Intelligent scheduling method and device

Abstract

The present application provides an intelligent scheduling method and device that can be used in the financial field or other fields. The method includes: obtaining real-time branch data collected by the operation and maintenance system; analyzing the real-time branch data through a scheduling prediction model constructed by time series decomposition and graph neural network algorithm to obtain the number of scheduled shifts; and generating branch scheduling data based on the number of scheduled shifts. The present application realizes intelligent prediction of branch scheduling through an innovative scheduling prediction module based on time series decomposition and graph neural network, solving the problem that previous manual scheduling relied too much on intuition and was inaccurate.

Description

Technical Field

[0001] The present application relates to the field of network analysis technology and can be applied to the financial field and other fields, and particularly refers to an intelligent scheduling method and device. Background Art

[0002] Artificial intelligence is a new strategic technology that will lead the future and is a key force driving a new round of technological revolution and product transformation. Currently, AI-related technologies continue to evolve, with industrialization and commercialization accelerating, and are rapidly integrating into numerous industries. Machine learning models are playing a crucial role in areas such as computer vision, natural language processing, speech recognition, intelligent risk control, precision marketing, and smart cities. In the banking industry, bank branches are key touchpoints between banks and customers.

[0003] However, scheduling bank branch staff has long been a major problem. Traditionally, bank branches have relied on subjective judgment and historical experience to schedule staff, resulting in a high degree of arbitrariness in staff vacation arrangements. This often results in staff being unable to cope with peak customer traffic, leading to delayed employee departures and a negative impact on customer service experience. Summary of the Invention

[0004] In response to the problems existing in the prior art, the main purpose of the embodiments of the present application is to provide an intelligent scheduling method and device to solve the problem that the current scheduling work of outlets relies too much on human intuition, resulting in inaccurate scheduling of outlet personnel, often leading to insufficient number of employees, poor customer service experience, or arranging too many employees to cause waste of resources.

[0005] In order to achieve the above-mentioned purpose, the present application provides an intelligent scheduling method, which includes: obtaining real-time network data collected by the operation and maintenance system; analyzing the real-time network data through a scheduling prediction model constructed by time series decomposition and graph neural network algorithm to obtain the number of schedules; and generating network scheduling data based on the number of schedules.

[0006] In the above-mentioned intelligent scheduling method, optionally, analyzing the real-time network data through a scheduling prediction model constructed by time series decomposition and graph neural network algorithm to obtain the number of scheduled shifts includes: inputting the real-time network data into the time series decomposition model in the scheduling prediction model to obtain influencing parameters; and obtaining the number of scheduled shifts through analysis of the time series graph neural network model in the scheduling prediction model based on the real-time network data and the influencing parameters.

[0007] In the above-mentioned intelligent scheduling method, optionally, the real-time branch data includes: the number of users received by the branch, the number of waiting users, the number of reception staff, special matters and geographical location.

[0008] In the above-mentioned intelligent scheduling method, optionally, inputting the real-time network point data into the time series decomposition model in the scheduling prediction model to obtain influencing parameters includes: analyzing the real-time network point data through the time series decomposition model to obtain influencing factors, trend items, cycle items and error items; and using the influencing factors, the trend items, the cycle items and the error items as influencing parameters.

[0009] In the above-mentioned intelligent scheduling method, optionally, the method further includes: obtaining historical network data, and obtaining influencing parameters using a time series decomposition algorithm based on the historical network data; and training an initial scheduling model using a graph neural network algorithm using the influencing parameters and the historical network data to obtain the scheduling prediction model.

[0010] In the above-mentioned intelligent scheduling method, optionally, the scheduling prediction model is obtained by training the initial scheduling model through the graph neural network algorithm using the influencing parameters and the historical network data, including: generating a graph neural network through the graph neural network algorithm according to the influencing parameters and the historical network data; and obtaining the scheduling prediction model by training the initial scheduling model through a time recursive neural network according to the graph neural network.

[0011] In the above-mentioned intelligent scheduling method, optionally, generating network scheduling data according to the scheduling quantity includes: extracting pre-stored network personnel information according to the scheduling quantity; generating candidate personnel data according to the network personnel information, and obtaining scheduling data by screening according to the scheduling quantity and the candidate personnel data.

[0012] The present application also provides an intelligent scheduling device, which includes: an acquisition module for acquiring real-time network point data collected by the operation and maintenance system; an analysis module for analyzing the real-time network point data through a scheduling prediction model constructed by time series decomposition and graph neural network algorithm to obtain the number of schedules; and a processing module for generating network point scheduling data based on the number of schedules.

[0013] The present application also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the above method is implemented when the processor executes the program.

[0014] The present application also provides a computer-readable storage medium, which stores a computer program for executing the above method.

[0015] The present application also provides a computer program product, comprising a computer program / instruction, which implements the steps of the above method when the computer program / instruction is executed by a processor.

[0016] The intelligent branch scheduling system based on time series decomposition and graph neural network involved in this application obtains information such as the number of customers received, the number of waiting customers, the number of reception employees, special matters, geographical location, etc. at the branch at a certain time, and realizes intelligent prediction of branch scheduling through an innovative scheduling prediction module based on time series decomposition and graph neural network, thereby solving the problem that previous manual scheduling relies too much on intuition and is inaccurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 This is a flow chart of an intelligent scheduling method according to an embodiment of the present application;

[0019] Figure 2 This is a schematic diagram of the generation logic of the scheduling prediction model in the embodiment of the present application;

[0020] Figure 3 This is a schematic diagram of the process of obtaining the scheduling prediction model in the embodiment of this application;

[0021] Figure 4 This is a schematic diagram of the structure of the knowledge graph in the embodiment of this application;

[0022] Figure 5 This is a structural diagram of a timing diagram neural model according to an embodiment of the present application;

[0023] Figure 6 This is a structural diagram of a scheduling prediction model in another embodiment of the present application;

[0024] Figure 7 This is a schematic diagram of the process of generating the number of shifts in the embodiment of the present application;

[0025] Figure 8 This is a schematic diagram of the process of generating network scheduling data in another embodiment of the present application;

[0026] Figure 9 This is a schematic diagram of the application process of the intelligent scheduling method in another embodiment of the present application;

[0027] Figure 10 This is a structural diagram of an intelligent scheduling device in another embodiment of the present application;

[0028] Figure 11 This is a schematic diagram of the application structure of the intelligent scheduling device in another embodiment of the present application;

[0029] Figure 12 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0030] The embodiments of the present application provide an intelligent operation and maintenance method and device, which can be used in the financial field and other fields. It should be noted that the intelligent operation and maintenance method and device of the present application can be used in the financial field, and can also be used in any field other than the financial field. The application field of the intelligent operation and maintenance method and device of the present application is not limited.

[0031] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0032] like Figure 1 The figure shows a flow chart of an intelligent scheduling method according to an embodiment of the present application, the method comprising:

[0033] S101 obtains real-time network point data collected by the operation and maintenance system;

[0034] S102 analyzes the real-time network point data to obtain the number of scheduled shifts using a scheduling prediction model constructed by a time series decomposition model and a time series graph neural network model;

[0035] S103 generates network scheduling data according to the scheduling quantity.

[0036] In the above embodiment, features are constructed by acquiring information such as the number of customers received, the number of customers waiting, the number of reception staff, special events, and geographic location at a given time. This is achieved through an innovative scheduling prediction module based on time series decomposition and graph neural networks, enabling intelligent prediction of branch scheduling, improving customer experience, effectively utilizing staff resources, and reducing waste of staff resources. The real-time branch data may include information such as the number of customers received, the number of customers waiting, the number of reception staff, special events, and geographic location.

[0037] Please refer to Figure 2 As shown, in one embodiment of the present application, the method further includes:

[0038] S201 acquires historical network point data, and obtains influencing parameters using a time series decomposition model based on the historical network point data;

[0039] S202 trains the initial scheduling model using the influencing parameters and the historical network data through a graph neural network algorithm to obtain the scheduling prediction model.

[0040] For further information, please refer to Figure 3 As shown, the scheduling prediction model is obtained by training the initial scheduling model through the time series graph neural network model using the influencing parameters and the historical network data, including:

[0041] S301 generates a graph neural network based on the influencing parameters and the historical network point data through the time-series graph neural network model;

[0042] S302 trains an initial scheduling model through a time recursive neural network according to the graph neural network to obtain the scheduling prediction model.

[0043] Specifically, in actual work, it includes a time series decomposition model and a time series graph neural network model. The specific construction logic is as follows:

[0044] Timing decomposition model: ;

[0045] in, is the value of a certain indicator at time t, which can be the number of customers received by the branch on a certain day, the number of customers consulting a certain business, or the total waiting time of customers in the branch on a certain day. It is broken down into the following 4 items:

[0046] The value of the special influencing factors at time t (such as the impact of holidays, sports events, major events, etc.);

[0047] is the trend item value at time t;

[0048] is the value of the periodic item at time t (such as seasonal impact);

[0049] is the error term value at time t.

[0050] The calculation method of each value is as follows:

[0051] The value of the special item influencing factor is calculated based on whether the special item factor occurs at time t, which indicates the value of the influencing factor brought by the special item at time t. The input value indicates the possible impact of the event, 0 means no special event occurs, and the upper limit is 10; then a generalized linear model is used to calculate .Right now, , where g is the scale transformation function, w is the weight coefficient matrix, and b is the bias vector, which is fitted in subsequent calculations.

[0052] The trend item value is based on the period before time t. and Determine, indicating the trend value at time t. The calculation method is , where T is the set time window length, which is fitted in subsequent calculations.

[0053] The value of the periodic term p(t) is calculated based on the period before time t. 、 、 Determine, represents the value of the periodic term at time t. The calculation method is , where N is the set time window length, and is the coefficient, will Perform fitting.

[0054] The error term value represents the error size in the time series decomposition at time t, and is calculated as .

[0055] therefore,

[0056] ;

[0057] ;

[0058] ;

[0059] ;

[0060] .

[0061] Through the above decomposition method, the numerical value at a certain time t can be decomposed into the numerical value of the special item influencing factor, the numerical value of the trend item, the numerical value of the cycle item, and the numerical value of the error item; the numerical values ​​involved are all helpful for subsequent calculations.

[0062] Regarding temporal graph neural networks, knowledge graphs are essentially composed of nodes and edges. Each node represents an entity in the real world, while each edge represents the relationship between entities. Simply put, a knowledge graph is a relational network that connects all different types of information, and this relational network will evolve over time.

[0063] In this scenario, each outlet is regarded as a node. The geographical location, size, and organizational level of each outlet are used as the attributes of the node. At the same time, the special item influencing factor values, trend item values, cycle item values, and error item values ​​of various data (such as the number of people queuing at the outlet, the number of employees at the outlet) obtained by the time series decomposition method at time t are also used as the attribute information of the graph node at time t. In addition, the distance between each outlet is used as the edge weight in each graph node. Afterwards, a time series graph neural network algorithm is used to analyze the changes and trends of the graph structure over time to predict how many employees should be scheduled to work on a certain day. For details, please refer to Figure 4 As shown, the differences in nodes and edges represent different types. It can be seen that the structure of the graph changes significantly from time t to time t+1.

[0064] For reference Figure 5 As shown, in the temporal graph neural network (T-GNN) model, in order to achieve temporal dynamic learning, this application provides the following two points: 1. A graph neural network (GCN) is learned separately for each time slice, and the different inputs of each GCN are reflected in the different adjacency matrices of the graph; 2. In order to consider the dynamic graph connection, the LSTM algorithm is used to string together the GCN model parameters of each time slice for sequence learning.

[0065] T-GNN uses the LSTM model to learn parameters in series. The hidden state of the LSTM model uses the parameters of the previous moment. , and its input also uses the parameters of the previous moment , The update formula is as follows:

[0066]

[0067] Therefore, Node Embedding and The update formula is as follows:

[0068]

[0069] The model structure can be referenced Figure 6 As shown in the figure, GCN is a graph convolutional neural network, which actually has the same function as the convolutional neural network CNN, that is, a feature extractor, except that its object is graph data.

[0070] In graph data, there are N nodes, each with its own unique features. These node features form an N×D-dimensional matrix X. The relationships between nodes also form an N×N-dimensional matrix A, also known as the adjacency matrix. X and A are the inputs to the model.

[0071] GCN is also a neural network layer. The propagation method between its layers is:

[0072]

[0073] In this formula:

[0074] A wave = A + I, I is the identity matrix;

[0075] Wave D is the degree matrix of wave A, and the formula is ;

[0076] H is the feature of each layer. For the input layer, H is X.

[0077] σ is a non-linear activation function.

[0078] The above two models are trained uniformly in an end-to-end manner, with the original feature data as input and the prediction results of the time series graph neural network as the final output; the model is trained with the goal of reducing the mean absolute error (MAE), and the fitting of each parameter to be fitted is completed during the training process.

[0079] The calculation method of MAE is:

[0080] ,

[0081] in The number of employees that should be assigned to the outlets as predicted by the model, The number of employees who should be assigned to manually label the outlets is n, and n is the number of training samples.

[0082] Please refer to Figure 7 As shown, in one embodiment of the present application, the scheduling prediction model constructed by time series decomposition and graph neural network algorithm analyzes the real-time network data to obtain the scheduling quantity, which includes:

[0083] S701: input the real-time network point data into the time series decomposition model in the scheduling prediction model to obtain the influencing parameters;

[0084] S702 obtains the number of shifts to be scheduled based on the real-time network point data and the influencing parameters through the time series graph neural network model analysis in the shift prediction model.

[0085] Among them, inputting the real-time network data into the time series decomposition model in the scheduling prediction model to obtain influencing parameters may include: analyzing the real-time network data through the time series decomposition model to obtain influencing factors, trend items, cycle items and error items; and using the influencing factors, the trend items, the cycle items and the error items as influencing parameters.

[0086] Please refer to Figure 8 As shown, in one embodiment of the present application, generating network scheduling data according to the number of scheduled shifts includes:

[0087] S801 extracts pre-stored network personnel information according to the number of shifts;

[0088] S802 generates candidate personnel data based on the network personnel information, and obtains the scheduling data based on the number of scheduled shifts and the candidate personnel data.

[0089] In this embodiment, the outlet personnel information can be pre-stored in a designated location, and then when the number of shifts is determined, the corresponding outlet personnel information can be retrieved, and the corresponding number of personnel can be selected as the scheduling data according to the number of shifts; of course, in actual use, a scheduling rule can also be preset, and the outlet personnel information can be screened based on the scheduling rule. Relevant technical personnel in this field can choose to use it according to actual needs, and this application does not make further limitations here.

[0090] To better understand the overall process of the intelligent scheduling method provided by this application, please refer to the following Figure 9 As shown, the above embodiments are generally described.

[0091] The process of feature information acquisition, time series decomposition and graph neural network prediction, and scheduling in the intelligent scheduling system for network points based on time series decomposition and graph neural network includes the following four steps:

[0092] Step 1: Acquire feature data. Build features by acquiring information such as the number of customers received, the number of waiting customers, the number of reception staff, special events, and geographic location at a particular time.

[0093] Step 2: Model training. Using historical feature data and the manually labeled number of branch staff to be assigned, train the time series decomposition model and the time series graph neural network model. (The model training process has been described in detail in the previous embodiment and will not be detailed here.)

[0094] Step 3: Use the latest feature data obtained in step 1 and the intelligent scheduling model trained in step 2 to predict the number of employees that should be assigned to the branch at a certain time in the future.

[0095] Step 4: Arrange the number of branch employees based on the forecast value obtained in step 3.

[0096] Please refer to Figure 10As shown, the present application also provides an intelligent scheduling device, which includes: an acquisition module for acquiring real-time network point data collected by the operation and maintenance system; an analysis module for analyzing the real-time network point data to obtain the number of schedules through a scheduling prediction model constructed by a time series decomposition model and a time series graph neural network model; and a processing module for generating network point scheduling data according to the number of schedules.

[0097] The specific structure can be referred to Figure 11 As shown, in actual work, the acquisition module, or feature information acquisition, primarily collects information, including the number of customers received at the branch at time t, the number of customers waiting at the branch at time t, the number of employees receiving services at the branch at time t, special events (holidays) at time t, the branch's geographic location, and other information. This is followed by an analysis module constructed using a time series decomposition model and a time series graph neural network to analyze the data and obtain the corresponding predicted headcount. Finally, the processing module completes the process by obtaining the predicted number of employees scheduled for work at a certain point in the future and the branch's shift schedule. The process is as follows:

[0098] S01 Feature Information Acquisition Unit: Construct features based on information such as the number of customers received, the number of waiting customers, the number of reception staff, special matters, and geographical location at a certain time; and then pass this information to the S02 module.

[0099] S02 Outbound Call Time Series Decomposition and Graph Neural Network Module: After receiving the relevant feature information from the S01 module, the S02 module transmits the prediction results to the S03 module through an innovative scheduling prediction model based on time series decomposition and graph neural network.

[0100] S03 subsequent processing unit: The S03 module receives the predicted number of employees to be scheduled for work at a certain time in the future sent by the S02 unit, and schedules the work tasks of the branch employees based on the prediction results.

[0101] The intelligent branch scheduling system based on time series decomposition and graph neural network involved in this application obtains information such as the number of customers received, the number of waiting customers, the number of reception employees, special matters, geographical location, etc. at the branch at a certain time, and realizes intelligent prediction of branch scheduling through an innovative scheduling prediction module based on time series decomposition and graph neural network, thereby solving the problem that previous manual scheduling relies too much on intuition and is inaccurate.

[0102] The present application also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above method when executing the program.

[0103] The present application also provides a computer program product, comprising a computer program / instruction, which implements the steps of the above method when the computer program / instruction is executed by a processor.

[0104] The present application also provides a computer-readable storage medium, which stores a computer program for executing the above method.

[0105] like Figure 12 As shown, the electronic device 600 may further include: a communication module 110, an input unit 120, an audio processor 130, a display 160, and a power supply 170. It is worth noting that the electronic device 600 does not necessarily have to include Figure 12 In addition, the electronic device 600 may also include all components shown in Figure 12 For components not shown, reference may be made to the prior art.

[0106] like Figure 12 As shown, the central processing unit 100 is sometimes also referred to as a controller or an operation control unit, and may include a microprocessor or other processor device and / or logic device. The central processing unit 100 receives inputs and controls the operations of various components of the electronic device 600 .

[0107] Memory 140 may be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It may store the aforementioned failure-related information and may also store programs that execute the relevant information. The CPU 100 may execute the programs stored in memory 140 to implement information storage or processing.

[0108] The input unit 120 provides input to the CPU 100. The input unit 120 may be, for example, a keypad or touch input device. The power supply 170 is used to provide power to the electronic device 600. The display 160 is used to display objects such as images and text. The display may be, for example, an LCD display, but is not limited thereto.

[0109] The memory 140 may be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), or SIM card. Alternatively, it may be a memory that retains information even when power is off, can be selectively erased, and can be stored with more data. Examples of such memory are sometimes referred to as EPROMs. The memory 140 may also be some other type of device. The memory 140 includes a buffer memory 141 (sometimes referred to as a buffer). The memory 140 may include an application / function storage unit 142 for storing application programs and function programs, or processes used by the central processing unit 100 to execute operations of the electronic device 600.

[0110] The memory 140 may also include a data storage unit 143 for storing data, such as contacts, digital data, images, sounds, and / or any other data used by the electronic device. The driver storage unit 144 of the memory 140 may include various driver programs for communication functions of the electronic device and / or for executing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0111] The communication module 110 is a transmitter / receiver 110 that transmits and receives signals via an antenna 111. The communication module (transmitter / receiver) 110 is coupled to the central processor 100 to provide input signals and receive output signals, which may be the same as in a conventional mobile communication terminal.

[0112] Based on different communication technologies, multiple communication modules 110 may be provided in the same electronic device, such as cellular network modules, Bluetooth modules, and / or wireless local area network modules. The communication module (transmitter / receiver) 110 is also coupled to a speaker 131 and a microphone 132 via an audio processor 130, providing audio output via the speaker 131 and receiving audio input from the microphone 132, thereby implementing common telecommunication functions. The audio processor 130 may include any suitable buffer, decoder, amplifier, etc. Furthermore, the audio processor 130 is coupled to the central processing unit 100, enabling local recording via the microphone 132 and playback of stored audio via the speaker 131.

[0113] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0114] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0115] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0116] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0117] Specific embodiments are used in this application to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. An intelligent scheduling method, characterized in that: The method comprises: Obtain real-time network point data collected by the operation and maintenance system; The scheduling prediction model constructed by the time series decomposition model and the time series graph neural network model analyzes the real-time network point data to obtain the number of scheduled shifts; Generate network scheduling data according to the number of shifts; The method further comprises: Acquire historical network point data, and obtain influencing parameters using a time series decomposition model based on the historical network point data; Generate a graph neural network using a graph neural network algorithm according to the influencing parameters and the historical network point data; The scheduling prediction model is obtained by training an initial scheduling model through a time recursive neural network according to the graph neural network; The time series decomposition model is: ;in, is the value of a certain indicator at time t; is the value of the special item influencing factor at time t; is the trend item value at time t; is the value of the periodic term at time t; is the error term value at time t; described It is calculated based on whether there is a special factor at time t, which indicates the value of the influencing factor brought by the special factor at time t; the input is The input value indicates the degree of impact of the event; it is calculated by the generalized linear model ,Right now, , where g is the scale transformation function; w is the weight coefficient matrix; b is the bias vector; described According to the period before time t and Determine, indicating the trend value at time t; described According to the period before time t 、 、 Determine, representing the value of the periodic term at time t; described Indicates the error size in the time series decomposition at time t, which is calculated as .

2. The intelligent scheduling method according to claim 1, characterized in that: The scheduling prediction model constructed by the time series decomposition model and the time series graph neural network model analyzes the real-time network data to obtain the scheduling quantity, including: Inputting the real-time network point data into the time series decomposition model in the scheduling prediction model to obtain influencing parameters; The number of shifts is obtained by analyzing the time series graph neural network model in the shift prediction model according to the real-time network point data and the influencing parameters.

3. The intelligent scheduling method according to claim 2, characterized in that: The real-time branch data includes: the number of users received by the branch, the number of users waiting, the number of reception staff, special matters and geographical location.

4. The intelligent scheduling method according to claim 2, characterized in that: Inputting the real-time network point data into the time series decomposition model in the scheduling prediction model to obtain influencing parameters includes: Analyze the real-time network point data through the time series decomposition model to obtain influencing factors, trend items, cycle items and error items; The influencing factors, the trend term, the period term and the error term are used as influencing parameters.

5. The intelligent scheduling method according to claim 1, characterized in that: Generating the network scheduling data according to the scheduling quantity includes: Extracting pre-stored branch personnel information based on the number of shifts; The candidate personnel data is generated according to the network personnel information, and the scheduling data is obtained by screening according to the number of schedules and the candidate personnel data.

6. An intelligent scheduling device, characterized in that: The device comprises: The collection module is used to obtain real-time network point data collected by the operation and maintenance system; An analysis module is used to analyze the real-time network point data to obtain the number of scheduled shifts through a scheduling prediction model constructed by a time series decomposition model and a time series graph neural network model; A processing module, configured to generate network scheduling data according to the number of scheduled shifts; The device is further configured to: obtain historical network point data, and obtain influencing parameters using a time series decomposition model based on the historical network point data; generate a graph neural network using a graph neural network algorithm based on the influencing parameters and the historical network point data; and train an initial scheduling model using a time recursive neural network based on the graph neural network to obtain the scheduling prediction model; The time series decomposition model is: ;in, is the value of a certain indicator at time t; is the value of the special item influencing factor at time t; is the trend item value at time t; is the value of the periodic term at time t; is the error term value at time t; described It is calculated based on whether there is a special factor at time t, which indicates the value of the influencing factor brought by the special factor at time t; the input is The input value indicates the degree of impact of the event; it is calculated by the generalized linear model ,Right now, , where g is the scale transformation function; w is the weight coefficient matrix; b is the bias vector; described According to the period before time t and Determine, indicating the trend value at time t; described According to the period before time t 、 、 Determine, indicating the value of the periodic term at time t; Indicates the error size in the time series decomposition at time t, which is calculated as .

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program for executing the method according to any one of claims 1 to 5.

9. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images