, it is determined that the telecommunications operation system has experienced a fault at the target time.
[0051] In one embodiment, the processor 101 may be used to invoke a fault determination program stored in the memory 102 and perform the following operations:
[0052] Obtain the first sample data and clean it.
[0053] The target data for a preset time period is obtained from the first sample data after cleaning to predict the fault type at the preset time.
[0054] Update the preset time period and obtain the predicted fault type to construct target sample data;
[0055] The operational parameter data of the target sample data is input into the input layer of the preset long short-term memory model, and the network hierarchy of the long short-term memory model is adjusted.
[0056] The number of fully connected layers in the long short-term memory model is determined based on the number of fault types, and the activation parameter of the prediction network model is set to softmax.
[0057] The prediction model is obtained by adjusting the parameters of the long short-term memory model based on the fault types in the first sample data.
[0058] In one embodiment, the processor 101 may be used to invoke a fault determination program stored in the memory 102 and perform the following operations:
[0059] Remove abnormal data from the first sample data and fill in missing values to complete the cleaning of the first sample data.
[0060] In one embodiment, the processor 101 may be used to invoke a fault determination program stored in the memory 102 and perform the following operations:
[0061] The operational parameter data is input into the input layer of the XGBoost classification model;
[0062] The XGBoost classification model, which uses a decision tree as a base classifier, classifies the prediction results of the operational parameter data.
[0063] The output layer of the XGBoost classification model outputs the classification with the highest score.
[0064] In one embodiment, the processor 101 may be used to invoke a fault determination program stored in the memory 102 and perform the following operations:
[0065] Obtain the second sample data and label the second sample data;
[0066] The operational parameter data of the labeled second sample data is input into a preset XGBoost classification model for training, thereby obtaining the XGBoost classification model.
[0067] In one embodiment, the processor 101 may be used to invoke a fault determination program stored in the memory 102 and perform the following operations:
[0068] Retrieve the staff members corresponding to the fault type from the association table;
[0069] Obtain the contact information of the aforementioned staff member;
[0070] The notification information, carrying the fault type and the target time, will be sent to the staff through the aforementioned contact method to notify the staff to handle the situation.
[0071] In one embodiment, the processor 101 may be used to invoke a fault determination program stored in the memory 102 and perform the following operations:
[0072] Obtain third sample data and analyze the third sample data using a statistical model;
[0073] The association table is constructed based on the analysis results.
[0074] Based on the hardware architecture of the fault determination device described above, an embodiment of the fault determination method of the present invention is proposed.
[0075] Reference Figure 2 , Figure 2 This is a first embodiment of the fault determination method of the present invention, the fault determination method comprising the following steps:
[0076] Step S10: Obtain the operation parameter data of the telecommunications operation system during a preset time period, input the operation parameter data into the input layer of the prediction model, and obtain the fault type predicted by the prediction model at the target time.
[0077] The preset time period is the time period before the target time. In this embodiment, operational parameter data within the preset time period can be obtained at intervals and input into the input layer of the prediction model. For example, when the target time is T6, the target time period is from T1 to T5. Operational parameter data obtained at times T1, T2, T3, T4, and T5 are obtained respectively, and the operational parameter data is used as parameters to predict the fault type at the target time.
[0078] At least one operational parameter data exists, which may include: time, values corresponding to business characteristics (e.g., values corresponding to business characteristic 1, business characteristic 2, ..., business characteristic n), values corresponding to traffic (where traffic may include different values, such as values corresponding to traffic 1, traffic 2, ..., traffic n), and values corresponding to capacity (e.g., values corresponding to capacity 1, capacity 2, ..., capacity n). It is understood that in this embodiment, the operational parameter data may also be values corresponding to other data characteristics, which will not be described in detail here.
[0079] After obtaining the operational parameter data for the predicted time period of the telecommunications operation system, the operational parameter data is input into the input layer of the prediction model, and the prediction model is controlled to predict the fault type at the target time based on the operational parameter data input from the input layer.
[0080] In this embodiment, the fault types include: the types corresponding to the occurrence of faults such as first fault (D1), second fault (D2), third fault (D3), and fourth fault (D4), and the type of fault that did not occur (D0).
[0081] Step S20: When the fault type is the type corresponding to the occurrence of the fault, the operation parameter data is input into the input layer of the XGBoost classification model to obtain the classification result of the XGBoost classification model;
[0082] Step S30: When the classification result indicates that a fault occurred at the target time, it is determined that a fault occurred in the telecommunications operation system at the target time.
[0083] In this embodiment, after obtaining the fault type predicted by the prediction model, it is determined whether the fault type corresponds to the type of fault that has occurred. For example, if the fault type at the target time output by the prediction model is D4, where D4 is the type corresponding to the fault that has occurred, then it is determined that the prediction model predicts that the telecommunications operation system will experience a fault at the target time.
[0084] Furthermore, the operational parameter data is input into the XGBoost classification model, and the output result of the XGBoost classification model is obtained. If the output result indicates that a fault has occurred, it is determined that a fault has occurred in the telecommunications operation system at the target time.
[0085] The step of inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of the XGBoost classification model includes:
[0086] Step S31: Input the operational parameter data into the input layer of the XGBoost classification model;
[0087] Step S32: The prediction results of the operational parameter data are classified using the XGBoost classification model with decision tree as the base classifier;
[0088] Step S33: Output the classification with the highest score in the output layer of the XGBoost classification model.
[0089] In this embodiment, when the XGBoost classification model classifies the prediction results of operational parameter data, it outputs a classification result indicating whether a fault has occurred, i.e., either a fault has occurred or no fault has occurred. The decision tree assigns a score to the prediction structure of the operational parameter data based on these two classification results, and the output layer of the XGBoost classification model outputs the classification with the highest score. For example, if the score for "no fault" is higher than the score for "fault" is higher, the output layer of the XGBoost classification model will output a classification result indicating "no fault has occurred"; conversely, if the score for "fault" is higher than the score for "no fault" is higher, the output layer of the XGBoost classification model will output a classification result indicating "fault has occurred".
[0090] In this embodiment, operational parameter data of the telecommunications operating system during a preset time period is obtained. This data is then input into the input layer of a prediction model to obtain the fault type predicted by the model at the target time. When the fault type matches the predicted fault type, the operational parameter data is input into the input layer of an XGBoost classification model to obtain the classification result. If the classification result indicates a fault has occurred, it is determined that a fault has occurred in the telecommunications operating system at the target time. By determining that the fault type of the telecommunications operating system at the target time is the type corresponding to a fault occurrence through the prediction model, and then using the classification result of the XGBoost classification model to determine when a fault has occurred, the accuracy of determining that a fault has occurred in the telecommunications operating system at the target time is improved.
[0091] Based on the first embodiment, this application proposes a second embodiment of a fault determination method, wherein the method includes the following steps before step S20:
[0092] Before the step of inputting the operational parameter data into the input layer of the prediction model to obtain the fault type predicted by the prediction model at the target time, the following steps are included:
[0093] Step S01: Obtain the first sample data and clean the first sample data;
[0094] Step S02: Obtain the target data for the preset time period from the cleaned first sample data to predict the fault type at the preset time.
[0095] Step S03: Update the preset time period and obtain the predicted fault type to construct the second sample data;
[0096] Step S04: Input the operational parameter data of the second sample data into the input layer of the preset long short-term memory model, and adjust the network layer of the long short-term memory model;
[0097] Step S05: Determine the number of fully connected layers in the long short-term memory model based on the number of fault types, and set the activation parameter of the prediction network model to softmax;
[0098] Step S06: Adjust the parameters of the long short-term memory model according to the fault type in the first sample data to obtain the prediction model.
[0099] In this embodiment, first sample data is obtained from the original sample and then cleaned.
[0100] Optionally, in this embodiment, outlier data is acquired and removed from the first sample data during the cleaning process. Specifically, outlier data is identified based on outliers in each operational parameter data, and this outlier data is removed to prevent it from reducing the accuracy of the prediction results. The removed operational parameter data is then supplemented after the outlier data is removed.
[0101] Optionally, in this embodiment, the original samples may be as shown in Table 1:
[0102]
[0103] Table 1
[0104] Table 1 contains text that the model cannot recognize. During the cleaning process, the text in Table 1 needs to be converted into numerical values. Optionally, word2vec can be used to train word vectors to convert Chinese characters into numerical values. It is understood that the original samples contain fault types used to determine when a fault occurs in a telecommunications operating system, which are not shown in Table 1.
[0105] Optionally, in this embodiment, the first sample data after cleaning is shown in Table 2 below:
[0106] Table 2
[0107]
[0108]
[0109] The target data for a preset time period is obtained from the cleaned first sample data to predict the fault type at the preset time. For example, when the preset time is T6, the time period from T1 to T5 before T6 is determined as the preset time period. The operational parameter data within this time period is obtained, the fault type at T6 is predicted, and the preset time period is updated. The fault types predicted at different preset times are obtained, and the target sample data is shown in Table 3 below:
[0110] Table 3
[0111]
[0112] The operational parameters of the target sample data are input into the input layer of a preset Long Short-Term Memory (LSTM) model. The network hierarchy of the LTM model is adjusted, and the number of fully connected layers is determined based on the number of fault types. The activation parameter of the prediction network model is set to softmax. The parameters of the LTM model are adjusted according to the fault types in the first sample data to obtain the prediction model. In this embodiment, the prediction model is obtained by training with the target sample data, enabling the prediction of fault types in the e-commerce telecommunications operation system at a target time.
[0113] Based on the first embodiment, this application proposes a third embodiment. Before the step of inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of the XGBoost classification model when the fault type is the type corresponding to the occurrence of a fault, the method includes:
[0114] Step S21: Obtain the second sample data and label the second sample data;
[0115] Step S22: Input the operational parameter data of the labeled second sample data into the preset XGBoost classification model for training to obtain the XGBoost classification model.
[0116] In this embodiment, second sample data is obtained. The second sample data can be obtained by cleaning and transforming the original sample data in Table 1. In this embodiment, the specific steps for obtaining the second sample data will not be described in detail.
[0117] The preset XGBoost classification model is the initial XGBoost classification model. After training, the XGBoost classification model used for classification in this application is obtained. Optionally, the running parameters of the second sample data are input into the preset XGBoost classification model for training to obtain the XGBoost classification model.
[0118] The XGBoost classification model outputs two results: one indicating a fault has occurred, and the other indicating no fault has occurred. In this embodiment, after obtaining the second sample data, the second sample data is manually labeled. Optionally, label 1 represents a classification result of a fault occurring, and label 2 represents a classification result of no fault occurring. The labeled second sample data is then input into a preset XGBoost classification model for training to obtain the XGBoost classification model. In this embodiment, by using the second sample data to train the preset XGBoost classification model, an XGBoost classification model is obtained to determine whether a fault has occurred at a target time, thus realizing the determination of whether a fault exists in the telecommunications operation system based on the classification results of the XGBoost classification model.
[0119] Reference Figure 3 , Figure 3 This is a detailed flowchart following S30 of the fourth embodiment of this application. Based on the first embodiment, this application proposes a fourth embodiment. After the step of determining that the telecommunications operating system has failed at the target time when the classification result indicates a failure, the method further includes:
[0120] Step S40: Obtain the staff member corresponding to the fault type from the association table;
[0121] Step S50: Obtain the contact information of the staff member;
[0122] Step S60: Send notification information carrying the fault type and the target time to the staff through the contact method to notify the staff to handle the issue.
[0123] In this embodiment, third sample data is obtained from the original sample data. The third sample data includes data on fault type and fault handler, and may also include operational parameter data from other dimensions. This application does not limit the third sample data.
[0124] The statistical model used is the TF-IDF analysis model. After obtaining the third sample data, the statistical model is used to analyze the third sample data to identify the staff corresponding to each fault type when a fault occurs, and to calculate the frequency of staff handling different fault types, thus constructing an association table. In this embodiment, calculating the frequency of staff handling different fault types and constructing an association table allows for direct identification of the corresponding staff based on the fault type when a fault is determined to occur.
[0125] After identifying the staff member corresponding to the fault type, the system further obtains the staff member's contact information and sends a notification message containing the fault type and target time to the staff member to notify them to take action, thus achieving intelligent early warning.
[0126] Reference Figure 4 , Figure 4 This is a schematic diagram of the modules in this application. The present invention also provides a fault determination device, the fault determination device comprising:
[0127] The first acquisition module 10 acquires the operation parameter data of the telecommunications operation system during a preset time period, inputs the operation parameter data into the input layer of the prediction model, and obtains the fault type predicted by the prediction model at the target time.
[0128] The second acquisition module 20 is used to input the operation parameter data into the input layer of the XGBoost classification model when the fault type is the type corresponding to the occurrence of the fault, and to obtain the classification result of the XGBoost classification model.
[0129] The determination module 30 is used to determine that the telecommunications operation system has failed at the target time when the classification result indicates that a failure has occurred.
[0130] In one embodiment, before inputting the operational parameter data into the input layer of the prediction model to obtain the fault type predicted by the prediction model at the target time, the first acquisition module 10 is specifically used for:
[0131] Obtain the first sample data and clean it.
[0132] The target data for a preset time period is obtained from the first sample data after cleaning to predict the fault type at the preset time.
[0133] Update the preset time period and obtain the predicted fault type to construct target sample data;
[0134] The operational parameter data of the target sample data is input into the input layer of the preset long short-term memory model, and the network hierarchy of the long short-term memory model is adjusted.
[0135] The number of fully connected layers in the long short-term memory model is determined based on the number of fault types, and the activation parameter of the prediction network model is set to softmax.
[0136] The prediction model is obtained by adjusting the parameters of the long short-term memory model based on the fault types in the first sample data.
[0137] In one embodiment, in terms of cleaning the first sample data, the first acquisition module 10 is specifically used for:
[0138] Remove abnormal data from the first sample data and fill in missing values to complete the cleaning of the first sample data.
[0139] In one embodiment, in terms of inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of the XGBoost classification model, the second acquisition module 10 is specifically used for:
[0140] The operational parameter data is input into the input layer of the XGBoost classification model;
[0141] The XGBoost classification model, which uses a decision tree as a base classifier, classifies the prediction results of the operational parameter data.
[0142] The output layer of the XGBoost classification model outputs the classification with the highest score.
[0143] In one embodiment, when the fault type is the type corresponding to the occurrence of a fault, before inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of the XGBoost classification model, the second acquisition module 10 is specifically used for:
[0144] Obtain the second sample data and label the second sample data;
[0145] The operational parameter data of the labeled second sample data is input into a preset XGBoost classification model for training, thereby obtaining the XGBoost classification model.
[0146] In one embodiment, when the classification result indicates a fault has occurred, after determining that the telecommunications operation system has failed at the target time, the determining module 30 is specifically used for:
[0147] Retrieve the staff members corresponding to the fault type from the association table;
[0148] Obtain the contact information of the aforementioned staff member;
[0149] The notification information, carrying the fault type and the target time, will be sent to the staff through the aforementioned contact method to notify the staff to handle the situation.
[0150] In one embodiment, before retrieving the staff member corresponding to the fault type from the association table, the determining module 30 is specifically used for:
[0151] Obtain third sample data and analyze the third sample data using a statistical model;
[0152] The association table is constructed based on the analysis results.
[0153] The present invention also provides a fault determination device, the fault determination device including a memory, a processor, and a fault determination program stored in the memory and executable on the processor, wherein when the fault determination program is executed by the processor, it implements the various steps of the fault determination method as described in the above embodiments.
[0154] The present invention also provides a computer-readable storage medium storing a fault determination program, which, when executed by a processor, implements the various steps of the fault determination method as described in the above embodiments.
[0155] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0156] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, telecommunications operating system, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, telecommunications operating system, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, telecommunications operating system, article, or apparatus that includes that element.
[0157] Through the above description of the embodiments, those skilled in the art can clearly understand that the telecommunications operation system described above can be implemented by means of software plus necessary general-purpose hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, parking management device, air conditioner, or network device, etc.) to execute the telecommunications operation system described in the various embodiments of the present invention.
[0158] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.
Claims
1. A method for determining a fault, characterized in that, The method for determining the fault includes: The system acquires operational parameter data of the telecommunications operation system during a preset time period, inputs the operational parameter data into the input layer of the prediction model, and obtains the fault type predicted by the prediction model at the target time. The fault type includes the type corresponding to the occurrence of a fault and the type of no fault occurrence. When the fault type is the type corresponding to the occurrence of a fault, the operational parameter data is input into the input layer of the XGBoost classification model to obtain the classification result of whether a fault has occurred output by the XGBoost classification model. When the classification result indicates a fault has occurred, it is determined that the telecommunications operation system has experienced a fault at the target time. Retrieve the staff members corresponding to the fault type from the association table; Obtain the contact information of the aforementioned staff member; The notification information, carrying the fault type and the target time, will be sent to the staff through the aforementioned contact method to notify the staff to handle the situation. Prior to the step of obtaining the staff member corresponding to the fault type from the association table, the following steps are included: Obtain third sample data, analyze the third sample data using the TF-IDF analysis model, analyze the staff corresponding to each fault type using the TF-IDF analysis model, calculate the frequency of staff handling different fault types, and construct an association table.
2. The fault determination method as described in claim 1, characterized in that, Before the step of inputting the operational parameter data into the input layer of the prediction model to obtain the fault type predicted by the prediction model at the target time, the following steps are included: Obtain the first sample data and clean it. The target data for a preset time period is obtained from the first sample data after cleaning to predict the fault type at the preset time. Update the preset time period and obtain the predicted fault type to construct target sample data; The operational parameter data of the target sample data is input into the input layer of the preset long short-term memory model, and the network hierarchy of the long short-term memory model is adjusted. The number of fully connected layers in the long short-term memory model is determined based on the number of fault types, and the activation parameter of the prediction network model is set to softmax. The prediction model is obtained by adjusting the parameters of the long short-term memory model based on the fault types in the first sample data.
3. The fault determination method as described in claim 2, characterized in that, The step of cleaning the first sample data includes: Remove abnormal data from the first sample data and fill in missing values to complete the cleaning of the first sample data.
4. The fault determination method as described in claim 1, characterized in that, The step of inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of whether a fault has occurred output by the XGBoost classification model includes: The operational parameter data is input into the input layer of the XGBoost classification model; The XGBoost classification model, which uses a decision tree as a base classifier, classifies the prediction results of the operational parameter data. The output layer of the XGBoost classification model outputs the classification with the highest score.
5. The fault determination method as described in claim 4, characterized in that, Before the step of inputting the operational parameter data into the input layer of the XGBoost classification model and obtaining the classification result of whether a fault has occurred output by the XGBoost classification model when the fault type is the type corresponding to the occurrence of a fault, the following steps are included: Obtain the second sample data and label the second sample data; The operational parameter data of the labeled second sample data is input into a preset XGBoost classification model for training, thereby obtaining the XGBoost classification model.
6. A fault determination device, characterized in that, The fault determination device includes: The first acquisition module is used to acquire the operation parameter data of the telecommunications operation system during a preset time period, input the operation parameter data into the input layer of the prediction model, and obtain the fault type predicted by the prediction model at the target time. The fault type includes the type corresponding to the occurrence of a fault and the type of no fault occurrence. The second acquisition module is used to input the operation parameter data into the input layer of the XGBoost classification model when the fault type is the type corresponding to the occurrence of a fault, and to obtain the classification result of whether a fault has occurred output by the XGBoost classification model. The determination module is used to determine that the telecommunications operation system has failed at the target time when the classification result indicates that a failure has occurred. Retrieve the staff members corresponding to the fault type from the association table; Obtain the contact information of the aforementioned staff member; The notification information, carrying the fault type and the target time, will be sent to the staff through the aforementioned contact method to notify the staff to handle the situation. Prior to the step of obtaining the staff member corresponding to the fault type from the association table, the following steps are included: Obtain third sample data, analyze the third sample data using the TF-IDF analysis model, analyze the staff corresponding to each fault type using the TF-IDF analysis model, calculate the frequency of staff handling different fault types, and construct an association table.
7. A fault determination device, characterized in that, The fault determination device includes a memory, a processor, and a fault determination program stored in the memory and executable on the processor, wherein the fault determination program, when executed by the processor, implements the steps of the fault determination method as described in any one of claims 1-5.
8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a fault determination program, which, when executed by a processor, implements the steps of the fault determination method as described in any one of claims 1-5.