Network disconnection event processing method and device

Abstract

The invention discloses a network disconnection event processing method and device. The method comprises the following steps: acquiring a plurality of passive network disconnection events, and determining a user group influenced by each event; arranging and combining the events to generate a plurality of event sequences; for each sequence, determining a processing strategy, and preferentially arranging repairing personnel to process an earliest repairable event; utilizing the pre-training model to predict the influence of each event on the data traffic of the user during the duration, and calculating the influence score of a single event in combination with the user attribute; summing the influence scores of all events in the sequence to obtain a comprehensive influence score of the sequence; and finally, selecting a strategy corresponding to the sequence with the minimum comprehensive influence score to process all events. According to the invention, the technical problem that the network experience of a user is influenced due to unreasonable arrangement of repair personnel for the network disconnection event is solved.

Description

Technical Field

[0001] The present application relates to the field of communication network technology, and in particular to a method and device for handling network disconnection events. Background Art

[0002] Telecommunications operators ensure stable data transmission by building network infrastructure and transmission lines. However, due to cost or environmental constraints, some networks rely on single-route technology, which can impact user experience if interrupted. Despite measures like redundant backups, equipment upgrades and cutovers can still cause communication interruptions, significantly impacting users, including complaints and network disconnection. To address network outages, operators generate troubleshooting work orders and dispatch them to relevant departments to assess the impact and arrange for repairs. However, this approach is based solely on the type of fault and the number of affected devices, without considering users' network habits and attributes, and thus cannot accurately reflect the actual impact. In reality, the impact of the same outage on different users can vary significantly. For example, differences in time, location, and other factors can lead to significant variations in the impact. Therefore, ignoring these factors can lead to user dissatisfaction and even more serious consequences.

[0003] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0004] The embodiments of the present application provide a method and apparatus for handling network disconnection events, so as to at least solve the technical problem that the user network experience is affected by the unreasonable arrangement of repair personnel for network disconnection events.

[0005] According to one aspect of an embodiment of the present application, a method for processing network disconnection events is provided, including: obtaining multiple passive network disconnection events to be processed, and determining the user group affected by each passive network disconnection event; arranging and combining the multiple passive network disconnection events to obtain multiple passive network disconnection event sequences; for each passive network disconnection event sequence, determining an event processing strategy corresponding to the passive network disconnection event sequence, wherein the event processing strategy is used to sequentially assign a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest to each passive network disconnection event in the passive network disconnection event sequence; for each passive network disconnection event in the passive network disconnection event sequence, using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event processing strategy is executed, and determining an impact score of the passive network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence as the comprehensive impact score corresponding to the passive network disconnection event sequence; and processing the multiple passive network disconnection events according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0006] Optionally, determining an event processing strategy corresponding to a passive network disconnection event sequence includes: determining the position of each passive network disconnection event corresponding to the passive network disconnection event sequence, the initial position of each repair personnel, and the processing time for each repair personnel to handle different passive network disconnection events; initializing an event record table and a time record table for each repair personnel, wherein the event record table is used to record the passive network disconnection events handled by the corresponding repair personnel, and the time record table is used to record the repair completion time of each passive network disconnection event handled by the corresponding repair personnel; looping through the following steps on the passive network disconnection event sequence until the passive network disconnection event sequence is empty: determining the first passive network disconnection event in the passive network disconnection event sequence as the target passive network disconnection event; for each repair personnel, determining the last passive network disconnection event in the event record table corresponding to the repair personnel; The first position of the network disconnection event and the second position of the target passive network disconnection event determine the transfer time of the repair personnel, and the sum of the last repair completion time and the transfer time in the time record table corresponding to the repair personnel is determined as the repair start time of the target passive network disconnection event, wherein, when the event record table corresponding to the repair personnel is empty, the initial position of the repair personnel is taken as the first position; the repair personnel with the smallest corresponding repair start time is determined as the target repair personnel who handles the target passive network disconnection event; the target passive network disconnection event is removed from the passive network disconnection event sequence, and the target passive network disconnection event is added to the event record table corresponding to the target repair personnel, and the sum of the repair start time corresponding to the target repair personnel and the processing time of the target repair personnel handling the target passive network disconnection event is added to the time record table corresponding to the target repair personnel.

[0007] Optionally, a pre-trained data traffic prediction model is used to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event handling strategy is executed, including: for each user in the user group, obtaining user information of the user, wherein the user information includes at least one of the following: user identification information, user address information; determining multiple first time periods corresponding to the duration of the passive network disconnection event when the event handling strategy is executed, and determining an initial first time period corresponding to the moment when the passive network disconnection event occurs, obtaining weather status information for the initial first time period, and obtaining traffic usage data of the user within a preset number of second time periods before the initial first time period, wherein a year is evenly divided into multiple time periods; using the data traffic prediction model to analyze the user information, the initial first time period, the weather status information and the traffic usage data to obtain the sub-data traffic required by the user in each first time period; and determining the data traffic required by the user during the duration of the passive network disconnection event based on each sub-data traffic.

[0008] Optionally, a data traffic prediction model is used to analyze user information, an initial first time period, weather status information, and traffic usage data to obtain the sub-data traffic required by the user in each first time period, including: for each first time period, respectively encoding and splicing the user information, the first time period, weather status information, a preset number of second time periods before the first time period, and / or the traffic usage data within the first time period to obtain a first feature vector corresponding to the first time period; and analyzing the first feature vector using a data traffic prediction model to obtain the sub-data traffic required by the user in the first time period.

[0009] Optionally, the data flow required by the user during the duration of the passive network disconnection event is determined based on each sub-data flow, including: for an initial first time period, determining a first ratio of a first duration from the time the passive network disconnection event occurs to the end time of the initial first time period to a second duration of the initial first time period, and updating the sub-data flow corresponding to the initial first time period by multiplying the sub-data flow corresponding to the initial first time period and the first ratio; for an ending first time period corresponding to a time when the passive network disconnection event is repaired, determining a second ratio of a third duration from the start time of the ending first time period to the time when the passive network disconnection event is repaired to the second duration of the ending first time period, and updating the sub-data flow corresponding to the ending first time period by multiplying the sub-data flow corresponding to the ending first time period and the second ratio; summing the sub-data flows corresponding to all first time periods to obtain the data flow required by the user during the duration of the passive network disconnection event.

[0010] Optionally, the training process of the data traffic prediction model includes: constructing an initial model, wherein the initial model is a fully connected neural network; obtaining user information of multiple users, traffic usage data of each user in multiple consecutive historical time periods, and weather status information for each historical time period; constructing multiple training samples and corresponding sample labels, wherein each training sample includes: user information of a user, traffic usage data of the user in a preset number of historical time periods, a target historical time period after the preset number of historical time periods, and weather status information for the target historical time period, and each sample label includes the traffic usage data of the corresponding user in the corresponding target historical time period; using multiple training samples and corresponding sample labels to iteratively train the initial model to obtain a data traffic prediction model.

[0011] Optionally, the impact score of the passive network disconnection event is determined based on the attribute information and data traffic of each user, including: for each user affected by the passive network disconnection event, obtaining attribute information of multiple dimensions of the user, wherein the dimensions include at least one of the following: user type, consumption level, time online, and fault reporting rate; determining the attribute weight corresponding to the attribute information of each dimension; and determining the total product of multiple attribute weights and the data traffic required by the user during the duration of the passive network disconnection event as the impact score of the passive network disconnection event.

[0012] Optionally, the method also includes: during the processing of multiple passive network disconnection events according to the event processing strategy, if a new passive network disconnection event is received, the new passive network disconnection event and the unrepaired passive network disconnection event are rearranged and combined to obtain multiple new passive network disconnection event sequences; determining the event processing strategy and comprehensive impact score corresponding to each new passive network disconnection event sequence, and processing each passive network disconnection event in the new passive network disconnection event sequence according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0013] According to another aspect of an embodiment of the present application, a method for handling network disconnection events is also provided, including: obtaining multiple active network disconnection events to be executed and a preset execution time period, and determining the user group affected by each active network disconnection event; arranging and combining the multiple active network disconnection events to obtain multiple active network disconnection event sequences; for each active network disconnection event sequence, determining an event handling strategy corresponding to the active network disconnection event sequence, wherein the event handling strategy is used to sequentially determine the execution start time of each active network disconnection event, and the event handling strategy simultaneously satisfies: all active network disconnection events can be completed within the preset execution time period, the impact score of each active network disconnection event is minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when executing the event handling strategy, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence; and handling the multiple active network disconnection events according to the event handling strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0014] Optionally, an event processing strategy corresponding to an active network disconnection event sequence is determined, including: determining the initial position of the repair personnel, the position of each active network disconnection event corresponding to the active network disconnection event sequence, and the processing time of each active network disconnection event; looping the following steps on the active network disconnection event sequence until the active network disconnection event sequence is empty: determining the first active network disconnection event in the active network disconnection event sequence as the target active network disconnection event; determining the sum of the processing time of each active network disconnection event in the active network disconnection event sequence and the sum of the transfer time between the positions of each active network disconnection event as the consumed time, and determining the end time and consumed time of the preset execution time period The difference between the two is the target deadline, wherein the sum of the transfer times of the first cycle includes the transfer time from the initial position of the repair personnel to the position of the target active network disconnection event; the execution completion time of the previous target active network disconnection event executed before the target active network disconnection event is determined as the target start time, wherein the target start time of the first cycle is the start time of the preset execution time period; the execution start time of the target active network disconnection event is determined between the target start time and the target deadline, wherein the impact score corresponding to the target active network disconnection event is the smallest during the duration starting from the execution start time; and the target active network disconnection event is removed from the active network disconnection event sequence.

[0015] According to another aspect of an embodiment of the present application, a network disconnection event processing device is further provided, comprising: a first acquisition module, configured to acquire multiple passive network disconnection events to be processed and determine a user group affected by each passive network disconnection event; a first combination module, configured to arrange and combine the multiple passive network disconnection events to obtain multiple passive network disconnection event sequences; a first analysis module, configured to determine, for each passive network disconnection event sequence, an event processing strategy corresponding to the passive network disconnection event sequence, wherein the event processing strategy is configured to sequentially assign, to each passive network disconnection event in the passive network disconnection event sequence, a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest; for each passive network disconnection event in the passive network disconnection event sequence, using a pre-trained data traffic prediction model to predict, when executing the event processing strategy, the data traffic required by each user in the user group affected by the passive network disconnection event during the duration, and determining an impact score of the passive network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of the multiple passive network disconnection events in the passive network disconnection event sequence as a comprehensive impact score corresponding to the passive network disconnection event sequence; and a first processing module, configured to process the multiple passive network disconnection events according to the event processing strategy corresponding to the passive network disconnection event sequence having the smallest comprehensive impact score.

[0016] According to another aspect of the embodiment of the present application, a network disconnection event processing device is also provided, including: a second acquisition module, used to obtain multiple active network disconnection events to be executed and a preset execution time period, and determine the user group affected by each active network disconnection event; a second combination module, used to arrange and combine multiple active network disconnection events to obtain multiple active network disconnection event sequences; a second analysis module, used to determine, for each active network disconnection event sequence, an event processing strategy corresponding to the active network disconnection event sequence, wherein the event processing strategy is used to determine the execution start time of each active network disconnection event in turn, and the event processing strategy also satisfies: all active network disconnection events can be executed within the preset execution time period The impact score of each active network disconnection event is completed and minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when the event processing strategy is executed, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence; a second processing module is used to process multiple active network disconnection events according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0017] According to another aspect of an embodiment of the present application, a computer program product is further provided, the computer program product comprising: a computer program, wherein when the computer program is executed by a processor, the above-mentioned network disconnection event processing method is implemented.

[0018] According to another aspect of an embodiment of the present application, an electronic device is further provided, comprising: a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the above-mentioned network disconnection event processing method through the computer program.

[0019] In the embodiment of the present application, a data traffic prediction model based on a fully connected neural network is introduced. It integrates user attributes (such as user type, consumption level, etc.), network usage habits and real-time weather information to predict the future traffic demand of individual users affected by network outages. By evaluating the permutations and combinations of multiple passive network outage events, the dynamic programming algorithm can find the optimal solution for the repair personnel to repair the events in the shortest possible time, ensuring that each event can be repaired in the shortest possible time. At the same time, it takes into account the location, transfer time and processing time of the repair personnel, thereby minimizing the overall impact of the network outage on users. During the repair process, once a new network outage event occurs, the algorithm can quickly adjust the repair strategy. Based on the latest location information and processing progress of the existing repair personnel, a repair personnel arrangement plan that includes the new event is immediately planned to ensure the continuous optimization of the repair work. Similar to passive network outage events, active network outage events also use the data traffic prediction model to predict the network demand of the affected user group and evaluate the impact score of the event in combination with user attribute information. However, on this basis, the processing strategy of active events focuses more on optimizing the execution order and time. In the processing scheme of active network outage events, the dynamic programming algorithm not only considers the relative order between events, but also ensures that all events can be completed within the predetermined time period. Through recursion, the algorithm determines the optimal execution start time for each event, in order to minimize the overall impact on users within the execution cycle, thereby solving the technical problem of users' network experience being affected by unreasonable arrangements for repair personnel for network outages. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0021] Figure 1 1 is a flow chart of an optional method for handling a network disconnection event according to an embodiment of the present application;

[0022] Figure 2 is an optional schematic diagram of sub-flow prediction according to an embodiment of the present application;

[0023] Figure 3 is a schematic structural diagram of an optional data flow prediction model according to an embodiment of the present application;

[0024] Figure 4 1 is a flow chart of an optional method for handling a network disconnection event according to an embodiment of the present application;

[0025] Figure 5 1 is a schematic structural diagram of an optional network disconnection event processing device according to an embodiment of the present application;

[0026] Figure 61 is a schematic structural diagram of an optional network disconnection event processing device according to an embodiment of the present application;

[0027] Figure 7 It is a schematic structural diagram of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0029] It should be noted that the terms "first", "second", etc. in the specification, claims, and drawings of the present application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0030] The information collected in the embodiments of the present application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0031] Example 1

[0032] According to an embodiment of the present application, a method for handling a network disconnection event is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] Figure 1 FIG. 1 is a flow chart of a method for handling a network disconnection event according to an embodiment of the present application. Figure 1 As shown, the method includes the following steps:

[0034] Step S102: obtaining multiple passive network disconnection events to be processed, and determining the user group affected by each passive network disconnection event;

[0035] Step S104, performing permutations and combinations on the multiple passive network disconnection events to obtain multiple passive network disconnection event sequences;

[0036] Step S106: For each passive network disconnection event sequence, determine an event handling strategy corresponding to the passive network disconnection event sequence, wherein the event handling strategy is used to sequentially assign to each passive network disconnection event in the passive network disconnection event sequence a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest; for each passive network disconnection event in the passive network disconnection event sequence, use a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event handling strategy is executed, and determine an impact score of the passive network disconnection event based on the attribute information and data traffic of each user; determine the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence as the comprehensive impact score corresponding to the passive network disconnection event sequence;

[0037] Step S108 : processing the multiple passive network disconnection events according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0038] The following describes the steps of the method for handling a network disconnection event in conjunction with a specific implementation process.

[0039] Obtain multiple passive network disconnection events to be processed and determine the user group affected by each passive network disconnection event.

[0040] For example, network outages can be detected in real time through the operator's network monitoring system. Once a network outage is detected, detailed information about the outage is immediately recorded, including but not limited to the time, location, and cause of the outage (such as equipment failure or line damage). Based on the geographic location and network topology of the outage, all users who lost network connectivity due to the outage are identified. This step may involve querying a database for user information and their network access points to determine which users were actually affected. For example, the affected user set can be determined by associating the physical devices affected by the outage with user terminals.

[0041] After obtaining a plurality of passive network disconnection events, the plurality of passive network disconnection events are arranged and combined to obtain a plurality of passive network disconnection event sequences.

[0042] For example, all pending passive network disconnection events are taken as a set. For example, if there are n network disconnection events, the event set is E = {e1, e2, ..., e n}, and permutate and combine them to obtain n! (n factorial) possible passive disconnection event sequences Q. For example: If E = {e1, e2, e3}, then all permutations include 6 combinations: [e1, e2, e3], [e1, e3, e2], [e2, e1, e3], [e2, e3, e1], [e3, e1, e2], [e3, e2, e1]. Each event sequence Q is a list that represents the order in which passive disconnection events are processed. For example, Q = [e1, e2, e3] means that e1 is processed first, then e2, and finally e3.

[0043] After obtaining multiple passive network disconnection event sequences, for each passive network disconnection event sequence, an event processing strategy corresponding to the passive network disconnection event sequence is determined, wherein the event processing strategy is used to sequentially assign a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest to each passive network disconnection event in the passive network disconnection event sequence; for each passive network disconnection event in the passive network disconnection event sequence, a pre-trained data traffic prediction model is used to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event processing strategy is executed, and an impact score of the passive network disconnection event is determined based on the attribute information and data traffic of each user; and the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence is determined as the comprehensive impact score corresponding to the passive network disconnection event sequence.

[0044] As an optional implementation, the following steps may be used to determine the event handling strategy corresponding to the passive network disconnection event sequence:

[0045] Determine the location of each passive network disconnection event in the passive network disconnection event sequence, the initial location of each repair person, and the processing time for each repair person to handle different passive network disconnection events; initialize an event record table and a time record table for each repair person, wherein the event record table is used to record the passive network disconnection events handled by the corresponding repair person, and the time record table is used to record the repair completion time of each passive network disconnection event handled by the corresponding repair person;

[0046] The following steps are performed cyclically on the passive network disconnection event sequence until the passive network disconnection event sequence is empty: the first passive network disconnection event in the passive network disconnection event sequence is determined to be the target passive network disconnection event; for each repair personnel, the transfer time of the repair personnel is determined based on the first position of the last passive network disconnection event in the event record table corresponding to the repair personnel and the second position of the target passive network disconnection event, and the sum of the last repair completion time and the transfer time in the time record table corresponding to the repair personnel is determined to be the repair start time of the target passive network disconnection event, wherein, when the event record table corresponding to the repair personnel is empty, the initial position of the repair personnel is used as the first position; the repair personnel with the smallest corresponding repair start time is determined to be the target repair personnel who handles the target passive network disconnection event; the target passive network disconnection event is removed from the passive network disconnection event sequence, the target passive network disconnection event is added to the event record table corresponding to the target repair personnel, and the sum of the repair start time corresponding to the target repair personnel and the processing time of the target repair personnel handling the target passive network disconnection event is added to the time record table corresponding to the target repair personnel.

[0047] For example, each repairer has two tables, including: event record table ω i : Record the time list processed by the i-th repairer; time record table c i : Record the end time of the i-th repairman handling the passive network disconnection event. When the repair record of the repairman is empty, the event record table is initialized to [O i ] (the initial position of the repairer), the time record table is initialized to [t0] (the current timestamp). The list P composed of the event processing records of each repairer is [w1, w2, ..., w m ], the list of the end time of the events handled by each repairer Rep_t=[c1,c2,…,c m ].

[0048] For each event in the time series Q, the following steps are performed in sequence:

[0049] Determine the target time: take the first event e from the left end of Q as the current target event;

[0050] The initial states of multiplexing P and Rep_t are recorded as:

[0051] Fix the event loop: As long as there are events waiting to be processed in Q, execute the following steps:

[0052] Select the event to be repaired: Take an event (e) from the left end of (Q), which represents the next event in the repair sequence.

[0053] Find the optimal repairer: For each repairer i, calculate the transfer time from the most recently processed event location to e plus the repair time, and find the repairer who can start the repair of e earliest: i = argmin 1≤i≤m (Rep_t ′ [i].right+Tr(P ′ [i].right,e)), where Rep_t ′ [i].right refers to the end time of the event recently handled by repairer i, Tr(P ′ [i].right,e) is the transfer time from the restorer’s most recent event location to event e: P ′ [i].add(e),Rep_t ′ [i].add(Rep_t ′ [i].right+Tr(P ′ [i].right,e)+R(e))

[0054] Update the repair personnel's event record and repair time: assign event e to maintenance personnel i, and update their event record table and repair time list.

[0055] As an optional implementation, a pre-trained data traffic prediction model is used to predict the data traffic required by each user in a user group affected by a passive network disconnection event during the duration when an event handling strategy is executed, including: for each user in the user group, obtaining user information of the user, wherein the user information includes at least one of the following: user identification information, user address information; determining multiple first time periods corresponding to the duration of the passive network disconnection event when the event handling strategy is executed, and determining an initial first time period corresponding to the moment when the passive network disconnection event occurs, obtaining weather status information for the initial first time period, and obtaining traffic usage data of the user within a preset number of second time periods before the initial first time period, wherein a year is evenly divided into multiple time periods; using a data traffic prediction model to analyze user information, the initial first time period, weather status information and traffic usage data to obtain the sub-data traffic required by the user in each first time period; and determining the data traffic required by the user during the duration of the passive network disconnection event based on each sub-data traffic.

[0056] For example, from the user knowledge graph, extract the user information of each user in the affected user group, which includes but is not limited to user identification information (such as user ID) and user address information (such as the latitude and longitude coordinates of the installation location or a detailed address description). The completeness and accuracy of the user information directly affect the accuracy of subsequent predictions. Determine the duration range of the passive network disconnection event and split it into multiple first time periods (for example, each half hour is a time period), and obtain weather status information at the start time of the network disconnection event (initial first time period), including temperature, humidity, wind speed, rainfall, etc. Weather conditions may affect user behavior patterns and thus affect data traffic requirements. Therefore, this part of information is also an important input to the prediction model. Collect the user's data traffic usage in a series of second time periods before the network disconnection event occurs. These time periods should be close to the initial first time period of the network disconnection event so that the model can use recent usage habits to make more accurate predictions. These time periods are close in time, so weather changes are not considered.

[0057] Among them, as an optional implementation method, a data traffic prediction model is used to analyze user information, an initial first time period, weather status information and traffic usage data to obtain the sub-data traffic required by the user in each first time period. The following steps can be taken: for each first time period, the user information, the first time period, weather status information, a preset number of second time periods before the first time period and / or the traffic usage data within the first time period are encoded and spliced to obtain a first feature vector corresponding to the first time period; the first feature vector is analyzed using a data traffic prediction model to obtain the sub-data traffic required by the user in the first time period.

[0058] When building a data traffic prediction model, the encoding layer is responsible for processing and converting various input information into a form that the model can understand. This information includes user ID, user address, current time period, weather information, and the user's recent traffic usage history. The following describes how to encode this information one by one and ultimately splice it into the "first eigenvector" required for model input:

[0059] Convert the user number into a binary vector, such as user number 980 is encoded as 0000000000001111010100 (a total of 22 bits). Use the Word2Vec model to convert the user address into a low-dimensional vector. This vector can reflect the similarity in geographical location and help the model capture the common characteristics of the user group. The user number code and the user address code are spliced as part of the feature representation of the user information; extract the features related to the target prediction time period and splice them into a vector with a dimension of 73 to represent the time period, such as year, month, week number, and weekday. The number of days, whether it is a holiday, and the specific time period number are represented by one-hot encoding. For example, March 12, 2024, at 13:20, is the third month of the year, the third week of March, the second day of the week, a non-holiday, and the 27th time period of the day (divided into half-hour time periods). Therefore, the information can be extracted and converted into vectors [0, 0, 1, 0, 0, 0, 0, 0, 0, 0], [0, 0, 1, 0, 0], [0, 1, 0, 0, 0, 0], [0] and [0, 0, 0, 0, …, 0, 1, 0 …] ∈ R 48 , concatenating these vectors is the information encoding for the current time period; weather status encoding includes z-score standardization of temperature and category coding of air quality, rain and snow conditions. The standardization formula for temperature and wind speed is: Where μ and δ are the mean and standard deviation of temperature (wind speed), respectively. This ensures that all weather-related values are uniformly scaled when input to the model, preventing magnitude differences from affecting the forecast results. Air quality and rain / snow conditions are coded using binary vectors. For example, excellent air quality is coded as 001, and no rain / snow is coded as 000. These codes are concatenated to form the overall weather information vector. The previous time traffic encoding uses the z-score normalization method to calculate the data traffic used in the three time periods preceding the current time period. For example, if the current timestamp is 20:13 and the time interval is half an hour, the first three time periods are [18:30, 19:00), [19:00, 19:30), and [19:30, 20:00]. The data traffic used in each time period is normalized to 50G, 13G, and 5G, respectively, and then concatenated into a vector.

[0060] The user identification, address, weather conditions and historical traffic data are standardized and converted into a model-readable format, which usually includes encoding and normalization operations. A pre-trained data traffic prediction model (DTPM) is used to predict the sub-data traffic demand of each user in each first time period based on the information of each user, the weather conditions of the initial first time period, and the historical traffic usage data.

[0061] The pre-trained data traffic prediction model (DTPM) is used to predict the user's data traffic demand in each first time period. In this step, the DTPM model takes user information, current time period, weather conditions, and previous traffic data as input, and outputs the user's possible data traffic demand (sub-data traffic) in each time period. The formula is expressed as:

[0062] D c (T i )=DTPM(c,T i ,W,[D c (T-1),D c (T-2),…,D c (Tj)]

[0063] Among them, D c (T i ) represents the sub-data traffic required in the first time period i, c is user information, T i Refers to the first time period of i, W is the weather status information, D c (Tj) are the data traffic used in the jth time period before the predicted time period.

[0064] For example, let's analyze a passive network outage event that begins at 2:00 PM on November 20, 2024, and lasts for four hours. The outage is broken down into eight first time periods, each half an hour long. User A's user information includes their ID, address, and customer type. Weather information indicates sunny weather, a temperature of 20°C, and no wind or rain. Historical traffic data shows that user A used 5 GB, 3 GB, and 2 GB of data in the three second time periods before the outage: 1:30 PM to 2:00 PM, 1:00 PM to 1:30 PM, and 12:30 PM to 1:00 PM, respectively. The DTPM model is used to predict user A's sub-data traffic demand in each first time period. Assume the following prediction results are obtained:

[0065] D c (T1)=2GB,D c (T2) = 3GB, D c (T3)=2GB,…,D c (T8) = 1GB

[0066] Among them, as an optional implementation method, the data flow required by the user during the duration of the passive network disconnection event is determined based on each sub-data flow, and the following steps can be taken: for the initial first time period, determine a first ratio of the first duration from the time the passive network disconnection event occurs to the end time of the initial first time period to the second duration of the initial first time period, and use the product of the sub-data flow corresponding to the initial first time period and the first ratio to update the sub-data flow corresponding to the initial first time period; for the ending first time period corresponding to the time when the passive network disconnection event is repaired, determine a second ratio of the third duration from the start time of the ending first time period to the time when the passive network disconnection event is repaired to the second duration of the ending first time period, and use the product of the sub-data flow corresponding to the ending first time period and the second ratio to update the sub-data flow corresponding to the ending first time period; sum the sub-data flows corresponding to all first time periods to obtain the data flow required by the user during the duration of the passive network disconnection event.

[0067] The process can be understood as:

[0068] like Figure 2 As shown, for the initial first time period, the ratio of the first duration from the time the network disconnection event occurs (i.e., the start time of the time period) to the end time of the time period to the second duration of the time period itself is calculated. This means that the traffic usage in some time periods may be suppressed or enhanced by the network disconnection. The formula is expressed as:

[0069]

[0070] Where, P(t1) end represents the end time of the time period, t1 represents the time when the network disconnection event occurs (i.e., the start time of the time period), and δ represents the second duration of the time period itself.

[0071] The predicted sub-data traffic of the initial first time period is adjusted by multiplying the first ratio to reflect the traffic demand of the actual available period. The formula is expressed as:

[0072]

[0073] Where D c (P(t1)) represents the sub-data traffic in the initial first time period.

[0074] For the first time period ending at the time of repair completion, calculate the ratio of the third duration from the start of the time period to the time of repair completion to the second duration of the time period itself. This represents the possible traffic recovery status of users in the remaining time period after the repair is completed. The formula is expressed as:

[0075]

[0076] Where t2 represents the time when the network outage repair is completed, P(t2)start Indicates the start time of the first time period in which the repair is completed.

[0077] The sub-data traffic forecast ending in the first time period is adjusted by multiplying the second ratio, taking into account possible delays or bursts in user traffic usage after the repair. The formula is expressed as:

[0078]

[0079] Where D c (P(t2)) represents the sub-data flow rate within the first time period ending at the moment when the repair is completed.

[0080] The adjusted sub-data traffic in all the first time period is summed to obtain the total data traffic demand of the user during the entire network outage period. The formula is expressed as:

[0081]

[0082] For example, suppose a network outage occurs at 14:15 on November 20, 2024, and the period from 14:00 to 14:30 is the initial first time period. The network outage is repaired at 15:00, so 14:30 to 15:00 is the end of the first time period. For user A, the originally predicted sub-data traffic from 14:00 to 14:30 (the initial first time period P(t1)) is 3GB. Based on the time of the event, the first time period is calculated to be 15 minutes (from 14:15 to 14:30), so the first ratio is Update the sub-data traffic of the initial first time period to: D c (P(t1)) = 3GB × 0.5 = 1.5GB. The originally predicted sub-data traffic from 14:30 to 15:00 (ending the first time period P(t2)) is 4GB. The third time period is calculated to be 30 minutes (from 14:30 to 15:00), so the second ratio is The repair completion time here falls exactly at the end of the time window, resulting in a ratio of 1.0, meaning that the traffic forecast for that period remains unchanged. For the time periods between when the outage takes effect and when the repair is complete, the traffic forecast remains unchanged because it falls outside the outage or repair period.

[0083] As an optional implementation, the training process of the data traffic prediction model can take the following steps: construct an initial model, wherein the initial model is a fully connected neural network; obtain user information of multiple users, traffic usage data of each user in multiple consecutive historical time periods, and weather status information of each historical time period; construct multiple training samples and corresponding sample labels, wherein each training sample includes: user information of a user, traffic usage data of the user in a preset number of historical time periods, a target historical time period after the preset number of historical time periods, and weather status information of the target historical time period, and each sample label includes the traffic usage data of the corresponding user in the corresponding target historical time period; use multiple training samples and corresponding sample labels to iteratively train the initial model to obtain a data traffic prediction model.

[0084] For example, when constructing the initial model, a fully connected neural network is constructed as the initial model. This is based on its powerful feature combination capabilities and is suitable for processing complex and nonlinear datasets, such as user traffic prediction. The input layer is designed to accommodate user identification (after encoding), user address (after Word2Vec encoding), time period characteristics, weather information, and traffic data from the previous period as input. The hidden layer is set up with multiple fully connected layers. The depth can be adjusted based on the actual data complexity and experimental results. Each layer contains a certain number of neurons. The output layer is designed to output predicted traffic. Since traffic prediction is usually a regression task, the output layer may have only one node, and its activation function is linear or ReLU (Rectified Linear Unit). For data preprocessing and sample construction: obtain a large amount of historical data, including user information of multiple users, traffic usage data of each user in a series of historical time periods, and weather status information of the corresponding time period. In a training sample, select a user from the data set, extract the traffic usage data of the user in a preset number (such as 3) of historical time periods, as one of the inputs of the model, determine the target historical time period after this preset number of time periods, and the weather status information of the time period, also as the model input, the label (Ground Truth) is the actual traffic usage data of the user in the target historical time period. For model iterative training: randomly divide the collected training data into training set and validation set. The training set is used for model training, and the validation set is used for model performance evaluation and hyperparameter adjustment. Before each training iteration, the weights and bias values of the model are randomly initialized, which helps to avoid the training falling into a local optimal solution. The training sample is input into the model, and the predicted output is calculated through the forward propagation of the model. The structure of the data traffic prediction model is as follows: Figure 3 shown.

[0085] As an optional implementation method, the impact score of a passive network disconnection event is determined based on the attribute information and data traffic of each user. The following steps can be taken: for each user affected by the passive network disconnection event, the attribute information of the user in multiple dimensions is obtained, where the dimensions include at least one of the following: user type, consumption level, time online, and fault reporting rate; the attribute weight corresponding to the attribute information of each dimension is determined; and the total product of multiple attribute weights and the data traffic required by the user during the duration of the passive network disconnection event is determined as the impact score of the passive network disconnection event.

[0086] It can be understood that when assessing the impact of passive disconnection events on users, it is necessary not only to consider the loss of data traffic but also to combine various user attributes to more comprehensively understand the consequences of each disconnection event. The following explains how to quantify the impact score of passive disconnection events based on users' specific attributes and traffic requirements:

[0087] For each user affected by a passive network outage, we collect attribute information from multiple dimensions, including but not limited to: user type (such as public users, government and enterprise users, or emergency users); consumption level (reflecting the user's package level or average monthly consumption); online time (the duration of the user's use of network services, which can reflect their loyalty and stability); fault reporting rate (the frequency of fault reports by users in the past period of time, which indirectly reflects their sensitivity to service quality). This attribute information can be obtained from the user knowledge graph or through database queries. Based on business experience and analysis, we determine the weight of each attribute dimension to reflect the importance of different attributes in assessing the impact of network outages. For example, the impact of network outages on government and enterprise users may be given a higher weight because their business operations are highly dependent on stable network connections. The formula is expressed as follows:

[0088]

[0089] Where, F e (t1, t2) represents the impact score of the passive network disconnection event, n represents the total number of users affected by the passive network disconnection event, P i1 is the user type weight of the i-th user affected by the passive network disconnection event, P i2 is the consumption level weight of the i-th user affected by the passive network disconnection event, P i3 is the online time weight of the i-th user affected by the passive disconnection event, P i4 is the fault reporting rate weight of the i-th user affected by the passive network disconnection event.

[0090] Generally speaking, compared with public users, government and enterprise users rely more on stable networks for business operations, and the impact of network outages is usually greater. Therefore, the P1 of government and enterprise customers (such as 1.5) is larger than the P1 of public customers (such as 0.5), and the P1 of emergency users is the largest. Assign different weights to users according to their consumption level. Users of higher-end packages often have higher expectations for network services, and the impact of network outages is greater. Assign different weights based on the user's online time. For example, users who are approaching the end of the agreement need to have a greater user experience. The fault reporting rate can reflect the user's tolerance for network outages to a certain extent. According to experience, if the user reporting rate (the probability of the user reporting a fault when a network outage occurs) is low, it means that the network outage does not have a great impact on the user. The impact of the network outage is measured by assigning different weights to users affected by the network outage and then calculating the weighted traffic they will use. F e The larger (t1, t2) is, the greater the impact of the network disconnection event e during the period from t1 to t2, and the more attention it should receive.

[0091] As an optional implementation method, during the processing of multiple passive network disconnection events according to the event processing strategy, if a new passive network disconnection event is received, the new passive network disconnection event and the unrepaired passive network disconnection events are rearranged and combined to obtain multiple new passive network disconnection event sequences; the event processing strategy and comprehensive impact score corresponding to each new passive network disconnection event sequence are determined, and each passive network disconnection event in the new passive network disconnection event sequence is processed according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0092] After obtaining the comprehensive impact score, multiple passive network disconnection events are processed according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0093] In an embodiment of the present application, a data traffic prediction model based on a fully connected neural network is introduced to comprehensively consider user attributes (such as user type, consumption level, etc.), network usage habits and real-time weather information to predict the future traffic demand of individual users affected by network outages. By evaluating the permutations and combinations of multiple passive network outage events, the dynamic programming algorithm can find the optimal solution for the repair sequence of events by repair personnel, ensuring that each event can be repaired in the shortest time, while taking into account the location, transfer time and processing time of the repair personnel, thereby minimizing the overall impact of the network outage event on the user. During the repair process, once a new network outage event occurs, the algorithm can quickly adjust the repair strategy and immediately plan a repair personnel arrangement plan including the new event based on the latest location information and processing progress of the existing repair personnel, thereby ensuring the continuous optimization of the repair work, thereby solving the technical problem of the user's network experience being affected by the unreasonable arrangement of repair personnel for network outage events.

[0094] Example 2

[0095] According to an embodiment of the present application, a method for handling a network disconnection event is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0096] Figure 4 FIG. 1 is a flow chart of a method for handling a network disconnection event according to an embodiment of the present application. Figure 4 As shown, the method includes the following steps:

[0097] Step S402: obtaining multiple active network disconnection events to be executed and preset execution time periods, and determining the user groups affected by each active network disconnection event;

[0098] Step S404, performing permutations and combinations on the multiple active network disconnection events to obtain multiple active network disconnection event sequences;

[0099] Step S406: For each active network disconnection event sequence, determine an event handling strategy corresponding to the active network disconnection event sequence, wherein the event handling strategy is used to sequentially determine the execution start time of each active network disconnection event, and the event handling strategy simultaneously satisfies: all active network disconnection events can be completed within a preset execution time period, and the impact score of each active network disconnection event is minimized. The process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when the event handling strategy is executed, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; and determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence;

[0100] Step S408 : Processing multiple active network disconnection events according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0101] The following describes the steps of the method for handling a network disconnection event in conjunction with a specific implementation process.

[0102] Obtain multiple active network disconnection events to be executed and preset execution time periods, and determine the user group affected by each active network disconnection event.

[0103] First, identify and collect information on all proactive network outages planned for a pre-defined timeframe, and determine the user groups each event will impact. This phase requires a detailed analysis of the scope of the network upgrade or cutover to ensure all potentially affected users are considered.

[0104] After obtaining multiple active network disconnection events, the multiple active network disconnection events are arranged and combined to obtain multiple active network disconnection event sequences.

[0105] Next, the collected active disconnection events are permuted and combined in various ways to form multiple different event sequences. Each sequence represents a potential execution order for the disconnection events, which is the basis for finding the optimal solution.

[0106] After obtaining multiple active network disconnection event sequences, for each active network disconnection event sequence, an event processing strategy corresponding to the active network disconnection event sequence is determined, wherein the event processing strategy is used to determine the execution start time of each active network disconnection event in turn, and the event processing strategy simultaneously satisfies: all active network disconnection events can be completed within a preset execution time period, the impact score of each active network disconnection event is minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when executing the event processing strategy, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence.

[0107] Before evaluating the event sequence, it is necessary to clearly identify the initial location of the repair personnel, the location of each active disconnection event, and the duration required to handle it. This helps calculate the transition time between events and the duration of the entire sequence, ensuring that all events can be completed within the preset time period.

[0108] As an optional implementation, the event handling strategy corresponding to the active network disconnection event sequence may be determined by taking the following steps:

[0109] Determine the initial location of the repair personnel, the location of each active network disconnection event corresponding to the active network disconnection event sequence, and the processing time of each active network disconnection event;

[0110] The following steps are executed for the active network disconnection event sequence in a loop until the active network disconnection event sequence is empty:

[0111] Determine the first active network disconnection event in the active network disconnection event sequence as the target active network disconnection event;

[0112] Determine the sum of the processing time of each active network disconnection event in the active network disconnection event sequence and the sum of the transfer time between the locations of each active network disconnection event as the consumed time, and determine the difference between the end time of the preset execution time period and the consumed time as the target end time, wherein the sum of the transfer time in the first cycle includes the transfer time from the initial location of the repair personnel to the location of the target active network disconnection event;

[0113] Determine the completion time of the previous target active network disconnection event executed before executing the target active network disconnection event as the target start time, wherein the target start time of the first cycle is the start time of the preset execution time period;

[0114] Determining the execution start time of the target active network disconnection event between the target start time and the target end time, wherein the impact score corresponding to the target active network disconnection event is the smallest during the duration starting from the execution start time;

[0115] Remove the target active network disconnection event from the active network disconnection event sequence.

[0116] Active disconnection event sequence Q=(q 1, q 2, …,q n )(fixed order, no need for permutation and combination), preset time period [t start ,t end ], initial position O, processing time P(q i ), transfer time Tr(q i, q j, ). Define variables

[0117] For the first event q1, the feasible time t satisfies:

[0118]

[0119] For the i-th event, calculate the remaining consumption time:

[0120]

[0121] Feasible time window:

[0122] t∈[t prev_end ,t end -Remaining time]

[0123] Among them, t prev_end =prev t +P(q i-1 )+Tr(q i-1, q i, ), prev t q i-1 Execution time.

[0124] When determining the impact score of active disconnection events based on individual user attributes and data traffic, it can be understood that when assessing the impact of passive disconnection events on users, it is necessary not only to consider the loss of data traffic but also to combine multiple user attributes to more comprehensively understand the consequences of each disconnection event. The following explains how to quantify the impact score of passive disconnection events based on specific user attributes and traffic requirements:

[0125] For each user affected by a passive network outage, we collect attribute information from multiple dimensions, including but not limited to: user type (such as public users, government and enterprise users, or emergency users); consumption level (reflecting the user's package level or average monthly consumption); online time (the duration of the user's use of network services, which can reflect their loyalty and stability); fault reporting rate (the frequency of fault reports by users in the past period of time, which indirectly reflects their sensitivity to service quality). This attribute information can be obtained from the user knowledge graph or through database queries. Based on business experience and analysis, we determine the weight of each attribute dimension to reflect the importance of different attributes in assessing the impact of network outages. For example, the impact of network outages on government and enterprise users may be given a higher weight because their business operations are highly dependent on stable network connections. The formula is expressed as follows:

[0126]

[0127] Where, F e (t1, t2) represents the impact score of the passive network disconnection event, n represents the total number of users affected by the passive network disconnection event, P i1 is the user type weight of the i-th user affected by the passive network disconnection event, P i2 is the consumption level weight of the i-th user affected by the passive network disconnection event, P i3 is the online time weight of the i-th user affected by the passive disconnection event, P i4 is the fault reporting rate weight of the i-th user affected by the passive network disconnection event.

[0128] Generally speaking, compared with public users, government and enterprise users rely more on stable networks for business operations, and the impact of network outages is usually greater. Therefore, the P1 of government and enterprise customers (such as 1.5) is larger than the P1 of public customers (such as 0.5), and the P1 of emergency users is the largest. Assign different weights to users according to their consumption level. Users of higher-end packages often have higher expectations for network services, and the impact of network outages is greater. Assign different weights based on the user's online time. For example, users who are approaching the end of the agreement need to have a greater user experience. The fault reporting rate can reflect the user's tolerance for network outages to a certain extent. According to experience, if the user reporting rate (the probability of the user reporting a fault when a network outage occurs) is low, it means that the network outage does not have a great impact on the user. The impact of the network outage is measured by assigning different weights to users affected by the network outage and then calculating the weighted traffic they will use. F e The larger (t1, t2) is, the greater the impact of the network disconnection event e during the period from t1 to t2, and the more attention it should receive.

[0129] After obtaining multiple comprehensive impact scores, the multiple active network disconnection events are processed according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0130] For example, suppose worker A needs to execute four active disconnection events (e1 to e4) within the time limit [22:00, 24:00]. These events affect different user groups. Generate all possible execution sequence combinations, such as e1-e2-e3-e4, e1-e3-e2-e4, and other 24 possible arrangements. Select e1-e2-e3-e4 as an example sequence for analysis. Assume that it takes 10 minutes for A to move from its initial position to e1, and 15 minutes to process e1. It takes 30 minutes for A to move from e1 to e2, and 10 minutes to process e2. And so on.

[0131] Determine the target event: e1 becomes the target of the first loop.

[0132] Calculate the elapsed time: For the first cycle, consider the transfer time from the initial position A to e1 (10 minutes) plus the processing time of e1 (15 minutes).

[0133] Target deadline: 24:00 - (10 + 15 + 30 + 10 + ...) = latest start time.

[0134] Target start time: 22:00 for the first cycle.

[0135] Optimize the start time: Find the optimal start time for e1 between 22:00 and the latest start time to ensure that it has the least impact on the score.

[0136] Sequence update: Remove e1 and continue the same process for the remaining events (e2-e3-e4) until the sequence is empty.

[0137] After completing the above cyclic evaluation for all 24 sequences, the comprehensive impact score of each sequence is compared, and the sequence with the lowest score (least impact) is selected as the final event handling strategy.

[0138] In the embodiment of the present application, the active network disconnection event also uses the data traffic prediction model to predict the network demand of the affected user group, and combines the user attribute information to evaluate the impact score of the event. However, on this basis, the processing strategy of the active event focuses more on the optimization of the execution order and time. In the processing scheme of the active network disconnection event, the dynamic programming algorithm not only considers the relative order between events, but also ensures that all events can be completed within the predetermined time period. In a recursive manner, the algorithm determines the optimal execution start time for each event, in order to minimize the overall impact on users within the execution cycle, thereby solving the technical problem of the user's network experience being affected by the unreasonable arrangement of repair personnel for the network disconnection event.

[0139] Example 3

[0140] According to an embodiment of the present application, a network disconnection event processing device for implementing the network disconnection event processing method in embodiment 1 is also provided. Figure 5 As shown, the network disconnection event processing device at least includes: a first acquisition module 51, a first combination module 52, a first analysis module 53 and a first processing module 54, wherein:

[0141] A first acquisition module 51 is configured to acquire multiple passive network disconnection events to be processed and determine a user group affected by each passive network disconnection event;

[0142] A first combining module 52 is configured to arrange and combine multiple passive network disconnection events to obtain multiple passive network disconnection event sequences;

[0143] The first analysis module 53 is configured to determine, for each passive network disconnection event sequence, an event handling strategy corresponding to the passive network disconnection event sequence, wherein the event handling strategy is configured to sequentially assign, to each passive network disconnection event in the passive network disconnection event sequence, a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest; for each passive network disconnection event in the passive network disconnection event sequence, using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration of the event handling strategy, and determine an impact score for the passive network disconnection event based on the attribute information and data traffic of each user; and determine the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence as a comprehensive impact score corresponding to the passive network disconnection event sequence;

[0144] The first processing module 54 is configured to process multiple passive network disconnection events according to an event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0145] The functions of each module of the network disconnection event processing device are described below in conjunction with a specific implementation process.

[0146] The first acquisition module acquires multiple passive network disconnection events to be processed and determines the user group affected by each passive network disconnection event.

[0147] After obtaining a plurality of passive network disconnection events, the first combination module arranges and combines the plurality of passive network disconnection events to obtain a plurality of passive network disconnection event sequences.

[0148] After obtaining multiple passive network disconnection event sequences, the first analysis module determines, for each passive network disconnection event sequence, an event processing strategy corresponding to the passive network disconnection event sequence, wherein the event processing strategy is used to sequentially assign to each passive network disconnection event in the passive network disconnection event sequence a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest; for each passive network disconnection event in the passive network disconnection event sequence, a pre-trained data traffic prediction model is used to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event processing strategy is executed, and an impact score of the passive network disconnection event is determined based on the attribute information and data traffic of each user; and the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence is determined as the comprehensive impact score corresponding to the passive network disconnection event sequence.

[0149] As an optional implementation method, the event processing strategy corresponding to the passive network disconnection event sequence can be determined by the following steps: determining the position of each passive network disconnection event corresponding to the passive network disconnection event sequence, the initial position of each repair personnel, and the processing time of each repair personnel for handling different passive network disconnection events; initializing an event record table and a time record table for each repair personnel, wherein the event record table is used to record the passive network disconnection events handled by the corresponding repair personnel, and the time record table is used to record the repair completion time of each passive network disconnection event handled by the corresponding repair personnel; looping through the following steps on the passive network disconnection event sequence until the passive network disconnection event sequence is empty: determining the first passive network disconnection event in the passive network disconnection event sequence as the target passive network disconnection event; for each repair personnel, according to the event record table corresponding to the repair personnel, The first position of the last passive network disconnection event and the second position of the target passive network disconnection event are used to determine the transfer time of the repair personnel, and the sum of the last repair completion time and the transfer time in the time record table corresponding to the repair personnel is determined as the repair start time of the target passive network disconnection event, wherein, when the event record table corresponding to the repair personnel is empty, the initial position of the repair personnel is used as the first position; the repair personnel with the smallest corresponding repair start time is determined as the target repair personnel who handles the target passive network disconnection event; the target passive network disconnection event is removed from the passive network disconnection event sequence, and the target passive network disconnection event is added to the event record table corresponding to the target repair personnel, and the sum of the repair start time corresponding to the target repair personnel and the processing time of the target repair personnel handling the target passive network disconnection event is added to the time record table corresponding to the target repair personnel.

[0150] As an optional implementation, a pre-trained data traffic prediction model is used to predict the data traffic required by each user in a user group affected by a passive network disconnection event during the duration when an event handling strategy is executed, including: for each user in the user group, obtaining user information of the user, wherein the user information includes at least one of the following: user identification information, user address information; determining multiple first time periods corresponding to the duration of the passive network disconnection event when the event handling strategy is executed, and determining an initial first time period corresponding to the moment when the passive network disconnection event occurs, obtaining weather status information for the initial first time period, and obtaining traffic usage data of the user within a preset number of second time periods before the initial first time period, wherein a year is evenly divided into multiple time periods; using a data traffic prediction model to analyze user information, the initial first time period, weather status information and traffic usage data to obtain the sub-data traffic required by the user in each first time period; and determining the data traffic required by the user during the duration of the passive network disconnection event based on each sub-data traffic.

[0151] Among them, as an optional implementation method, a data traffic prediction model is used to analyze user information, an initial first time period, weather status information and traffic usage data to obtain the sub-data traffic required by the user in each first time period. The following steps can be taken: for each first time period, the user information, the first time period, weather status information, a preset number of second time periods before the first time period and / or the traffic usage data within the first time period are encoded and spliced to obtain a first feature vector corresponding to the first time period; the first feature vector is analyzed using a data traffic prediction model to obtain the sub-data traffic required by the user in the first time period.

[0152] Among them, as an optional implementation method, the data flow required by the user during the duration of the passive network disconnection event is determined based on each sub-data flow, and the following steps can be taken: for the initial first time period, determine a first ratio of the first duration from the time the passive network disconnection event occurs to the end time of the initial first time period to the second duration of the initial first time period, and use the product of the sub-data flow corresponding to the initial first time period and the first ratio to update the sub-data flow corresponding to the initial first time period; for the ending first time period corresponding to the time when the passive network disconnection event is repaired, determine a second ratio of the third duration from the start time of the ending first time period to the time when the passive network disconnection event is repaired to the second duration of the ending first time period, and use the product of the sub-data flow corresponding to the ending first time period and the second ratio to update the sub-data flow corresponding to the ending first time period; sum the sub-data flows corresponding to all first time periods to obtain the data flow required by the user during the duration of the passive network disconnection event.

[0153] As an optional implementation, the training process of the data traffic prediction model can take the following steps: construct an initial model, wherein the initial model is a fully connected neural network; obtain user information of multiple users, traffic usage data of each user in multiple consecutive historical time periods, and weather status information of each historical time period; construct multiple training samples and corresponding sample labels, wherein each training sample includes: user information of a user, traffic usage data of the user in a preset number of historical time periods, a target historical time period after the preset number of historical time periods, and weather status information of the target historical time period, and each sample label includes the traffic usage data of the corresponding user in the corresponding target historical time period; use multiple training samples and corresponding sample labels to iteratively train the initial model to obtain a data traffic prediction model.

[0154] As an optional implementation method, the impact score of a passive network disconnection event is determined based on the attribute information and data traffic of each user. The following steps can be taken: for each user affected by the passive network disconnection event, the attribute information of the user in multiple dimensions is obtained, where the dimensions include at least one of the following: user type, consumption level, time online, and fault reporting rate; the attribute weight corresponding to the attribute information of each dimension is determined; and the total product of multiple attribute weights and the data traffic required by the user during the duration of the passive network disconnection event is determined as the impact score of the passive network disconnection event.

[0155] As an optional implementation method, during the processing of multiple passive network disconnection events according to the event processing strategy, if a new passive network disconnection event is received, the new passive network disconnection event and the unrepaired passive network disconnection events are rearranged and combined to obtain multiple new passive network disconnection event sequences; the event processing strategy and comprehensive impact score corresponding to each new passive network disconnection event sequence are determined, and each passive network disconnection event in the new passive network disconnection event sequence is processed according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0156] After obtaining the comprehensive impact score, the first processing module processes the multiple passive network disconnection events according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0157] It should be noted that each module in the network disconnection event processing device in the embodiment of the present application corresponds one-to-one to each implementation step of the network disconnection event processing method in Example 1. Since a detailed description has been given in Example 1, some details not reflected in this embodiment can be referred to Example 1 and will not be elaborated here.

[0158] Example 4

[0159] According to an embodiment of the present application, a network disconnection event processing device for implementing the network disconnection event processing method in Example 2 is also provided. Figure 6 As shown, the network disconnection event processing device at least includes: a second acquisition module 61, a second combination module 62, a second analysis module 63 and a second processing module 64, wherein:

[0160] A second acquisition module 61 is configured to acquire multiple active network disconnection events to be executed and preset execution time periods, and determine the user groups affected by each active network disconnection event;

[0161] A second combining module 62 is configured to arrange and combine multiple active network disconnection events to obtain multiple active network disconnection event sequences;

[0162] A second analysis module 63 is configured to determine, for each active network disconnection event sequence, an event handling strategy corresponding to the active network disconnection event sequence, wherein the event handling strategy is configured to sequentially determine the execution start time of each active network disconnection event, and the event handling strategy simultaneously satisfies the following requirements: all active network disconnection events can be completed within a preset execution time period, and the impact score of each active network disconnection event is minimized. Furthermore, the impact score of each active network disconnection event is determined by: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration of the event handling strategy, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; and determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence.

[0163] The second processing module 64 is configured to process multiple active network disconnection events according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0164] The functions of each module of the network disconnection event processing device are described below in conjunction with a specific implementation process.

[0165] The second acquisition module acquires multiple active network disconnection events to be executed and preset execution time periods, and determines the user group affected by each active network disconnection event.

[0166] After obtaining multiple active network disconnection events, the second combination module arranges and combines the multiple active network disconnection events to obtain multiple active network disconnection event sequences.

[0167] After obtaining multiple active network disconnection event sequences, the second analysis module determines the event processing strategy corresponding to each active network disconnection event sequence, wherein the event processing strategy is used to determine the execution start time of each active network disconnection event in turn, and the event processing strategy simultaneously satisfies: all active network disconnection events can be completed within a preset execution time period, the impact score of each active network disconnection event is minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when executing the event processing strategy, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence.

[0168] As an optional implementation, the event handling strategy corresponding to the active network disconnection event sequence may be determined by taking the following steps:

[0169] Determine the initial location of the repair personnel, the location of each active network disconnection event corresponding to the active network disconnection event sequence, and the processing time of each active network disconnection event;

[0170] The following steps are executed for the active network disconnection event sequence in a loop until the active network disconnection event sequence is empty:

[0171] Determine the first active network disconnection event in the active network disconnection event sequence as the target active network disconnection event;

[0172] Determine the sum of the processing time of each active network disconnection event in the active network disconnection event sequence and the sum of the transfer time between the locations of each active network disconnection event as the consumed time, and determine the difference between the end time of the preset execution time period and the consumed time as the target end time, wherein the sum of the transfer time in the first cycle includes the transfer time from the initial location of the repair personnel to the location of the target active network disconnection event;

[0173] Determine the completion time of the previous target active network disconnection event executed before executing the target active network disconnection event as the target start time, wherein the target start time of the first cycle is the start time of the preset execution time period;

[0174] Determining the execution start time of the target active network disconnection event between the target start time and the target end time, wherein the impact score corresponding to the target active network disconnection event is the smallest during the duration starting from the execution start time;

[0175] Remove the target active network disconnection event from the active network disconnection event sequence.

[0176] After obtaining a plurality of comprehensive impact scores, the second processing module processes the plurality of active network disconnection events according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0177] It should be noted that each module in the network disconnection event processing device in the embodiment of the present application corresponds one-to-one to each implementation step of the network disconnection event processing method in Example 2. Since a detailed description has been given in Example 2, some details not reflected in this embodiment can be referred to Example 2 and will not be elaborated here.

[0178] Example 5

[0179] According to an embodiment of the present application, a computer program product is further provided, which includes a computer program, wherein when the computer program is executed by a processor, the network disconnection event processing method in Example 1 or Example 2 is implemented.

[0180] According to an embodiment of the present application, a non-volatile storage medium is also provided, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the network disconnection event processing method in Example 1 or Example 2 by running the computer program.

[0181] According to an embodiment of the present application, a processor is further provided, which is used to run a computer program, wherein the network disconnection event processing method in Example 1 or Example 2 is executed when the computer program is running.

[0182] According to an embodiment of the present application, an electronic device is also provided, which includes: a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the network disconnection event processing method in Example 1 or Example 2 through the computer program.

[0183] Specifically, when the computer program is running, the following steps can be executed: obtaining multiple passive network disconnection events to be processed and determining the user group affected by each passive network disconnection event; arranging and combining the multiple passive network disconnection events to obtain multiple passive network disconnection event sequences; for each passive network disconnection event sequence, determining an event processing strategy corresponding to the passive network disconnection event sequence, wherein the event processing strategy is used to sequentially assign a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest to each passive network disconnection event in the passive network disconnection event sequence; for each passive network disconnection event in the passive network disconnection event sequence, using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event processing strategy is executed, and determining the impact score of the passive network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence as the comprehensive impact score corresponding to the passive network disconnection event sequence; and processing multiple passive network disconnection events according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

[0184] Specifically, when the computer program is running, the following steps can be executed: obtaining multiple active network disconnection events to be executed and a preset execution time period, and determining the user group affected by each active network disconnection event; arranging and combining the multiple active network disconnection events to obtain multiple active network disconnection event sequences; for each active network disconnection event sequence, determining an event processing strategy corresponding to the active network disconnection event sequence, wherein the event processing strategy is used to sequentially determine the execution start time of each active network disconnection event, and the event processing strategy simultaneously satisfies: all active network disconnection events can be completed within the preset execution time period, the impact score of each active network disconnection event is minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when executing the event processing strategy, and determining the impact score of the active network disconnection event based on the attribute information and data traffic of each user; determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence; and processing the multiple active network disconnection events according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

[0185] As an optional implementation, the electronic device may be in the form of a mobile terminal, a computer terminal or a similar computing device. Figure 7 FIG1 shows a hardware structure block diagram of an electronic device for implementing a method for handling network disconnection events. Figure 7 As shown, the electronic device 70 may include one or more (shown as 702a, 702b, ..., 702n in the figure) processors 702 (the processor 702 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 704 for storing data, and a transmission device 706 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It will be understood by those skilled in the art that Figure 7 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 7 More or fewer components than shown, or with Figure 7 Different configurations shown.

[0186] It should be noted that the one or more processors 702 and / or other data processing circuits described above may generally be referred to herein as "data processing circuitry". The data processing circuitry may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuitry may be a single independent processing module, or may be incorporated in whole or in part into any of the other components of the electronic device 70. As described in the embodiments of the present application, the data processing circuitry serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0187] The memory 704 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the network disconnection event processing method in the embodiment of the present application. The processor 702 executes various functional applications and data processing by running the software programs and modules stored in the memory 704, that is, implementing the vulnerability detection method of the above-mentioned application. The memory 704 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 704 may further include a memory remotely located relative to the processor 702, and these remote memories may be connected to the electronic device 70 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0188] The transmission device 706 is used to receive or send data via a network. Specific examples of the aforementioned network may include a wireless network provided by the communications provider of the electronic device 70. In one embodiment, the transmission device 706 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In one embodiment, the transmission device 706 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0189] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the electronic device 70 .

[0190] The serial numbers of the above embodiments are for description only and do not represent the advantages or disadvantages of the embodiments.

[0191] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0192] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0193] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected to achieve the purpose of the present embodiment according to actual needs.

[0194] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0195] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program code.

[0196] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for handling a network disconnection event, characterized in that: include: Acquire multiple passive network disconnection events to be processed, and determine the user group affected by each of the passive network disconnection events; Arrange and combine the multiple passive network disconnection events to obtain multiple passive network disconnection event sequences; For each passive network disconnection event sequence, determine an event handling strategy corresponding to the passive network disconnection event sequence, wherein the event handling strategy is used to sequentially assign to each passive network disconnection event in the passive network disconnection event sequence a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest; for each passive network disconnection event in the passive network disconnection event sequence, use a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration when the event handling strategy is executed, and determine the impact score of the passive network disconnection event based on the attribute information of each user and the data traffic; determine the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence as the comprehensive impact score corresponding to the passive network disconnection event sequence; The multiple passive network disconnection events are processed according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

2. The method according to claim 1, characterized in that Determining an event handling strategy corresponding to the passive network disconnection event sequence includes: Determining the location of each passive network disconnection event corresponding to the passive network disconnection event sequence, the initial location of each repair worker, and the processing time for each repair worker to process different passive network disconnection events; Initialize an event record table and a time record table for each repair worker, wherein the event record table is used to record the passive network disconnection events handled by the corresponding repair worker, and the time record table is used to record the repair completion time of each passive network disconnection event handled by the corresponding repair worker; The following steps are performed cyclically on the passive network disconnection event sequence until the passive network disconnection event sequence is empty: Determining the first passive network disconnection event in the passive network disconnection event sequence as a target passive network disconnection event; For each repair worker, determine the transfer time of the repair worker based on the first position of the last passive network disconnection event in the event record table corresponding to the repair worker and the second position of the target passive network disconnection event, and determine the sum of the last repair completion time in the time record table corresponding to the repair worker and the transfer time as the repair start time of the target passive network disconnection event, wherein, when the event record table corresponding to the repair worker is empty, the initial position of the repair worker is used as the first position; Determining the repair person with the smallest repair start time as the target repair person for handling the target passive network disconnection event; The target passive network disconnection event is removed from the passive network disconnection event sequence, the target passive network disconnection event is added to the event record table corresponding to the target repair personnel, and the sum of the repair start time corresponding to the target repair personnel and the processing time of the target repair personnel for handling the target passive network disconnection event is added to the time record table corresponding to the target repair personnel.

3. The method according to claim 1, characterized in that The pre-trained data traffic prediction model is used to predict the data traffic required by each user in the user group affected by the passive network disconnection event during the duration of the event processing strategy, including: For each user in the user group, obtaining user information of the user, wherein the user information includes at least one of the following: user identification information and user address information; determining, when executing the event handling strategy, a plurality of first time periods corresponding to the duration of the passive network disconnection event, determining an initial first time period corresponding to the moment when the passive network disconnection event occurs, obtaining weather status information for the initial first time period, and obtaining traffic usage data of the user within a preset number of second time periods before the initial first time period, wherein a year is evenly divided into the plurality of time periods; Analyzing the user information, the initial first time period, the weather condition information, and the traffic usage data using the data traffic prediction model to obtain the sub-data traffic required by the user in each of the first time periods; The data flow required by the user during the duration of the passive network disconnection event is determined according to each of the sub-data flows.

4. The method according to claim 3, characterized in that Analyzing the user information, the initial first time period, the weather status information, and the traffic usage data using the data traffic prediction model to obtain the sub-data traffic required by the user in each of the first time periods includes: For each first time period, respectively encode and concatenate the user information, the first time period, the weather status information, a preset number of second time periods before the first time period, and / or traffic usage data within the first time period to obtain a first feature vector corresponding to the first time period; The first feature vector is analyzed using the data traffic prediction model to obtain the sub-data traffic required by the user in the first time period.

5. The method according to claim 3, characterized in that Determining the data flow required by the user during the duration of the passive network disconnection event based on each of the sub-data flows includes: For the initial first time period, determining a first ratio of a first duration from the time when the passive network disconnection event occurs to the time when the initial first time period ends to a second duration of the initial first time period, and updating the sub-data flow corresponding to the initial first time period by multiplying the sub-data flow corresponding to the initial first time period by the first ratio; For a first time period corresponding to the moment when the passive network disconnection event is repaired, determining a second ratio of a third duration from the start time of the first time period to the moment when the passive network disconnection event is repaired to the second duration of the first time period, and updating the sub-data flow rate corresponding to the first time period by multiplying the sub-data flow rate corresponding to the first time period by the second ratio; The sub-data flows corresponding to all the first time periods are summed to obtain the data flow required by the user during the duration of the passive network disconnection event.

6. The method according to claim 3, characterized in that The training process of the data traffic prediction model includes: Constructing an initial model, wherein the initial model is a fully connected neural network; Obtain user information of multiple users, traffic usage data of each user in multiple consecutive historical time periods, and weather status information for each historical time period; Constructing multiple training samples and corresponding sample labels, wherein each training sample includes: user information of a user, traffic usage data of the user within a preset number of historical time periods, a target historical time period after the preset number of historical time periods, and weather status information for the target historical time period; and each sample label includes the traffic usage data of the corresponding user within the corresponding target historical time period; The initial model is iteratively trained using a plurality of the training samples and corresponding sample labels to obtain the data flow prediction model.

7. The method according to claim 1, characterized in that Determining the impact score of the passive network disconnection event based on the attribute information of each user and the data traffic includes: For each user affected by the passive network disconnection event, obtaining attribute information of the user in multiple dimensions, wherein the dimensions include at least one of the following: user type, consumption level, online time, and fault reporting rate; Determine the attribute weight corresponding to the attribute information of each dimension; The total product of the plurality of attribute weights and the data traffic required by the user during the duration of the passive network disconnection event is determined as the impact score of the passive network disconnection event.

8. The method according to claim 1, characterized in that The method further comprises: During the processing of the plurality of passive network disconnection events according to the event processing strategy, if a new passive network disconnection event is received, the new passive network disconnection event and the unrepaired passive network disconnection event are rearranged and combined to obtain a plurality of new passive network disconnection event sequences; Determine the event processing strategy and comprehensive impact score corresponding to each new passive network disconnection event sequence, and process each passive network disconnection event in the new passive network disconnection event sequence according to the event processing strategy corresponding to the passive network disconnection event sequence with the smallest comprehensive impact score.

9. A method for handling a network disconnection event, characterized in that: include: Acquire multiple active network disconnection events to be executed and preset execution time periods, and determine the user groups affected by each of the active network disconnection events; Arrange and combine the multiple active network disconnection events to obtain multiple active network disconnection event sequences; For each active network disconnection event sequence, determine the event processing strategy corresponding to the active network disconnection event sequence, wherein the event processing strategy is used to sequentially determine the execution start time of each active network disconnection event, and the event processing strategy simultaneously satisfies: all active network disconnection events can be completed within the preset execution time period, the impact score of each active network disconnection event is minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when the event processing strategy is executed, and determining the impact score of the active network disconnection event based on the attribute information of each user and the data traffic; determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence; The multiple active network disconnection events are processed according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

10. The method according to claim 9, characterized in that Determining an event handling strategy corresponding to the active network disconnection event sequence includes: Determining the initial location of the repair personnel, the location of each active network disconnection event corresponding to the active network disconnection event sequence, and the processing time of each active network disconnection event; The following steps are performed cyclically on the active network disconnection event sequence until the active network disconnection event sequence is empty: Determining the first active network disconnection event in the active network disconnection event sequence as a target active network disconnection event; Determining the sum of the processing durations of each active network disconnection event in the active network disconnection event sequence and the sum of the transfer times between the locations of each active network disconnection event as the consumed time, and determining the difference between the end time of the preset execution time period and the consumed time as the target end time, wherein the sum of the transfer times in the first cycle includes the transfer time from the initial location of the repair personnel to the location of the target active network disconnection event; Determine the completion time of the last target active network disconnection event executed before executing the target active network disconnection event as the target start time, wherein the target start time of the first cycle is the start time of the preset execution time period; Determining an execution start time of the target active network disconnection event between the target start time and the target end time, wherein the impact score corresponding to the target active network disconnection event is the smallest during the duration starting from the execution start time; The target active network disconnection event is removed from the active network disconnection event sequence.

11. A network disconnection event processing device, characterized in that: include: A first acquisition module is configured to acquire a plurality of passive network disconnection events to be processed and determine a user group affected by each of the passive network disconnection events; A first combining module is used to arrange and combine the multiple passive network disconnection events to obtain multiple passive network disconnection event sequences; A first analysis module is configured to determine, for each passive network disconnection event sequence, an event handling strategy corresponding to the passive network disconnection event sequence, wherein the event handling strategy is configured to sequentially assign to each passive network disconnection event in the passive network disconnection event sequence a repair personnel who can enable the corresponding passive network disconnection event to be repaired the earliest; for each passive network disconnection event in the passive network disconnection event sequence, use a pre-trained data traffic prediction model to predict, when the event handling strategy is executed, the data traffic required by each user in the user group affected by the passive network disconnection event during the duration, and determine an impact score of the passive network disconnection event based on attribute information of each user and the data traffic; and determine the sum of the impact scores of multiple passive network disconnection events in the passive network disconnection event sequence as a comprehensive impact score corresponding to the passive network disconnection event sequence; The first processing module is configured to process the plurality of passive network disconnection events according to an event processing strategy corresponding to a passive network disconnection event sequence having a minimum comprehensive impact score.

12. A network disconnection event processing device, characterized in that: include: A second acquisition module is used to acquire multiple active network disconnection events to be executed and preset execution time periods, and determine the user group affected by each of the active network disconnection events; A second combining module is used to arrange and combine the multiple active network disconnection events to obtain multiple active network disconnection event sequences; The second analysis module is used to determine, for each active network disconnection event sequence, an event processing strategy corresponding to the active network disconnection event sequence, wherein the event processing strategy is used to sequentially determine the execution start time of each active network disconnection event, and the event processing strategy simultaneously satisfies: all active network disconnection events can be completed within the preset execution time period, the impact score of each active network disconnection event is minimized, and the process of determining the impact score of each active network disconnection event includes: using a pre-trained data traffic prediction model to predict the data traffic required by each user in the user group affected by the active network disconnection event during the duration when the event processing strategy is executed, and determining the impact score of the active network disconnection event based on the attribute information of each user and the data traffic; and determining the sum of the impact scores of multiple active network disconnection events in the active network disconnection event sequence as the comprehensive impact score corresponding to the active network disconnection event sequence; The second processing module is configured to process the multiple active network disconnection events according to the event processing strategy corresponding to the active network disconnection event sequence with the smallest comprehensive impact score.

13. A computer program product, characterized in that include: A computer program, wherein when the computer program is executed by a processor, the method for handling a network disconnection event according to any one of claims 1 to 10 is implemented.

14. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the network disconnection event processing method according to any one of claims 1 to 10 through the computer program.

Images