Deep learning-based population migration flow generation method, system and terminal

Abstract

The invention discloses a population migration flow generation method, system and terminal based on deep learning, and the method comprises the steps: obtaining mobile phone signaling data of a target resident, carrying out the classification according to meteorological data, and tracking the mobile phone signaling data of the target resident in a target region; constructing and training a deep learning model according to the mobile phone signaling data and the collected multi-source data to obtain a population migration model; and inputting the current mobile phone signaling data into the population migration model, and outputting the population migration flow in the target area. By integrating multi-source data and a deep learning technology, the population migration model for predicting the population migration flow is constructed, dynamic changes of residents in a disaster period can be accurately captured, migration flow of different communities or regions in a city can be depicted more finely, the data precision and timeliness are improved, and the population migration flow prediction method can be used for predicting the population migration flow of the residents in the disaster period. Accurate prediction of the urban resident population migration flow during the disaster period is realized.

Description

Technical Field

[0001] The present invention relates to the field of information analysis technology, and in particular to a method, system, terminal and computer-readable storage medium for generating a population migration flow based on deep learning. Background Art

[0002] Population migration is affected by many factors, including changes in statistical caliber, accelerated urbanization process, unbalanced economic development, etc. These factors work together to lead to the trend of population migration from rural to urban areas, while also bringing about the acceleration of urbanization process and changes in social and economic structure.

[0003] However, existing technologies mainly rely on historical data and expert experience to analyze population migration flows during disasters. This method has problems such as low data accuracy, lack of real-time and insufficient spatial precision. At the same time, traditional methods cannot make full use of modern data sources (such as mobile phone signaling data, remote sensing image data, etc.) to carry out refined disaster population migration flow predictions.

[0004] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0005] The main purpose of the present invention is to provide a method, system, terminal and computer-readable storage medium for generating population migration flows based on deep learning, aiming to solve the problem that the analysis of population migration flows in the prior art lacks data accuracy and real-time performance, and has insufficient spatial precision and a single data source, thus leading to inaccurate analysis of population migration flows.

[0006] To achieve the above object, the present invention provides a method for generating a population migration flow based on deep learning, and the method for generating a population migration flow based on deep learning comprises the following steps:

[0007] Acquire multiple initial mobile phone signaling data of each resident in the target area, screen out multiple target residents according to all the initial mobile phone signaling data, and pre-process all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident;

[0008] Acquire the meteorological data of the target area, divide all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and acquire a plurality of third mobile phone signaling data of all the target residents;

[0009] Constructing a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, constructing a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, acquiring multi-source data, and constructing a deep learning model according to the multi-source data, the first matrix, and the second matrix;

[0010] Performing feature extraction on a plurality of the first mobile phone signaling data and a plurality of the second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographical distances, and inputting all the feature vectors and all the geographical distances into the deep learning model for training to obtain a population migration model;

[0011] A plurality of current mobile phone signaling data of a plurality of residents in the target area are obtained and input into the population migration model, and the population migration flow in the target area is output.

[0012] Optionally, the method for generating a population migration flow based on deep learning, wherein the step of obtaining a plurality of initial mobile phone signaling data of each resident in the target area, screening out a plurality of target residents according to all the initial mobile phone signaling data, and preprocessing all the initial mobile phone signaling data of all the target residents to obtain a plurality of mobile phone signaling data of each of the target residents, specifically includes:

[0013] Obtain all initial mobile phone signaling data of all residents in the target area within a preset time. If there are multiple complete initial mobile phone signaling data of a resident, the resident is taken as the target resident, wherein the complete initial mobile phone signaling data means that the initial mobile phone signaling data of the target resident every day within the preset time and within the target area is complete;

[0014] The formats of all initial mobile phone signaling data of all the target residents are uniformly processed, and the data with abnormal activity trajectories are deleted to obtain multiple mobile phone signaling data of each target resident within a preset time.

[0015] Optionally, the method for generating a population migration flow based on deep learning, wherein the step of obtaining the meteorological data of the target area, dividing all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtaining a plurality of third mobile phone signaling data of all the target residents, specifically comprises:

[0016] Acquire meteorological data of the target area within the preset time, and divide the preset time into a first preset time and a second preset time according to the meteorological data;

[0017] Marking all mobile phone signaling data within the first preset time as a plurality of first mobile phone signaling data, and marking all mobile phone signaling data within the second preset time as a plurality of second mobile phone signaling data;

[0018] A plurality of third mobile phone signaling data of all the target residents are obtained, wherein the third mobile phone signaling data represent the mobile phone signaling data of the target residents after they migrate to the destination area after the preset time.

[0019] Optionally, the method for generating a population migration flow based on deep learning, wherein the first matrix is ​​constructed according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, the second matrix is ​​constructed according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, and multi-source data is acquired, and a deep learning model is constructed according to the multi-source data, the first matrix and the second matrix, specifically comprising:

[0020] Determine the first target number of the target residents within the target area within the first preset time according to all the first mobile phone signaling data;

[0021] Determine the second target number of the target residents within the target area within the second preset time according to all the second mobile phone signaling data;

[0022] Determine the third target number of the target residents in the target area after the preset time according to all the third mobile phone signaling data;

[0023] Acquire multi-source data in the target area, construct a first matrix according to the first target number of residents and the third target number of residents, and construct a second matrix according to the second target number of residents and the third target number of residents;

[0024] A deep learning model is constructed based on the multi-source data, the first matrix and the second matrix.

[0025] Optionally, in the method for generating population migration flow based on deep learning, the feature vector includes: migration frequency and average migration distance;

[0026] The feature extraction of the plurality of the first mobile phone signaling data and the plurality of the second mobile phone signaling data is performed to obtain a plurality of feature vectors and a plurality of geographical distances, and all the feature vectors and all the geographical distances are input into the deep learning model for training to obtain a population migration model, specifically including:

[0027] Performing feature extraction on all the first mobile phone signaling data to obtain first coordinate data of all the target residents in the target area within the first preset time;

[0028] Performing feature extraction on all the second mobile phone signaling data to obtain second coordinate data of all the target residents in the target area within the second preset time;

[0029] Calculate the migration frequency, the average migration distance and the geographical distance of all the target residents according to all the first coordinate data and all the second coordinate data;

[0030] The migration frequency, the average migration distance and the geographical distance are input into the deep learning model for training to obtain a population migration model.

[0031] Optionally, the method for generating a population migration flow based on deep learning, wherein the step of obtaining a plurality of current mobile phone signaling data of a plurality of residents in the target area and inputting the data into the population migration model, and outputting the population migration flow in the target area, specifically comprises:

[0032] Acquire multiple current mobile phone signaling data of multiple residents in the target area, and extract current coordinate information from each current mobile phone signaling data, wherein the current coordinate information indicates the current position of the resident corresponding to the current mobile phone signaling data;

[0033] Each current coordinate information is input into the population migration model for prediction, and the population migration flow of the target area is output, wherein the population migration flow represents the predicted population number of residents in the target area migrating to the destination area.

[0034] Optionally, the method for generating a population migration flow based on deep learning, wherein the step of inputting each current coordinate information into the population migration model for prediction and outputting the population migration flow of the target area specifically includes:

[0035] Input each of the current coordinate information into the feedforward neural network of the population migration model, the feedforward neural network analyzes the current coordinate information through an activation function, and outputs a predicted score of the target area;

[0036] The population migration model converts the predicted score into the population migration flow through a preset function and outputs the result.

[0037] In addition, to achieve the above-mentioned purpose, the present invention also provides a population migration flow generation system based on deep learning, wherein the population migration flow generation system based on deep learning includes:

[0038] A data preprocessing module is used to obtain multiple initial mobile phone signaling data of each resident in the target area, screen out multiple target residents according to all the initial mobile phone signaling data, and preprocess all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident;

[0039] A data classification module, used for obtaining meteorological data of the target area, classifying all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtaining a plurality of third mobile phone signaling data of all the target residents;

[0040] A model building module, used to build a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, build a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, and obtain multi-source data, and build a deep learning model according to the multi-source data, the first matrix and the second matrix;

[0041] A model training module, used for extracting features from a plurality of the first mobile phone signaling data and a plurality of the second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographical distances, and inputting all the feature vectors and all the geographical distances into the deep learning model for training to obtain a population migration model;

[0042] The migration prediction module is used to obtain multiple current mobile phone signaling data of multiple residents in the target area, input the data into the population migration model, and output the population migration flow in the target area.

[0043] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a deep learning-based population migration flow generation program stored on the memory and run on the processor, and when the deep learning-based population migration flow generation program is executed by the processor, the steps of the deep learning-based population migration flow generation method as described above are implemented.

[0044] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a population migration flow generation program based on deep learning, and when the population migration flow generation program based on deep learning is executed by a processor, the steps of the population migration flow generation method based on deep learning as described above are implemented.

[0045] In the present invention, a plurality of initial mobile phone signaling data of each resident in the target area are obtained, a plurality of target residents are screened out according to all the initial mobile phone signaling data, and all the initial mobile phone signaling data of all the target residents are preprocessed to obtain a plurality of mobile phone signaling data of each target resident; meteorological data of the target area are obtained, all the mobile phone signaling data are divided into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and a plurality of third mobile phone signaling data of all the target residents are obtained; a first matrix is ​​constructed according to the plurality of the first mobile phone signaling data and the plurality of the third mobile phone signaling data, and a plurality of the third mobile phone signaling data of each target resident are obtained according to the plurality of the first mobile phone signaling data and the plurality of the third mobile phone signaling data. A second matrix is ​​constructed with multiple second mobile phone signaling data and multiple third mobile phone signaling data, and multi-source data is obtained, and a deep learning model is constructed according to the multi-source data, the first matrix and the second matrix; feature extraction is performed on multiple first mobile phone signaling data and multiple second mobile phone signaling data to obtain multiple feature vectors and multiple geographical distances, and all the feature vectors and all the geographical distances are input into the deep learning model for training to obtain a population migration model; multiple current mobile phone signaling data of multiple residents in the target area are obtained, and input into the population migration model, and the population migration flow in the target area is output. The present invention constructs a population migration model for predicting population migration flow by integrating multi-source data and deep learning technology, which can accurately capture the dynamic changes of residents during disasters, and more finely characterize the migration flow of different communities or regions within the city, thereby improving data accuracy and timeliness, and realizing accurate prediction of urban residents' population migration flow during disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of a preferred embodiment of the method for generating population migration flow based on deep learning of the present invention;

[0047] Figure 2 It is a migration schematic diagram of a preferred embodiment of the method for generating a population migration flow based on deep learning of the present invention;

[0048] Figure 3 It is a structural diagram of a preferred embodiment of a population migration flow generation system based on deep learning of the present invention;

[0049] Figure 4 Schematic diagram of the operating environment of a preferred embodiment of the terminal of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0051] The method for generating population migration flow based on deep learning described in the preferred embodiment of the present invention is as follows: Figure 1 As shown, the population migration flow generation method based on deep learning includes the following steps:

[0052] Step S10, obtaining multiple initial mobile phone signaling data of each resident in the target area, screening out multiple target residents based on all the initial mobile phone signaling data, and preprocessing all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident.

[0053] Among them, by obtaining the initial mobile phone signaling data of each resident in the target area, the change characteristics of each resident's residence within a certain period of time (such as during a disaster) are tracked, and the frequency of acquisition can be selected according to the needs of the user (for example, once every 15 minutes), so as to identify the changes in the residential community of residents at a high frequency, capture the spatiotemporal migration characteristics of residents, improve the real-time nature of information acquisition, and provide reliable samples for subsequent model training, thereby improving the accuracy of migration flow prediction.

[0054] Specifically, all initial mobile phone signaling data of all residents in the target area within a preset time are obtained. If there are multiple complete initial mobile phone signaling data of a resident, the resident is taken as the target resident, wherein the complete initial mobile phone signaling data means that the initial mobile phone signaling data of the target resident every day within the preset time and within the target area are complete; the formats of all initial mobile phone signaling data of all the target residents are uniformly processed, and the data with abnormal activity trajectories are deleted to obtain multiple mobile phone signaling data of each target resident within the preset time.

[0055] Among them, the boundaries of each urban community plot in the target area are identified through relevant data and ArcGIS (ArcGIS Platform, geographic information system) spatial analysis; then the initial mobile phone signaling data of residents in the target area within a preset time is obtained, including micro-population spatiotemporal activity data with geographic coordinate information, and the data is pre-processed in a unified format to delete device information without tag ID and data with abnormal activity trajectories; according to the geographic information coordinates of highways and main roads, the relevant coordinate data is deleted, and finally the mobile phone signaling data is obtained; among them, in order to protect the privacy of residents and the uniformity of data, the time interval for collecting coordinate information is set to every 15 minutes, and only the sample of mobile phone users whose spatiotemporal coordinates are fully tracked every day is retained.

[0056] Furthermore, if the mobile phone signaling data of a certain resident is incomplete for a few days within the preset time, the mobile phone signaling data of the resident will not be considered and will be deleted, thereby improving the integrity of the data and avoiding the impact of incomplete data on subsequent model training. Based on the mobile phone signaling data of each user, the spatial grid that each mobile phone user visits most frequently every day is identified to represent their main activity area, and the vector<p,tp> Represents each coordinate data, where p represents the spatial grid coordinate of the resident at time tp. By using high-frequency mobile phone signaling data and real-time remote sensing image data, the dynamic changes of residents during disasters can be accurately captured, improving data accuracy and timeliness.

[0057] Step S20, obtaining meteorological data of the target area, dividing all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtaining a plurality of third mobile phone signaling data of all the target residents.

[0058] Specifically, meteorological data of the target area within the preset time is obtained, and the preset time is divided into a first preset time and a second preset time according to the meteorological data; all mobile phone signaling data within the first preset time is marked as a plurality of first mobile phone signaling data, and all mobile phone signaling data within the second preset time is marked as a plurality of second mobile phone signaling data; multiple third mobile phone signaling data of all the target residents are obtained, wherein the third mobile phone signaling data represents the mobile phone signaling data of the target residents after migrating to the destination area after the preset time.

[0059] Among them, after pre-processing all mobile phone signaling data, the preset time is divided into three categories according to the obtained meteorological data. For example, the disaster period is divided into the pre-disaster period, the disaster period and the post-disaster period through meteorological data. According to these three types of time, all mobile phone signaling data can be divided into three categories, namely, the first mobile phone signaling data in the pre-disaster period, the second mobile phone signaling data in the disaster period, and the third mobile phone signaling data in the post-disaster period.

[0060] Step S30, constructing a first matrix based on multiple first mobile phone signaling data and multiple third mobile phone signaling data, constructing a second matrix based on multiple second mobile phone signaling data and multiple third mobile phone signaling data, and acquiring multi-source data, and constructing a deep learning model based on the multi-source data, the first matrix and the second matrix.

[0061] Specifically, the first target number of target residents within the target area within the first preset time is determined based on all the first mobile phone signaling data; the second target number of target residents within the target area within the second preset time is determined based on all the second mobile phone signaling data; the third target number of target residents within the destination area after the preset time is determined based on all the third mobile phone signaling data; multi-source data within the target area is acquired, a first matrix is ​​constructed based on the first target number of residents and the third target number of residents, and a second matrix is ​​constructed based on the second target number of residents and the third target number of residents; a deep learning model is constructed based on the multi-source data, the first matrix and the second matrix.

[0062] Among them, the target area represents the area when the disaster occurs, and the destination area represents the destination where the residents in the target area migrate when the disaster occurs. The first mobile phone signaling data and the third mobile phone signaling data can be used to determine the number of residents in the pre-disaster period and the pre- and post-disaster period. The first matrix is ​​constructed based on these two data, which represents the number of residents in the target area who migrated from one city to another city during the pre-disaster period. Similarly, the second matrix represents the number of residents in the target area who migrated from one city to another city during the disaster. Furthermore, in order to compare the linear correlation of the matrices of these two periods, the Pearson correlation coefficient can be used to quantify their linear relationship to provide better samples for subsequent model training.

[0063] Step S40: extract features from the plurality of the first mobile phone signaling data and the plurality of the second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographic distances, and input all the feature vectors and all the geographic distances into the deep learning model for training to obtain a population migration model.

[0064] The feature vector includes migration frequency and average migration distance. The migration frequency indicates how many residents migrate out of the target area every day, and the average migration distance indicates the distance that residents migrate in one day. Based on the user's mobile phone signaling data, in addition to obtaining feature vectors and geographic distances, community space information and socioeconomic data in the target area can also be obtained for model training, thereby providing a more detailed depiction of migration flows in different communities or regions within the city and enhancing the spatial resolution of the model.

[0065] Specifically, feature extraction is performed on all the first mobile phone signaling data to obtain the first coordinate data of all the target residents in the target area within the first preset time; feature extraction is performed on all the second mobile phone signaling data to obtain the second coordinate data of all the target residents in the target area within the second preset time; the migration frequency, the average migration distance and the geographical distance of all the target residents are calculated based on all the first coordinate data and all the second coordinate data; the migration frequency, the average migration distance and the geographical distance are input into the deep learning model for training to obtain a population migration model.

[0066] Among them, a deep learning model is constructed, and the first coordinate data is extracted according to the first mobile phone signaling data of the residents in the target area during the pre-disaster period, and the second coordinate data is extracted according to the second mobile phone signaling data of the residents in the target area during the disaster period, so as to calculate the migration frequency, average migration distance and geographical distance in the target area (that is, the geographical distance of each resident migrating from the starting position in the target area to the destination area), and input into the constructed deep learning model for training to obtain a population migration model for generating population migration flow. In the training process, cross entropy can be used as a loss function, and multiple cycles (for example, 20 cycles) of training are performed. The optimizer selects the RMSProp (Root Mean Square Propagation) algorithm, and the momentum of the training process can be set to 0.9, the learning rate is 5×10^-6, and the batch size is 64 starting positions. Furthermore, a negative sampling method can be used to randomly select up to 512 target positions at each starting position, thereby reducing the training time.

[0067] Furthermore, during the training process, residents’ mobile phone signaling data, community spatial information, and socioeconomic data are integrated to establish a multi-dimensional and multi-level disaster population migration flow prediction model, making the prediction results more accurate and reliable.

[0068] Step S50: obtaining a plurality of current mobile phone signaling data of a plurality of residents in the target area, and inputting the data into the population migration model, and outputting the population migration flow in the target area.

[0069] Specifically, multiple current mobile phone signaling data of multiple residents in the target area are obtained, and the current coordinate information in each current mobile phone signaling data is extracted, wherein the current coordinate information represents the current position of the resident corresponding to the current mobile phone signaling data; each current coordinate information is input into the population migration model for prediction, and the population migration flow of the target area is output, wherein the population migration flow represents the predicted population number of residents in the target area migrating to the destination area.

[0070] According to the current mobile phone signaling data of residents in the target area, the input feature vector (current coordinate information) is constructed, including the feature vector of the starting position, the feature vector of the destination position and the geographical distance between the two. After inputting into the population migration model, the population migration flow of the target area is output, such as Figure 2 As shown; the generated migration flow data can provide accurate data support for urban emergency management and post-disaster reconstruction, help formulate more effective disaster response strategies and resource allocation plans, and improve the city’s disaster response capabilities and resilience.

[0071] Furthermore, each current coordinate information is input into the feedforward neural network of the population migration model, and the feedforward neural network analyzes the current coordinate information through an activation function to output a predicted score of the target area; the population migration model converts the predicted score into the population migration flow through a preset function and outputs it.

[0072] The model uses input features to calculate the migration scores of each destination in a specific area, and outputs a migration score vector that can be used for subsequent migration flow analysis or probability conversion. Specifically, the model calculation process includes three steps: 1. Constructing the input feature vector; 2. The Kolmogorov–Arnold network in the model realizes complex nonlinear mapping through single variable function decomposition and combination; 3. The migration scores output by the model are further normalized or ranked as needed to analyze the trend of migration flow.

[0073] Among them, the feedforward neural network is built using the Kolmogorov–Arnold network (Kolmogorov-ArnoldNetworks, a neural network model designed based on the Kolmogorov-Arnold representation theorem), with 15 hidden layers, of which the bottom 6 layers are 256 dimensions and the remaining 9 layers are 128 dimensions. Among them, the width parameter (width parameter) of the Kolmogorov–Arnold network can be set to [input_dim, 128, 64, n_destinations], the grid parameter can be set to 5 to control the complexity of the edge function, and the k parameter can be set to 3 to control the number of connections for each node.

[0074] After the feedforward neural network outputs the predicted score of the target area, an activation function (such as softmax, normalized exponential function) is used to convert the predicted score into a probability value to generate a population migration flow. The activation function is a learnable edge function, and each edge uses an independent activation function. The network finally outputs a migration score vector s = [s1, s2, ..., sn] for each destination. These scores reflect the attractiveness or migration trend of each destination.

[0075] The present invention integrates multi-source data and deep learning technology to construct a population migration model for predicting population migration flows. It can accurately capture the dynamic changes of residents during disasters, and more finely characterize the migration flows of different communities or regions within the city, thereby improving data accuracy and timeliness, and achieving accurate prediction of urban residents' population migration flows during disasters.

[0076] Furthermore, if Figure 3 As shown, based on the above-mentioned population migration flow generation method based on deep learning, the present invention also provides a population migration flow generation system based on deep learning, wherein the population migration flow generation system based on deep learning includes:

[0077] The data preprocessing module 51 is used to obtain multiple initial mobile phone signaling data of each resident in the target area, screen out multiple target residents according to all the initial mobile phone signaling data, and preprocess all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident;

[0078] A data classification module 52 is used to obtain the meteorological data of the target area, classify all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtain a plurality of third mobile phone signaling data of all the target residents;

[0079] A model building module 53 is used to build a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, build a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, and obtain multi-source data, and build a deep learning model according to the multi-source data, the first matrix and the second matrix;

[0080] A model training module 54 is used to extract features from the plurality of first mobile phone signaling data and the plurality of second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographical distances, and input all the feature vectors and all the geographical distances into the deep learning model for training to obtain a population migration model;

[0081] The migration prediction module 55 is used to obtain a plurality of current mobile phone signaling data of a plurality of residents in the target area, input the data into the population migration model, and output the population migration flow in the target area.

[0082] Furthermore, if Figure 4 As shown, based on the above-mentioned deep learning-based population migration flow generation method and system, the present invention also provides a terminal accordingly, and the terminal includes a processor 10, a memory 20 and a display 30. Figure 4Only some components of the terminal are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0083] The memory 20 may be an internal storage unit of the terminal in some embodiments, such as a hard disk or memory of the terminal. The memory 20 may also be an external storage device of the terminal in other embodiments, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the terminal. Further, the memory 20 may also include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code of the installation terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, a population migration flow generation program 40 based on deep learning is stored on the memory 20, and the population migration flow generation program 40 based on deep learning can be executed by the processor 10, thereby realizing the population migration flow generation method based on deep learning in the present application.

[0084] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run the program code or process data stored in the memory 20, such as executing the deep learning-based population migration flow generation method.

[0085] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface. The components 10-30 of the terminal communicate with each other via a system bus.

[0086] In one embodiment, when the processor 10 executes the population migration flow generation program 40 based on deep learning in the memory 20, the following steps are implemented:

[0087] Acquire multiple initial mobile phone signaling data of each resident in the target area, screen out multiple target residents according to all the initial mobile phone signaling data, and pre-process all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident;

[0088] Acquire the meteorological data of the target area, divide all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and acquire a plurality of third mobile phone signaling data of all the target residents;

[0089] Constructing a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, constructing a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, acquiring multi-source data, and constructing a deep learning model according to the multi-source data, the first matrix, and the second matrix;

[0090] Performing feature extraction on a plurality of the first mobile phone signaling data and a plurality of the second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographical distances, and inputting all the feature vectors and all the geographical distances into the deep learning model for training to obtain a population migration model;

[0091] A plurality of current mobile phone signaling data of a plurality of residents in the target area are obtained and input into the population migration model, and the population migration flow in the target area is output.

[0092] The method of obtaining multiple initial mobile phone signaling data of each resident in the target area, screening out multiple target residents according to all the initial mobile phone signaling data, and preprocessing all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident specifically includes:

[0093] Obtain all initial mobile phone signaling data of all residents in the target area within a preset time. If there are multiple complete initial mobile phone signaling data of a resident, the resident is taken as the target resident, wherein the complete initial mobile phone signaling data means that the initial mobile phone signaling data of the target resident every day within the preset time and within the target area is complete;

[0094] The formats of all initial mobile phone signaling data of all the target residents are uniformly processed, and the data with abnormal activity trajectories are deleted to obtain multiple mobile phone signaling data of each target resident within a preset time.

[0095] The step of obtaining the meteorological data of the target area, dividing all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtaining a plurality of third mobile phone signaling data of all the target residents specifically includes:

[0096] Acquire meteorological data of the target area within the preset time, and divide the preset time into a first preset time and a second preset time according to the meteorological data;

[0097] Marking all mobile phone signaling data within the first preset time as a plurality of first mobile phone signaling data, and marking all mobile phone signaling data within the second preset time as a plurality of second mobile phone signaling data;

[0098] A plurality of third mobile phone signaling data of all the target residents are obtained, wherein the third mobile phone signaling data represent the mobile phone signaling data of the target residents after they migrate to the destination area after the preset time.

[0099] Among them, the constructing a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, constructing a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, and acquiring multi-source data, and constructing a deep learning model according to the multi-source data, the first matrix and the second matrix, specifically includes:

[0100] Determine the first target number of the target residents within the target area within the first preset time according to all the first mobile phone signaling data;

[0101] Determine the second target number of the target residents within the target area within the second preset time according to all the second mobile phone signaling data;

[0102] Determine the third target number of the target residents in the target area after the preset time according to all the third mobile phone signaling data;

[0103] Acquire multi-source data in the target area, construct a first matrix according to the first target number of residents and the third target number of residents, and construct a second matrix according to the second target number of residents and the third target number of residents;

[0104] A deep learning model is constructed based on the multi-source data, the first matrix and the second matrix.

[0105] Wherein, the characteristic vector includes: migration frequency and average migration distance;

[0106] The feature extraction of the plurality of the first mobile phone signaling data and the plurality of the second mobile phone signaling data is performed to obtain a plurality of feature vectors and a plurality of geographical distances, and all the feature vectors and all the geographical distances are input into the deep learning model for training to obtain a population migration model, specifically including:

[0107] Performing feature extraction on all the first mobile phone signaling data to obtain first coordinate data of all the target residents in the target area within the first preset time;

[0108] Performing feature extraction on all the second mobile phone signaling data to obtain second coordinate data of all the target residents in the target area within the second preset time;

[0109] Calculate the migration frequency, the average migration distance and the geographical distance of all the target residents according to all the first coordinate data and all the second coordinate data;

[0110] The migration frequency, the average migration distance and the geographical distance are input into the deep learning model for training to obtain a population migration model.

[0111] The step of obtaining a plurality of current mobile phone signaling data of a plurality of residents in the target area, inputting the data into the population migration model, and outputting the population migration flow in the target area specifically includes:

[0112] Acquire multiple current mobile phone signaling data of multiple residents in the target area, and extract current coordinate information from each current mobile phone signaling data, wherein the current coordinate information indicates the current position of the resident corresponding to the current mobile phone signaling data;

[0113] Each current coordinate information is input into the population migration model for prediction, and the population migration flow of the target area is output, wherein the population migration flow represents the predicted population number of residents in the target area migrating to the destination area.

[0114] The step of inputting each current coordinate information into the population migration model for prediction and outputting the population migration flow of the target area specifically includes:

[0115] Input each of the current coordinate information into the feedforward neural network of the population migration model, the feedforward neural network analyzes the current coordinate information through an activation function, and outputs a predicted score of the target area;

[0116] The population migration model converts the predicted score into the population migration flow through a preset function and outputs the result.

[0117] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a population migration flow generation program based on deep learning, and when the population migration flow generation program based on deep learning is executed by a processor, the steps of the population migration flow generation method based on deep learning as described above are implemented.

[0118] In summary, the present invention provides a method for generating a population migration flow based on deep learning and related equipment, the method comprising: obtaining multiple initial mobile phone signaling data of each resident in the target area, screening out multiple target residents based on all the initial mobile phone signaling data, and preprocessing all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident; obtaining meteorological data of the target area, dividing all the mobile phone signaling data into multiple first mobile phone signaling data and multiple second mobile phone signaling data according to the meteorological data, and obtaining multiple third mobile phone signaling data of all the target residents; filtering out multiple target residents based on the multiple first mobile phone signaling data and multiple second mobile phone signaling data; ... The first matrix is ​​constructed based on the third mobile phone signaling data, the second matrix is ​​constructed based on multiple second mobile phone signaling data and multiple third mobile phone signaling data, and multi-source data is obtained, and a deep learning model is constructed based on the multi-source data, the first matrix and the second matrix; feature extraction is performed on multiple first mobile phone signaling data and multiple second mobile phone signaling data to obtain multiple feature vectors and multiple geographical distances, and all the feature vectors and all the geographical distances are input into the deep learning model for training to obtain a population migration model; multiple current mobile phone signaling data of multiple residents in the target area are obtained, and input into the population migration model, and the population migration flow in the target area is output. The present invention constructs a population migration model for predicting population migration flow by integrating multi-source data and deep learning technology, which can accurately capture the dynamic changes of residents during disasters, and more finely characterize the migration flow of different communities or regions within the city, thereby improving data accuracy and timeliness, and realizing accurate prediction of urban residents' population migration flow during disasters.

[0119] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or terminal including the element.

[0120] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the program can be stored in a computer-readable storage medium that can be read by a computer, and the program can include the processes of the above-mentioned method embodiments when executed. The computer-readable storage medium can be a memory, a disk, an optical disk, etc.

[0121] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A method for generating population migration flow based on deep learning, characterized in that: The population migration flow generation method based on deep learning includes: Acquire multiple initial mobile phone signaling data of each resident in the target area, screen out multiple target residents according to all the initial mobile phone signaling data, and pre-process all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident; Acquire the meteorological data of the target area, divide all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and acquire a plurality of third mobile phone signaling data of all the target residents; Constructing a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, constructing a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, acquiring multi-source data, and constructing a deep learning model according to the multi-source data, the first matrix, and the second matrix; Performing feature extraction on a plurality of the first mobile phone signaling data and a plurality of the second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographical distances, and inputting all the feature vectors and all the geographical distances into the deep learning model for training to obtain a population migration model; A plurality of current mobile phone signaling data of a plurality of residents in the target area are obtained and input into the population migration model, and the population migration flow in the target area is output.

2. The method for generating population migration flow based on deep learning according to claim 1, characterized in that: The step of obtaining multiple initial mobile phone signaling data of each resident in the target area, screening out multiple target residents according to all the initial mobile phone signaling data, and preprocessing all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident specifically includes: Obtain all initial mobile phone signaling data of all residents in the target area within a preset time. If there are multiple complete initial mobile phone signaling data of a resident, the resident is taken as the target resident, wherein the complete initial mobile phone signaling data means that the initial mobile phone signaling data of the target resident every day within the preset time and within the target area is complete; The formats of all initial mobile phone signaling data of all the target residents are uniformly processed, and the data with abnormal activity trajectories are deleted to obtain multiple mobile phone signaling data of each target resident within a preset time.

3. The method for generating population migration flow based on deep learning according to claim 2, characterized in that: The step of obtaining the meteorological data of the target area, dividing all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtaining a plurality of third mobile phone signaling data of all the target residents specifically includes: Acquire meteorological data of the target area within the preset time, and divide the preset time into a first preset time and a second preset time according to the meteorological data; Marking all mobile phone signaling data within the first preset time as a plurality of first mobile phone signaling data, and marking all mobile phone signaling data within the second preset time as a plurality of second mobile phone signaling data; A plurality of third mobile phone signaling data of all the target residents are obtained, wherein the third mobile phone signaling data represent the mobile phone signaling data of the target residents after they migrate to the destination area after the preset time.

4. The method for generating population migration flow based on deep learning according to claim 3, characterized in that: The step of constructing a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, constructing a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, acquiring multi-source data, and constructing a deep learning model according to the multi-source data, the first matrix, and the second matrix specifically includes: Determine the first target number of the target residents within the target area within the first preset time according to all the first mobile phone signaling data; Determine the second target number of the target residents within the target area within the second preset time according to all the second mobile phone signaling data; Determine the third target number of the target residents in the target area after the preset time according to all the third mobile phone signaling data; Acquire multi-source data in the target area, construct a first matrix according to the first target number of residents and the third target number of residents, and construct a second matrix according to the second target number of residents and the third target number of residents; A deep learning model is constructed based on the multi-source data, the first matrix and the second matrix.

5. The method for generating population migration flow based on deep learning according to claim 4, characterized in that: The feature vector includes: migration frequency and average migration distance; The feature extraction of the plurality of the first mobile phone signaling data and the plurality of the second mobile phone signaling data is performed to obtain a plurality of feature vectors and a plurality of geographical distances, and all the feature vectors and all the geographical distances are input into the deep learning model for training to obtain a population migration model, specifically including: Performing feature extraction on all the first mobile phone signaling data to obtain first coordinate data of all the target residents in the target area within the first preset time; Performing feature extraction on all the second mobile phone signaling data to obtain second coordinate data of all the target residents in the target area within the second preset time; Calculate the migration frequency, the average migration distance and the geographical distance of all the target residents according to all the first coordinate data and all the second coordinate data; The migration frequency, the average migration distance and the geographical distance are input into the deep learning model for training to obtain a population migration model.

6. The method for generating population migration flow based on deep learning according to claim 1, characterized in that: The acquiring of a plurality of current mobile phone signaling data of a plurality of residents in the target area, inputting the data into the population migration model, and outputting the population migration flow in the target area specifically includes: Acquire multiple current mobile phone signaling data of multiple residents in the target area, and extract current coordinate information from each current mobile phone signaling data, wherein the current coordinate information indicates the current position of the resident corresponding to the current mobile phone signaling data; Each current coordinate information is input into the population migration model for prediction, and the population migration flow of the target area is output, wherein the population migration flow represents the predicted population number of residents in the target area migrating to the destination area.

7. The method for generating population migration flow based on deep learning according to claim 6, characterized in that: The step of inputting each current coordinate information into the population migration model for prediction and outputting the population migration flow of the target area specifically includes: Input each of the current coordinate information into the feedforward neural network of the population migration model, the feedforward neural network analyzes the current coordinate information through an activation function, and outputs a predicted score of the target area; The population migration model converts the predicted score into the population migration flow through a preset function and outputs the result.

8. A population migration flow generation system based on deep learning, characterized in that: The population migration flow generation system based on deep learning includes: A data preprocessing module is used to obtain multiple initial mobile phone signaling data of each resident in the target area, screen out multiple target residents according to all the initial mobile phone signaling data, and preprocess all the initial mobile phone signaling data of all the target residents to obtain multiple mobile phone signaling data of each target resident; A data classification module, used for obtaining meteorological data of the target area, classifying all the mobile phone signaling data into a plurality of first mobile phone signaling data and a plurality of second mobile phone signaling data according to the meteorological data, and obtaining a plurality of third mobile phone signaling data of all the target residents; A model building module, used to build a first matrix according to a plurality of the first mobile phone signaling data and a plurality of the third mobile phone signaling data, build a second matrix according to a plurality of the second mobile phone signaling data and a plurality of the third mobile phone signaling data, and obtain multi-source data, and build a deep learning model according to the multi-source data, the first matrix and the second matrix; A model training module, used for extracting features from a plurality of the first mobile phone signaling data and a plurality of the second mobile phone signaling data to obtain a plurality of feature vectors and a plurality of geographical distances, and inputting all the feature vectors and all the geographical distances into the deep learning model for training to obtain a population migration model; The migration prediction module is used to obtain multiple current mobile phone signaling data of multiple residents in the target area, input the data into the population migration model, and output the population migration flow in the target area.

9. A terminal, characterized in that: The terminal includes: a memory, a processor, and a deep learning-based population migration flow generation program stored in the memory and executable on the processor. When the deep learning-based population migration flow generation program is executed by the processor, the steps of the deep learning-based population migration flow generation method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a population migration flow generation program based on deep learning, and when the population migration flow generation program based on deep learning is executed by a processor, the steps of the population migration flow generation method based on deep learning as described in any one of claims 1 to 7 are implemented.

Images