Method, device and equipment for detecting disturbance work order and storage medium

Abstract

The application discloses a kind of detection methods, device and equipment of disturbance work order and storage medium, involve communication technical field, to solve the problem of poor real-time of manual screening.The method comprises: when generating work order, according to the pre-set data monitoring time length, the N-dimensional monitoring index data related to work order generated until the time when generating work order is obtained;N-dimensional monitoring index data is input into the pre-set disturbance work order detection model, according to the pre-set M-dimensional key feature, the N*M-dimensional key feature data corresponding to N-dimensional monitoring index data is extracted, wherein the key feature is the feature that can reflect the difference between normal state and non-normal state of monitoring index data;N*M-dimensional key feature data is processed by disturbance work order detection model, and the detection result whether work order is disturbance work order is obtained;Wherein, N and M are positive integers, and the values of N and M are set according to the application scene demand of work order.The technical scheme can be applied in wireless network.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, and particularly relates to a detection method and device for a disturbance work order, an equipment and a storage medium. BACKGROUND

[0002] In a wireless network system, a wireless network monitoring center collects wireless performance indexes of each cell according to a pre-set collection frequency, and a wireless network management system monitors the collected data in real time. When the monitoring finds data abnormality, the wireless network management system dispatches a work order according to the data abnormality, and notifies a maintenance personnel of the abnormality through the dispatched work order, so that the maintenance personnel can find and repair problems in time.

[0003] In actual use, the work order dispatched by the wireless network management system often includes a disturbance work order and a non-disturbance work order. The disturbance work order is a work order generated due to some special factors when the network does not have a fault, and the non-disturbance work order is a work order generated when the network has a fault. Due to the existence of the disturbance work order, the maintenance personnel need to identify whether the received work order is a disturbance work order and exclude it. At present, the disturbance work order can only be identified through manual screening. However, the manual screening method consumes a high labor cost, and due to the need for long-time observation and screening, the real-time performance is poor, which cannot meet the timeliness requirement of the network. SUMMARY

[0004] Embodiments of the present application provide a detection method, device, equipment and storage medium for a disturbance work order to solve the problem of poor real-time performance of manual screening.

[0005] In a first aspect, the embodiments of the present application provide a detection method for a disturbance work order, comprising:

[0006] When a work order is generated, N-dimensional monitoring index data related to the work order generated by a time point of generating the work order is acquired according to a pre-set data monitoring time length;

[0007] The N-dimensional monitoring index data is input into a pre-set disturbance work order detection model, N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data is extracted according to pre-set M-dimensional key features, wherein the key features are features capable of distinguishing normal state and abnormal state of the monitoring index data;

[0008] The N*M-dimensional key feature data is processed through the disturbance work order detection model to acquire a detection result of whether the work order is a disturbance work order;

[0009] wherein N and M are positive integers, and the values of N and M are set according to application scene requirements of the work order.

[0010] Optionally, before the step of inputting the N-dimensional monitoring index data into a pre-set disturbance work order detection model and extracting N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to a pre-set M-dimensional key feature when the work order is generated, according to a pre-set data monitoring time length, obtaining N-dimensional monitoring index data generated before the time when the work order is generated, the method further comprises the following steps of:

[0011] If the monitoring index data comprises X sub-monitoring index data, sampling the monitoring index data according to a pre-set data sampling rule to obtain x target sub-monitoring index data, wherein X and x are positive integers, and X≥2, x∈(1,X);

[0012] The step of inputting the N-dimensional monitoring index data into a pre-set disturbance work order detection model and extracting N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to a pre-set M-dimensional key feature comprises the following steps of:

[0013] Inputting the N-dimensional monitoring index data into a pre-set disturbance work order detection model, extracting x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to a pre-set M-dimensional key feature;

[0014] Splicing the x*N*M-dimensional key feature data to obtain target key feature data;

[0015] The step of processing the N*M-dimensional feature data by the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order comprises the following steps of:

[0016] Processing the target key feature data by the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order.

[0017] Optionally, the step of splicing the x*N*M-dimensional key feature data to obtain target key feature data comprises the following steps of:

[0018] According to the time sequence of the x target sub-monitoring index data, performing horizontal splicing on the x*N*M-dimensional key feature data to obtain the x*N*M-dimensional key feature data as the target key feature data; and / or,

[0019] Performing vertical splicing on the x*N*M-dimensional key feature data to obtain N*M-dimensional key feature data as the target key feature data.

[0020] Optionally, the step of processing the N*M-dimensional key feature data by the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order comprises the following steps of:

[0021] By means of the disturbance work order detection model, a LightGBM model is adopted to perform binary classification training on the N*M-dimensional key feature data, so as to obtain a probability value of the work order belonging to a disturbance work order;

[0022] The probability value is compared with a pre-set threshold value, and a detection result of whether the work order is a disturbance work order is obtained according to a comparison result.

[0023] Optionally, after the N*M-dimensional key feature data is processed by means of the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order, the method further comprises:

[0024] According to the detection result, the importance of the N-dimensional monitoring index data is sorted, and a sorting result is obtained;

[0025] According to the sorting result, n-dimensional target monitoring index data is selected from the N-dimensional monitoring index data, wherein n is a positive integer and n

[0026] Then, the N-dimensional monitoring index data is input into a pre-set disturbance work order detection model, and N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data is extracted according to pre-set M-dimensional key features, which are:

[0027] The n-dimensional target monitoring index data is input into a pre-set disturbance work order detection model, and n*M-dimensional key feature data corresponding to the n-dimensional target monitoring index data is extracted according to pre-set M-dimensional key features.

[0028] The N*M-dimensional key feature data is processed by means of the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order, which is:

[0029] The n*M-dimensional key feature data is processed by means of the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order.

[0030] Optionally, the key features include any one or combination of the following:

[0031] Statistical features, fitting features, abnormal features and business features;

[0032] The statistical features are multi-dimensional mathematical statistical features of the monitoring index data; the fitting features are one-dimensional features obtained by using a pre-set fitting algorithm to calculate the mean value of each monitoring index data; the abnormal features are abnormal results obtained according to a pre-set abnormality detection algorithm; and the business features are features obtained from rules or granularity codes corresponding to actual businesses.

[0033] In a second aspect, the embodiments of the present application also provide a detection device for a disturbance work order, comprising:

[0034] A first module is configured to, when a work order is generated, acquire N-dimensional monitoring index data related to the work order generated before a time point of generating the work order according to a pre-set data monitoring time length;

[0035] A second module is configured to input the N-dimensional monitoring index data into a pre-set disturbance work order detection model, extract N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to pre-set M-dimensional key features, wherein the key features are features capable of distinguishing normal state from abnormal state of the monitoring index data;

[0036] A third module is configured to process the N*M-dimensional key feature data through the disturbance work order detection model, and acquire a detection result of whether the work order is a disturbance work order.

[0037] Wherein, N and M are positive integers, and the values of N and M are set according to application scene requirements of the work order.

[0038] Optionally, the detection device for the disturbance work order further comprises:

[0039] A fourth module is configured to, if the monitoring index data comprises X sub-monitoring index data, sample the monitoring index data according to a pre-set data sampling rule, and acquire x target sub-monitoring index data, wherein X and x are positive integers, X≥2, and x∈(1, X).

[0040] The second module is further configured to input the N-dimensional monitoring index data into a pre-set disturbance work order detection model, extract x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to pre-set M-dimensional key features, and perform splicing processing on the x*N*M-dimensional key feature data to acquire target key feature data.

[0041] The third module is further configured to process the target key feature data through the disturbance work order detection model, and acquire a detection result of whether the work order is a disturbance work order.

[0042] Optionally, the second module is further configured to perform horizontal splicing processing on the x*N*M-dimensional key feature data according to a time sequence of generation of the x target sub-monitoring index data, acquire the x*N*M-dimensional key feature data as target key feature data, and / or perform vertical splicing processing on the x*N*M-dimensional key feature data to acquire the N*M-dimensional key feature data as target key feature data.

[0043] Optionally, the third module is further configured to train the N*M-dimensional key feature data by using a LightGBM model to perform binary classification, obtain a probability value of the work order being a disturbance work order, compare the probability value with a pre-set threshold, and obtain a detection result of whether the work order is a disturbance work order according to a comparison result.

[0044] Optionally, the detection device for the disturbance work order further includes:

[0045] The fifth module is configured to sort the importance of the N-dimensional monitoring indicator data according to the detection result, obtain a sorting result, and select n-dimensional target monitoring indicator data from the N-dimensional monitoring indicator data according to the sorting result, where n is a positive integer and n

[0046] The second module is further configured to input the n-dimensional target monitoring indicator data into a pre-set disturbance work order detection model, extract n*M-dimensional key feature data corresponding to the n-dimensional target monitoring indicator data according to pre-set M-dimensional key features.

[0047] The third module is further configured to process the n*M-dimensional key feature data by using the disturbance work order detection model, and obtain a detection result of whether the work order is a disturbance work order.

[0048] In a third aspect, an embodiment of the present application further provides a detection device for a disturbance work order, including a processor and a transceiver.

[0049] The transceiver is configured to, when a work order is generated, obtain N-dimensional monitoring indicator data related to the work order generated within a pre-set data monitoring time length.

[0050] The processor is configured to input the N-dimensional monitoring indicator data into a pre-set disturbance work order detection model, extract N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data according to pre-set M-dimensional key features, process the N*M-dimensional key feature data by using the disturbance work order detection model, and obtain a detection result of whether the work order is a disturbance work order, where the key features are features capable of distinguishing between normal and abnormal states of the monitoring indicator data, and N and M are positive integers, and the values of N and M are set according to application scene requirements of the work order.

[0051] Optionally, the processor is further configured to, if the monitoring indicator data comprises X pieces of sub-monitoring indicator data, sample the monitoring indicator data according to a preset data sampling rule to obtain x pieces of target sub-monitoring indicator data, wherein X and x are positive integers, X is greater than or equal to 2, and x is in the range of (1, X), input the N-dimensional monitoring indicator data into a preset disturbance ticket detection model, extract x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data according to a preset M-dimensional key feature, perform splicing processing on the x*N*M-dimensional key feature data to obtain target key feature data, and process the target key feature data by using the disturbance ticket detection model to obtain a detection result of whether the ticket is a disturbance ticket.

[0052] Optionally, the processor is further configured to perform horizontal splicing processing on the x*N*M-dimensional key feature data according to a time sequence in which the x pieces of target sub-monitoring indicator data are generated, obtain the x*N*M-dimensional key feature data as the target key feature data, and / or perform vertical splicing processing on the x*N*M-dimensional key feature data to obtain N*M-dimensional key feature data as the target key feature data.

[0053] Optionally, the processor is further configured to, by using the disturbance ticket detection model, train the N*M-dimensional key feature data by using a LightGBM model to perform binary classification, obtain a probability value of the ticket belonging to a disturbance ticket, compare the probability value with a preset threshold value, and obtain the detection result of whether the ticket is a disturbance ticket according to a comparison result.

[0054] Optionally, the processor is further configured to, according to the detection result, sort the importance of the N-dimensional monitoring indicator data to obtain a sorting result, select n-dimensional target monitoring indicator data from the N-dimensional monitoring indicator data according to the sorting result, wherein n is a positive integer and n is less than N, input the n-dimensional target monitoring indicator data into a preset disturbance ticket detection model, extract n*M-dimensional key feature data corresponding to the n-dimensional target monitoring indicator data according to a preset M-dimensional key feature, process the n*M-dimensional key feature data by using the disturbance ticket detection model, and obtain a detection result of whether the ticket is a disturbance ticket.

[0055] In a fourth aspect, an embodiment of the present application further provides a terminal device, which comprises a transceiver, a memory, a processor, and a program stored in the memory and executable on the processor, and the processor implements the steps in the detection method of the disturbance ticket when executing the program.

[0056] Fifthly, embodiments of this application also provide a readable storage medium for storing a program, which, when executed by a processor, implements the steps in the disturbance work order detection method described above.

[0057] In this embodiment, when a work order is generated, N-dimensional monitoring indicator data related to the work order up to the time the work order was generated can be obtained according to a pre-set data monitoring duration. Furthermore, the detection result of whether the work order is a disruptive work order is obtained based on the N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data. The entire process is automatically completed by a disruptive work order detection model, achieving real-time detection of disruptive work orders and solving the problem of poor real-time performance in the prior art due to manual screening of disruptive work orders. The detection process for disruptive work orders uses N-dimensional monitoring indicator data related to the work order, as well as M-dimensional key features that distinguish between normal and abnormal states of the monitoring indicator data. This allows the entire detection process to comprehensively analyze multiple important indicators and features that affect the work order, thereby achieving more accurate detection results and avoiding inaccurate results due to a single detection object. Moreover, since both the N-dimensional monitoring indicator data and the M-dimensional key features used are related to the work order, the influence of human factors is reduced, making the technical solution provided in this embodiment more versatile and reusable in various network scenarios, such as 4G and 5G. Attached Figure Description

[0058] Figure 1 This is one of the flowcharts of the disturbance work order detection method provided in the embodiments of this application;

[0059] Figure 2 This is the second flowchart of the disturbance work order detection method provided in the embodiments of this application;

[0060] Figure 3 yes Figure 2 The diagram shown is a schematic of horizontal splicing in the disturbance work order detection method provided in the embodiment of this application;

[0061] Figure 4 yes Figure 2 The diagram shown is a schematic of vertical stitching in the disturbance work order detection method provided in the embodiment of this application;

[0062] Figure 5 This is one of the structural diagrams of the disturbance work order detection device provided in the embodiments of this application;

[0063] Figure 6 This is the second structural diagram of the disturbance work order detection device provided in the embodiments of this application. Detailed Implementation

[0064] The term "and / or" in the embodiments of the present application describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after it.

[0065] The term "multiple" in the embodiments of the present application means two or more, and other quantifiers are similar.

[0066] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0067] Referring to Figure 1 , Figure 1 is a flowchart of the detection method of the disturbance work order provided in the embodiments of the present application, as shown in Figure 1 , comprising the following steps:

[0068] Step 101: When a work order is generated, according to a pre-set data monitoring time length, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated is acquired.

[0069] Step 102: The N-dimensional monitoring index data is input into a pre-set disturbance work order detection model, and N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data is extracted according to pre-set M-dimensional key features, wherein the key features are features that can reflect the difference between the normal state and the abnormal state of the monitoring index data.

[0070] Step 103: The N*M-dimensional key feature data is processed by the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order, wherein N and M are positive integers, and the values of N and M are set according to the application scene requirements of the work order.

[0071] In the embodiment of the present application, when the work order is generated, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated is obtained according to the pre-set data monitoring time length, and the detection result of whether the work order is a disturbance work order is obtained according to the N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data. The whole process is automatically completed by the disturbance work order detection model, which can achieve the purpose of real-time detection of disturbance work orders, and solves the problem that the existing technology uses manual screening of disturbance work orders, resulting in poor real-time performance of the screening operation. The detection process of the disturbance work order uses N-dimensional monitoring index data related to the work order and M-dimensional key features that can reflect the difference between the normal state and the abnormal state of the monitoring index data, so that the whole detection process comprehensively analyzes a plurality of important indicators and features that affect the generation of the work order, thereby achieving the purpose of more accurate detection results and avoiding the problem of inaccurate detection results caused by single detection object. Moreover, since the N-dimensional monitoring index data and the M-dimensional key features used are related to the work order, the influence of human factors is reduced, so that the technical solution provided by the embodiment of the present application has stronger universality and can be reused in various network scenarios such as 4G, 5G, etc.

[0072] Referring to Figure 2 , Figure 2 is a flowchart of the detection method of the disturbance work order provided by the embodiment of the present application, as shown in Figure 2 , comprising the following steps:

[0073] Step 201, when a work order is generated, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated is obtained according to a pre-set data monitoring time length.

[0074] In the embodiment, the unit of the data monitoring time length can be set according to the actual application scene requirement, for example, it can be in units of days, or in units of hours, which will not be described one by one. In order to facilitate the description, in the embodiment and each of the following embodiments, the unit of the data monitoring time length is described in units of days.

[0075] In the embodiment, the data monitoring time length can be set to 1 day. Considering the problem that the accuracy of the monitored data in one day is relatively low, the data monitoring time length can be set to multiple days. In the embodiment, the data monitoring time length is taken as 7 days for example.

[0076] In the embodiment, the N-dimensional monitoring indicator data is obtained according to the pre-set N-dimensional monitoring indicator. Specifically, the N-dimensional monitoring indicator is set according to KPI (Key Performance Indicator) and MR (Measurement Report) and other data associated with the work order. In the embodiment and each of the following embodiments, N is specifically 40 dimensions, and the 40-dimensional monitoring indicator is a key monitoring indicator, for example, the drop rate, the connection rate, and the like, which will not be described one by one. Of course, the above is only a specific example, and in actual use, the N value and the specific monitoring indicator content can be set according to the needs of the application scenario.

[0077] By setting the N-dimensional monitoring indicator, the detection method for the disturbance work order provided in the embodiment has stronger learning ability, and the detection result is more accurate.

[0078] It should be noted that, in the embodiment, the data monitoring duration is 7 days, and the unit of the data monitoring duration is day, so the N-dimensional monitoring indicator data obtained in step 201 includes 7 sub-monitoring indicator data, and the 7 sub-monitoring indicator data correspond to the sub-monitoring indicator data obtained on the first day, the sub-monitoring indicator data obtained on the second day, and the sub-monitoring indicator data obtained on the seventh day, respectively, wherein the sub-monitoring indicator data obtained on the first day is the sub-monitoring indicator data on the day when the work order is generated.

[0079] In step 202, if the monitoring indicator data includes X sub-monitoring indicator data, the monitoring indicator data is sampled according to the pre-set data sampling rule to obtain x target sub-monitoring indicator data, wherein X and x are positive integers, X≥2, and x∈(1,X).

[0080] According to the description of step 201, in the embodiment, the X sub-monitoring indicator data is specifically 7 sub-monitoring indicator data.

[0081] It should be noted that the data sampling rule is not limited in the embodiment, and the data sampling rule can be set according to the requirements of the actual application scene. In order to facilitate understanding, in the embodiment, the data sampling rule is taken as sampling the 1st day, the 3rd day and the 7th day as an example for description. Specifically, the sampling duration of the 1st day is [T-1day-12 hour, T], the sampling duration of the 3rd day is [T-3day-12 hour, T-3day+12 hour], and the sampling duration of the 7th day is [T-7day-12 hour, T-7day+12 hour], wherein T is the ticket generation time. At this time, the x target sub-monitoring indicator data is 3 target sub-monitoring indicator data, which are respectively: the 1st sub-monitoring indicator data, the 3rd sub-monitoring indicator data and the 7th sub-monitoring indicator data.

[0082] In step 203, the N-dimensional monitoring indicator data is input into the pre-set disturbance ticket detection model, and x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data is extracted according to the pre-set M-dimensional key feature.

[0083] The key feature is a feature that can reflect the difference between the normal state and the abnormal state of the monitoring indicator data, wherein the normal state represents the disturbance, and the abnormal state represents the non-disturbance.

[0084] In the embodiment, the key feature is obtained through a visual experiment of each feature of the monitoring indicator data. The key feature can be more clearly and intuitively obtained through the visual experiment. Of course, the above is only an example, and the key feature can also be obtained through other ways in actual use, and each case is not described here.

[0085] In the embodiment, the following four types of key features can be obtained through the visual experiment, which are:

[0086] 1. Statistical feature

[0087] The statistical feature is a multi-dimensional mathematical statistical feature of the monitoring indicator data, including: mean, variance, number of values greater than mean, sliding coefficient of variation, and autocorrelation, etc., a total of 36-dimensional statistical features.

[0088] Table 1 shows the difference between the normal state and the abnormal state of the monitoring indicator data indicated by part of the statistical features:

[0089]

[0090]

[0091] Table 1

[0092] 2. Fitting feature

[0093] The mean value of the monitoring index data of each dimension is obtained by using a preset fitting algorithm.

[0094] In the embodiment, the fitting algorithm can include a moving average, a linear weighted moving average, an exponential weighted moving average, a double exponential weighted moving average, and the like.

[0095] For the disturbance work order, the fitting can eliminate the burr of data fluctuation, and the final mean value is small; for the non-disturbance work order, the fitting still cannot eliminate the burr, and the final mean value is large.

[0096] 3, Abnormal feature

[0097] The abnormal result obtained according to a preset abnormality detection algorithm.

[0098] In the embodiment, the abnormality detection algorithm can include COPOD and IForest, and the intersection of the results obtained according to the two abnormality algorithms can obtain an abnormal point, and the abnormal result is constructed according to the abnormal point, wherein the abnormal result can include a continuous abnormal sequence length, an abnormal degree, and the like.

[0099] For the disturbance work order, the abnormal feature is that the number of abnormal points is small, the abnormal amplitude is small, and the abnormality is recovered quickly, and the non-disturbance work order is opposite.

[0100] 4, Service feature

[0101] The feature obtained according to a rule or granularity code corresponding to an actual service.

[0102] According to the analysis of the above four types of key features, in the embodiment, the M-dimensional key feature can be a 36-dimensional key feature, and it needs to be noted that the 36-dimensional key feature includes all the key features described above, and in the actual use process, only one type or more types can be included, and the embodiment only takes the key feature as an example.

[0103] Due to the use of the above four types of key features, the type of the key feature is more comprehensive and reliable, and the accuracy of the disturbance work order detection result is further ensured.

[0104] At this time, the x*N*M-dimensional key feature data obtained in step 203 is 3*40*36-dimensional key feature data.

[0105] Step 204, splicing the x*N*M-dimensional key feature data to obtain target key feature data.

[0106] In this embodiment, the x*N*M-dimensional key feature data can be spliced in two ways:

[0107] One way is to horizontally splice the x*N*M-dimensional key feature data according to the time sequence of the x target sub-monitoring index data to obtain x*N*M-dimensional key feature data as the target key feature data. Specifically, as shown in Figure 3 the target key feature data is 3*1440-dimensional key feature data.

[0108] The horizontal splicing processing can compare the features of the target sub-monitoring index data in different time periods.

[0109] Another way is to vertically splice the x*N*M-dimensional key feature data to obtain N*M-dimensional key feature data as the target key feature data. Specifically, as shown in Figure 4 the target key feature data is 1440-dimensional key feature data.

[0110] The vertical splicing processing can reflect the overall features of the target sub-monitoring index data.

[0111] In order to make the detection result of the disturbance work order more accurate, in this embodiment, the above two methods can be used to splice the x*N*M-dimensional key feature data, and in actual use process, only one of the methods can be used, which is not limited here.

[0112] Step 205, processing the target key feature data by the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order.

[0113] In this embodiment, step 205 can be implemented by the following steps:

[0114] First, by using the Light Gradient Boosting Machine (LightGBM) model, the target key feature data is trained by the disturbance work order detection model to obtain the probability value of the work order belonging to the disturbance work order.

[0115] Second, compare the probability value with the pre-set threshold value, and obtain the detection result of whether the work order is a disturbance work order according to the comparison result.

[0116] In the embodiment, when setting the threshold value, the recall of the non-disturbance work order is focused on, and it is ensured that all non-disturbance work orders can be identified, so that the accuracy of the disturbance work order can be improved, and the recalled disturbance work order is directly closed. On this basis, it is also ensured that the non-disturbance work order is recalled as much as possible, and balance is made between the two to determine the final threshold value.

[0117] Optionally, in order to avoid overfitting phenomenon caused by too many dimensions of the monitoring index data, after step 205, the method can further include the steps of: according to the detection result, sorting the importance of the N-dimensional monitoring index data to obtain a sorting result; and according to the sorting result, selecting n-dimensional target monitoring index data from the N-dimensional monitoring index data. Wherein, n is a positive integer, and n

[0118] Optionally, in order to further improve the accuracy of the detection result, the n-dimensional target monitoring index data can be further processed according to human experience, and other monitoring index data is selected from (N-n) dimensional monitoring index data for addition.

[0119] Through experimental verification, the n-dimensional target monitoring index data can include the example data shown in Table 2:

[0120]

[0121]

[0122] Table 2

[0123] Thereafter, the n-dimensional target monitoring index data can be used to detect the disturbance work order, and the detection steps are consistent with the steps of detecting the disturbance work order by using the N-dimensional monitoring index data, which will not be described here.

[0124] In the embodiment of the present application, when a work order is generated, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated is obtained according to a pre-set data monitoring time length, and a detection result of whether the work order is a disturbance work order is obtained according to N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data. The whole process is automatically completed by a disturbance work order detection model, which can achieve the purpose of real-time detection of disturbance work orders and solve the problem of poor real-time operation of manual selection of disturbance work orders in the prior art. The detection process of the disturbance work order uses N-dimensional monitoring index data related to the work order and M-dimensional key features that can reflect the difference between the normal state and the abnormal state of the monitoring index data, so that the whole detection process comprehensively analyzes a plurality of important indicators and features that affect the generation of the work order, thereby achieving the purpose of more accurate detection results and avoiding the problem of inaccurate detection results caused by single detection object. In addition, since the N-dimensional monitoring index data and the M-dimensional key features used are related to the work order, the influence of human factors is reduced, so that the technical solution provided by the embodiment of the present application has stronger universality and can be reused in various network scenarios such as 4G and 5G.

[0125] The embodiment of the present application also provides a disturbance work order detection device. Referring to FIG. 5, Figure 5 As shown in FIG. 5, the disturbance work order detection device 500 comprises:

[0126] A first module 501 is configured to, when a work order is generated, obtain N-dimensional monitoring index data related to the work order generated before the time when the work order is generated according to a pre-set data monitoring time length.

[0127] A second module 502 is configured to input the N-dimensional monitoring index data into a pre-set disturbance work order detection model, extract N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to pre-set M-dimensional key features, wherein the key features are features that can reflect the difference between the normal state and the abnormal state of the monitoring index data.

[0128] A third module 503 is configured to process the N*M-dimensional key feature data by using the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order.

[0129] Wherein, N and M are positive integers, and the values of N and M are set according to the application scene requirements of the work order.

[0130] Optionally, the disturbance work order detection device further comprises:

[0131] The fourth module is configured to sample the monitoring indicator data according to a preset data sampling rule to obtain x target sub-monitoring indicator data if the monitoring indicator data comprises X sub-monitoring indicator data, where X and x are positive integers, X is greater than or equal to 2, and x is in the range of 1 to X.

[0132] The second module is further configured to input the N-dimensional monitoring indicator data into a preset disturbance work order detection model, extract x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data according to a preset M-dimensional key feature, perform splicing processing on the x*N*M-dimensional key feature data, and obtain target key feature data.

[0133] The third module is further configured to process the target key feature data by using the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order.

[0134] Optionally, the second module is further configured to perform horizontal splicing processing on the x*N*M-dimensional key feature data according to a time sequence in which the x target sub-monitoring indicator data is generated, obtain the x*N*M-dimensional key feature data as the target key feature data, and / or perform vertical splicing processing on the x*N*M-dimensional key feature data to obtain N*M-dimensional key feature data as the target key feature data.

[0135] Optionally, the third module is further configured to use a LightGBM model to perform binary classification training on the N*M-dimensional key feature data by using the disturbance work order detection model, obtain a probability value of the work order belonging to a disturbance work order, compare the probability value with a preset threshold value, and obtain the detection result of whether the work order is a disturbance work order according to a comparison result.

[0136] Optionally, the disturbance work order detection apparatus further comprises:

[0137] The fifth module is configured to sort the importance of the N-dimensional monitoring indicator data according to the detection result to obtain a sorting result, and select n-dimensional target monitoring indicator data from the N-dimensional monitoring indicator data according to the sorting result, where n is a positive integer and n is less than N.

[0138] The second module is further configured to input the n-dimensional target monitoring indicator data into a preset disturbance work order detection model, and extract n*M-dimensional key feature data corresponding to the n-dimensional target monitoring indicator data according to a preset M-dimensional key feature.

[0139] The third module is further configured to process the n*M-dimensional key feature data by using the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order.

[0140] The specific implementation method of the detection device of the disturbance work order provided in the embodiments of the present application can be referred to the detection method of the disturbance work order provided in the embodiments of the present application, which will not be repeated here.

[0141] In the embodiments of the present application, when a work order is generated, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated can be obtained according to a pre-set data monitoring time length, and a detection result of whether the work order is a disturbance work order can be obtained according to N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data. The whole process is automatically completed through a disturbance work order detection model, which can achieve the purpose of real-time detection of disturbance work orders and solve the problem of poor real-time operation of manual screening of disturbance work orders in the prior art. The detection process of the disturbance work order uses N-dimensional monitoring index data related to the work order and M-dimensional key features that can reflect the difference between the normal state and the abnormal state of the monitoring index data, so that the whole detection process comprehensively analyzes a plurality of important indicators and features that affect the generation of the work order, thereby achieving the purpose of more accurate detection results and avoiding the problem of inaccurate detection results caused by single detection object. Moreover, since the N-dimensional monitoring index data and the M-dimensional key features are both related to the work order, the influence of human factors is reduced, so that the technical solution provided in the embodiments of the present application has stronger universality and can be reused in various network scenarios such as 4G and 5G.

[0142] The embodiments of the present application also provide a detection device of a disturbance work order, which can be referred to Figure 5 as shown in the figure, comprising a processor 601 and a transceiver 602;

[0143] The transceiver 602 is configured to, when a work order is generated, obtain N-dimensional monitoring index data related to the work order generated before the time when the work order is generated according to a pre-set data monitoring time length.

[0144] The processor 601 is configured to input the N-dimensional monitoring index data into a pre-set disturbance work order detection model, extract N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to pre-set M-dimensional key features, wherein the key features are features that can reflect the difference between the normal state and the abnormal state of the monitoring index data, and process the N*M-dimensional key feature data through the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order, wherein N and M are positive integers, and the values of N and M are set according to the application scenario requirements of the work order.

[0145] Optionally, the processor 601 is further configured to, if the monitoring indicator data includes X pieces of sub-monitoring indicator data, sample the monitoring indicator data according to a preset data sampling rule to obtain x pieces of target sub-monitoring indicator data, where X and x are positive integers, X is greater than or equal to 2, and x is an integer between 1 and X, input the N-dimensional monitoring indicator data into a preset disturbance work order detection model, extract x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data according to a preset M-dimensional key feature, perform splicing processing on the x*N*M-dimensional key feature data to obtain target key feature data, and process the target key feature data by using the disturbance work order detection model to obtain a detection result of whether the work order is a disturbance work order.

[0146] Optionally, the processor 601 is further configured to perform horizontal splicing processing on the x*N*M-dimensional key feature data according to a time sequence in which the x pieces of target sub-monitoring indicator data are generated, obtain the x*N*M-dimensional key feature data as target key feature data, and / or perform vertical splicing processing on the x*N*M-dimensional key feature data to obtain N*M-dimensional key feature data as target key feature data.

[0147] Optionally, the processor 601 is further configured to, by using the disturbance work order detection model, train the N*M-dimensional key feature data by using a LightGBM model to perform binary classification, obtain a probability value of the work order belonging to a disturbance work order, compare the probability value with a preset threshold value, and obtain the detection result of whether the work order is a disturbance work order according to a comparison result.

[0148] Optionally, the processor 601 is further configured to, according to the detection result, sort the importance of the N-dimensional monitoring indicator data to obtain a sorting result, select n-dimensional target monitoring indicator data from the N-dimensional monitoring indicator data according to the sorting result, where n is a positive integer and n is less than N, input the n-dimensional target monitoring indicator data into a preset disturbance work order detection model, extract n*M-dimensional key feature data corresponding to the n-dimensional target monitoring indicator data according to a preset M-dimensional key feature, process the n*M-dimensional key feature data by using the disturbance work order detection model, and obtain a detection result of whether the work order is a disturbance work order.

[0149] In the embodiment of the present application, when a work order is generated, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated is obtained according to a pre-set data monitoring time length, and a detection result of whether the work order is a disturbance work order is obtained according to N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data. The whole process is automatically completed through a disturbance work order detection model, which can achieve the purpose of real-time detection of disturbance work orders, and solve the problem that the real-time performance of the screening operation is poor due to manual screening of disturbance work orders in the prior art. The N-dimensional monitoring index data related to the work order and the M-dimensional key feature capable of reflecting the difference between the normal state and the abnormal state of the monitoring index data are used in the detection process of the disturbance work order, so that the whole detection process comprehensively analyzes a plurality of important indexes and features that affect the work order, thereby achieving the purpose of more accurate detection results and avoiding the problem of inaccurate detection results caused by single detection object. In addition, since the N-dimensional monitoring index data and the M-dimensional key feature are both related to the work order, the influence of human factors is reduced, and the technical solution provided by the embodiment of the present application has stronger universality and can be reused in various network scenarios such as 4G, 5G, etc.

[0150] The embodiment of the present application also provides a terminal device, which comprises a transceiver, a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps in the detection method of the disturbance work order are implemented.

[0151] In the embodiment of the present application, when a work order is generated, N-dimensional monitoring index data related to the work order generated before the time when the work order is generated is obtained according to a pre-set data monitoring time length, and a detection result of whether the work order is a disturbance work order is obtained according to N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data. The whole process is automatically completed through a disturbance work order detection model, which can achieve the purpose of real-time detection of disturbance work orders, and solve the problem that the real-time performance of the screening operation is poor due to manual screening of disturbance work orders in the prior art. The N-dimensional monitoring index data related to the work order and the M-dimensional key feature capable of reflecting the difference between the normal state and the abnormal state of the monitoring index data are used in the detection process of the disturbance work order, so that the whole detection process comprehensively analyzes a plurality of important indexes and features that affect the work order, thereby achieving the purpose of more accurate detection results and avoiding the problem of inaccurate detection results caused by single detection object. In addition, since the N-dimensional monitoring index data and the M-dimensional key feature are both related to the work order, the influence of human factors is reduced, and the technical solution provided by the embodiment of the present application has stronger universality and can be reused in various network scenarios such as 4G, 5G, etc.

[0152] The embodiment of the present application further provides a terminal device, comprising a transceiver, a memory, a processor, and a program stored in the memory and executable on the processor, and the processor implements the steps in the method for detecting a disturbed work order when executing the program.

[0153] It should be noted that the division of units in the embodiments of the present application is illustrative, and is only a logical functional division. In actual implementation, another division manner can be used. In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0154] When the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, the integrated unit can be stored in a processor-readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0155] The embodiment of the present application further provides a readable storage medium, and the readable storage medium stores a program. The program is executed by a processor to implement each process of the method for detecting a disturbed work order, and achieves the same technical effect. To avoid repetition, details are not described herein. The readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to a magnetic memory (for example, a floppy disk, a hard disk, a magnetic tape, a magneto-optical disk (MO), etc.), an optical memory (for example, a CD, a DVD, a BD, a HVD, etc.), and a semiconductor memory (for example, a ROM, an EPROM, an EEPROM, a non-volatile memory (NAND FLASH), a solid state disk (SSD), etc.).

[0156] It should be noted that, in the present document, the terms "comprises / comprising" or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises... a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0157] Through the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be realized by means of software and necessary general hardware platforms, of course, they can also be realized by hardware, but in many cases, the former is a better embodiment. According to such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, or an optical disk) and includes a number of instructions for causing a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in the various embodiments of the present application.

[0158] The embodiments of the present application are described above in combination with the accompanying drawings, but the present application is not limited to the above-mentioned specific embodiments, and the above-mentioned specific embodiments are only illustrative and not restrictive. Those skilled in the art can make many forms under the inspiration of the present application without departing from the scope of the present application and the scope protected by the claims.

Claims

1. A method for detecting disturbance work orders, characterized in that, include: When a work order is generated, N-dimensional monitoring indicator data related to the work order are obtained up to the time when the work order was generated, according to the preset data monitoring duration. The N-dimensional monitoring index data is input into a pre-set disturbance work order detection model. Based on the pre-set M-dimensional key features, N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data is extracted. The key features are features that can reflect the difference between the normal state and the abnormal state of the monitoring index data. The N*M dimensional key feature data are processed by the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order; Where N and M are both positive integers, and the values ​​of N and M are set according to the application scenario requirements of the work order.

2. The method according to claim 1, characterized in that, When a work order is generated, after obtaining N-dimensional monitoring indicator data related to the work order up to the time the work order was generated, according to a pre-set data monitoring duration, and before inputting the N-dimensional monitoring indicator data into a pre-set disturbance work order detection model and extracting N*M-dimensional key feature data corresponding to the N-dimensional monitoring indicator data based on pre-set M-dimensional key features, the method further includes: If the monitoring indicator data includes X sub-monitoring indicator data, the monitoring indicator data is sampled according to the pre-set data sampling rules to obtain x target sub-monitoring indicator data, where X and x are both positive integers, and X≥2, x∈(1,X); The step of inputting the N-dimensional monitoring index data into a pre-set disturbance work order detection model, and extracting N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data based on pre-set M-dimensional key features, includes: The N-dimensional monitoring index data is input into a pre-set disturbance work order detection model, and x*N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data is extracted based on the pre-set M-dimensional key features. The key feature data in x*N*M dimensions are concatenated to obtain the target key feature data; The step of processing the N*M dimensional feature data using the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order includes: The target key feature data is processed by the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order.

3. The method according to claim 2, characterized in that, The process of concatenating the x*N*M dimensional key feature data to obtain the target key feature data includes: Based on the time sequence of the generation of the x target sub-monitoring indicator data, the x*N*M dimensional key feature data are horizontally concatenated to obtain x*N*M dimensional key feature data as target key feature data; and / or The x*N*M dimensional key feature data is vertically concatenated to obtain N*M dimensional key feature data as target key feature data.

4. The method according to claim 1, characterized in that, The step of processing the N*M dimensional key feature data using the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order includes: The disturbance work order detection model uses the LightGBM model to perform binary classification training on the N*M dimensional key feature data to obtain the probability value of the work order belonging to the disturbance work order. The probability value is compared with a pre-set threshold, and the detection result of whether the work order is a disturbance work order is obtained based on the comparison result.

5. The method according to claim 1, characterized in that, After processing the N*M dimensional key feature data using the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order, the process further includes: Based on the detection results, the importance of the N-dimensional monitoring indicator data is ranked, and the ranking results are obtained. Based on the sorting results, n-dimensional target monitoring indicator data are selected from the N-dimensional monitoring indicator data, where n is a positive integer and n < N; The process involves inputting the N-dimensional monitoring index data into a pre-set disturbance work order detection model, and extracting the N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data based on pre-set M-dimensional key features: The n-dimensional target monitoring index data is input into the pre-set disturbance work order detection model, and the n*M-dimensional key feature data corresponding to the n-dimensional target monitoring index data is extracted according to the pre-set M-dimensional key features. The process of processing the N*M dimensional key feature data using the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order is as follows: The disturbance work order detection model is used to process the n*M dimensional key feature data to obtain the detection result of whether the work order is a disturbance work order.

6. The method according to any one of claims 1-5, characterized in that, The key features include any one or a combination of the following: Statistical characteristics, fitting characteristics, anomaly characteristics, and business characteristics; Wherein, the statistical features are the multidimensional mathematical statistical features of the monitoring indicator data; the fitting features are the one-dimensional features obtained by calculating the mean of each dimension of the monitoring indicator data using a pre-set fitting algorithm; the anomaly features are the anomaly results obtained according to a pre-set anomaly detection algorithm; and the business features are the features obtained by rules or granular coding corresponding to the actual business.

7. A device for detecting disturbance work orders, characterized in that, include: The first module is used to obtain N-dimensional monitoring index data related to the work order up to the time the work order was generated, according to a pre-set data monitoring duration, when a work order is generated. The second module is used to input the N-dimensional monitoring index data into a pre-set disturbance work order detection model, and extract N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data according to the pre-set M-dimensional key features. The key features are features that can reflect the difference between the normal state and the abnormal state of the monitoring index data. The third module is used to process the N*M dimensional key feature data through the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order; Where N and M are both positive integers, and the values ​​of N and M are set according to the application scenario requirements of the work order.

8. A device for detecting disturbance work orders, characterized in that, include: Processor and transceiver; The transceiver is used to, when a work order is generated, acquire N-dimensional monitoring index data related to the work order up to the time the work order was generated, according to a pre-set data monitoring duration. The processor is used to input the N-dimensional monitoring index data into a pre-set disturbance work order detection model, extract N*M-dimensional key feature data corresponding to the N-dimensional monitoring index data based on pre-set M-dimensional key features, wherein the key features are features that can reflect the difference between the normal and abnormal states of the monitoring index data, and process the N*M-dimensional key feature data through the disturbance work order detection model to obtain the detection result of whether the work order is a disturbance work order, wherein N and M are both positive integers, and the values ​​of N and M are set according to the application scenario requirements of the work order.

9. A terminal device, comprising: A transceiver, a memory, a processor, and a program stored in the memory and executable on the processor; characterized in that the processor is configured to read the program in the memory to implement the steps in the method for detecting disturbance work orders as described in any one of claims 1 to 6.

10. A readable storage medium for storing a program, characterized in that, When the program is executed by the processor, it implements the steps in the method for detecting disturbance work orders as described in any one of claims 1 to 6.

Images