Tourist number prediction method and system based on big data analysis

Abstract

The invention relates to the field of big data analysis, in particular to a tourist number prediction method and system based on big data analysis. The method comprises the steps of obtaining real-time tourism associated data in a database, and performing data preprocessing on the real-time tourism associated data to obtain target training tourism associated data; a target CNN-Transform mixed human traffic data prediction model is established; inputting the target training tourism associated data into a target CNN-Transform mixed visitor flow rate data prediction model for identification and prediction, obtaining associated features of the training tourism associated data by using a CNN convolutional neural network, obtaining global information of the training tourism associated data by Transform, and obtaining real-time tourism number prediction data; if the visitor flow rate is higher than the limit visitor flow rate data of the scenic spot, generating an alternative scenic spot tourism scheme, and displaying the alternative scenic spot tourism scheme to a mobile terminal of the user; advanced encryption algorithms and technical means are adopted, the security and privacy of data are ensured, and the real situation of the tourism market is reflected more accurately.

Description

Technical Field

[0001] The present invention relates to the field of big data analysis, and in particular to a method and system for predicting the number of tourists based on big data analysis. Background Art

[0002] As an important decision-making basis for the tourism industry, tourist number forecasting is of great significance for the rational allocation of tourism resources, precise marketing of the tourism market, and the scientific formulation of tourism policies. Although traditional tourist number forecasting methods, such as linear regression based on historical data and time series analysis, can reflect the changing trend of tourist number to a certain extent, they are often difficult to cope with the complex and changing tourism market environment due to their single data source and simple analysis methods. In addition, along with the application and mining of big data technology, the application of tourism data is usually accompanied by the leakage of personal privacy, which has a certain risk of public privacy information leakage. Therefore, how to improve the accuracy of tourist number forecasting while ensuring that the privacy of the public is not leaked is a technical problem that needs to be solved at this stage. Summary of the invention

[0003] The purpose of the present invention is to solve the above problems and to design a tourist number prediction method and system based on big data analysis.

[0004] To achieve the above object, the technical solution of the present invention is that, further, in the above-mentioned tourist number prediction method based on big data analysis, the tourist number prediction method comprises the following steps:

[0005] Acquire real-time tourism-related data in a database, perform data preprocessing on the real-time tourism-related data, and obtain training tourism-related data, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data;

[0006] Encrypting the training tourism related data by using RSA asymmetric encryption algorithm to obtain target training tourism related data;

[0007] Establish a CNN-Transformer hybrid human flow data prediction model, use Next-Vi T as a hybrid framework of the CNN-Transformer hybrid human flow data prediction model, and obtain a target CNN-Transformer hybrid human flow data prediction model;

[0008] The target training tourism associated data is input into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction, and the associated features of the training tourism associated data are obtained by using the CNN convolutional neural network. Transfome obtains the global information of the training tourism associated data to obtain real-time tourist number prediction data;

[0009] Based on the real-time tourist number forecast data, determine whether the flow of people at the scenic spot is higher than the preset limit flow of people at the scenic spot; if it is higher than the limit flow of people at the scenic spot, generate an alternative scenic spot travel plan, and display the alternative scenic spot travel plan to the user's mobile terminal;

[0010] The real-time tourist number feedback data of the user is obtained, and the real-time tourist number feedback data is compared with the real-time tourist number prediction data. If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid traffic data prediction model for training.

[0011] Furthermore, in the above-mentioned tourist number prediction method based on big data analysis, the real-time tourism-related data in the database is obtained, and the real-time tourism-related data is preprocessed to obtain training tourism-related data, and the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data, including:

[0012] Acquire real-time tourism-related data in the database based on the user's real-time scenic spot query needs, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data;

[0013] Segmenting the real-time tourism related data to obtain multiple layers of real-time tourism related data;

[0014] Performing binary conversion on the multi-layer real-time tourism related data to obtain binary real-time tourism related data;

[0015] The binary real-time tourism-related data is classified into data sets to obtain training tourism-related data.

[0016] Furthermore, in the above-mentioned tourist number prediction method based on big data analysis, it is characterized in that the training tourism associated data is encrypted using the RSA asymmetric encryption algorithm to obtain the target training tourism associated data, including:

[0017] Obtain training tourism association data, randomly select two prime numbers, and record the two prime numbers as P and Q respectively;

[0018] Set N to the product of two prime numbers, where N=P*Q, and N is a part of the public key and the private key;

[0019] Let M be the product of two prime numbers minus one, where M = (P-1)*(Q-1);

[0020] Set E as part of the public key, and E and M are relatively prime, where E and M are relatively prime odd numbers, and 1<E<M;

[0021] Let D be part of the private key, and D is the modular inverse element of E with respect to M, where E*Dmod M=1;

[0022] Assume P is the training tourism associated data, and C is the target training tourism associated data, where C = P^e mod n.

[0023] Further, in the above-mentioned tourist number prediction method based on big data analysis, the CNN-Transformer hybrid passenger flow data prediction model is established, and Next-Vi T is used as a hybrid framework of the CNN-Transformer hybrid passenger flow data prediction model to obtain a target CNN-Transformer hybrid passenger flow data prediction model, including:

[0024] Establishing a CNN-Transformer hybrid human flow data prediction model, wherein the CNN-Transformer hybrid human flow data prediction model at least includes an initial layer, an input layer, an output layer, a fully connected layer, and an average pooling layer;

[0025] A linear embedding layer of a patch is placed before each layer of the CNN-Transformer hybrid crowd flow data prediction model, and the linear embedding layer is used for channel matching and extracting patch information;

[0026] Using Next-V i T as a hybrid framework of the CNN-Transformer hybrid crowd flow data prediction model;

[0027] The NCB module and the NTB module in the CNN-Transformer hybrid human flow data prediction model are stacked in a (N+1)*L hybrid manner to obtain a target CNN-Transformer hybrid human flow data prediction model.

[0028] Furthermore, in the above-mentioned tourist number prediction method based on big data analysis, the target training tourism associated data is input into the target CNN-Transformer hybrid human flow data prediction model for identification and prediction, and the CNN convolutional neural network is used to obtain the associated features of the training tourism associated data, and Transfome obtains the global information of the training tourism associated data to obtain real-time tourist number prediction data, including:

[0029] Input the target training tourism-related data into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction;

[0030] Using a CNN convolutional neural network to obtain the associated features of the training tourism associated data, Transfome obtains the global information of the training tourism associated data;

[0031] Based on the MHCA multi-head convolutional attention mechanism as a token mixer for the target CNN-Transformer hybrid crowd flow data prediction model;

[0032] Performing deep learning on the data information of multiple subspaces in the target CNN-Transformer hybrid passenger flow data prediction model, acquiring data information from multiple parallel subspaces, and obtaining real-time tourist number prediction data;

[0033] The real-time tourist number prediction data at least includes: tourist vehicle flow data and tourist flow data.

[0034] Furthermore, in the above-mentioned tourist number prediction method based on big data analysis, the method of judging whether the flow of tourists at the scenic spot is higher than the preset limit flow of tourists at the scenic spot based on the real-time tourist number prediction data, and generating an alternative scenic spot travel plan if it is higher than the limit flow of tourists at the scenic spot, and displaying the alternative scenic spot travel plan to the user's mobile terminal includes:

[0035] Acquire the number of parking spaces and the number of tickets sold for the scenic spot, and generate the maximum passenger flow data of the scenic spot based on the number of parking spaces and the number of tickets sold for the scenic spot;

[0036] Determine whether the crowd flow data of the scenic spot is higher than the preset limit crowd flow data of the scenic spot according to the real-time tourist number prediction data;

[0037] If the real-time tourist number forecast data is higher than the maximum tourist flow data of the scenic spot, then obtaining a target scenic spot in the database that is similar to the real-time scenic spot type, and obtaining the tourist flow data of the target scenic spot;

[0038] If the passenger flow data of the target scenic spot is less than the real-time tourist number forecast data, an alternative scenic spot travel plan is generated and transmitted to the user's mobile terminal.

[0039] Furthermore, in the above-mentioned tourist number prediction method based on big data analysis, the real-time tourist number feedback data of the user is obtained, and the real-time tourist number feedback data is compared with the real-time tourist number prediction data. If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid traffic data prediction model for training, including:

[0040] Acquiring scenic spot survey feedback data based on the user's mobile terminal, wherein the scenic spot survey feedback data at least includes: scenic spot real-time image data, scenic spot number real-time survey data, and scenic spot traffic flow data;

[0041] Generating real-time tourist number feedback data based on the scenic spot survey feedback data, and comparing the real-time tourist number feedback data with the real-time tourist number forecast data;

[0042] If the difference between the two data is less than 20%, the parameters of the target CNN-Transformer hybrid human flow data prediction model are adjusted;

[0043] If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid passenger flow data prediction model for training.

[0044] The technical solution of the present invention to achieve the above-mentioned purpose is, further, a tourist number prediction system based on big data analysis, the tourist number prediction system comprising the following modules:

[0045] A data processing module is used to obtain real-time tourism-related data in a database, perform data preprocessing on the real-time tourism-related data, and obtain training tourism-related data, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data;

[0046] A data encryption module, used for encrypting the training tourism related data by using an RSA asymmetric encryption algorithm to obtain target training tourism related data;

[0047] A model building module is used to establish a CNN-Transformer hybrid human flow data prediction model, using Next-Vi T as a hybrid framework of the CNN-Transformer hybrid human flow data prediction model to obtain a target CNN-Transformer hybrid human flow data prediction model;

[0048] The number of people prediction module is used to input the target training tourism associated data into the target CNN-Transformer hybrid human flow data prediction model for identification and prediction, use the CNN convolutional neural network to obtain the associated features of the training tourism associated data, and Transformer obtains the global information of the training tourism associated data to obtain real-time tourist number prediction data;

[0049] A travel selection module, for determining whether the flow of people at a scenic spot is higher than a preset limit flow of people at the scenic spot based on the real-time tourist number forecast data, and generating an alternative scenic spot travel plan if it is higher than the limit flow of people at the scenic spot, and displaying the alternative scenic spot travel plan to a user's mobile terminal;

[0050] The prediction comparison module is used to obtain the user's real-time tourist number feedback data, compare the real-time tourist number feedback data with the real-time tourist number prediction data, and if the difference between the two data is greater than 20%, input the real-time tourist number feedback data into the target CNN-Transformer hybrid traffic data prediction model for training.

[0051] Furthermore, in the above-mentioned tourist number prediction system based on big data analysis, the data processing module includes the following submodules:

[0052] An acquisition submodule is used to acquire real-time tourism-related data in a database based on the user's real-time scenic spot query requirements, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot spot social media data;

[0053] A segmentation module, used for segmenting the real-time tourism related data to obtain multiple layers of real-time tourism related data;

[0054] A binary submodule, used for performing binary conversion on the multi-layer real-time tourism related data to obtain binary real-time tourism related data;

[0055] A submodule is obtained, which is used to classify the binary real-time tourism related data into data sets to obtain training tourism related data.

[0056] Furthermore, in the above-mentioned tourist number prediction system based on big data analysis, the model building module includes the following submodules:

[0057] Establish a submodule for establishing a CNN-Transformer hybrid human flow data prediction model, wherein the CNN-Transformer hybrid human flow data prediction model includes at least an initial layer, an input layer, an output layer, a fully connected layer, and an average pooling layer;

[0058] An embedding submodule, used for placing a linear embedding layer of a patch before each layer of the CNN-Transformer hybrid pedestrian flow data prediction model, wherein the linear embedding layer performs user channel matching and extracts patch information;

[0059] A hybrid submodule, for using Next-V i T as a hybrid framework of the CNN-Transformer hybrid crowd flow data prediction model;

[0060] The stacking submodule is used to stack the NCB module and the NTB module in the CNN-Transformer hybrid human flow data prediction model in a (N+1)*L hybrid manner to obtain a target CNN-Transformer hybrid human flow data prediction model.

[0061] Its beneficial effect is that by using advanced encryption algorithms and technical means in the process of collecting, storing and analyzing tourism data, the security and privacy of data are ensured. It not only protects consumers' personal information from being leaked, but also maintains the business secrets of tourism companies. And it realizes a comprehensive analysis of the tourism market. Whether it is the search engine query volume, social media user comments, or online travel platform booking data, they are all included in the analysis scope, thus more accurately reflecting the real situation of the tourism market. This comprehensive data analysis makes the prediction results more accurate and provides strong decision-making support for tourism companies. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.

[0063] Figure 1 This is a schematic diagram of a first embodiment of a tourist number prediction method based on big data analysis in an embodiment of the present invention;

[0064] Figure 2 A schematic diagram of a second embodiment of a tourist number prediction method based on big data analysis in an embodiment of the present invention;

[0065] Figure 3A schematic diagram of a third embodiment of a tourist number prediction method based on big data analysis in an embodiment of the present invention;

[0066] Figure 4 This is a schematic diagram of a first embodiment of a tourist number prediction system based on big data analysis in an embodiment of the present invention. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0068] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0069] The present invention will be described in detail below in conjunction with the accompanying drawings. Figure 1 As shown, a tourist number prediction method based on big data analysis includes the following steps:

[0070] Step 101: obtaining real-time tourism-related data in a database, performing data preprocessing on the real-time tourism-related data, and obtaining training tourism-related data, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data;

[0071] Specifically, in this embodiment, real-time tourism-related data in the database is obtained based on the user's real-time scenic spot query needs, and the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data; the real-time tourism-related data is segmented to obtain multi-layer real-time tourism-related data; the multi-layer real-time tourism-related data is binary-converted to obtain binary real-time tourism-related data; the binary real-time tourism-related data is classified into data sets to obtain training tourism-related data.

[0072] Step 102: Encrypt the training tourism related data using the RSA asymmetric encryption algorithm to obtain target training tourism related data;

[0073] Specifically, in this embodiment, training tourism associated data is obtained, two prime numbers are randomly selected, and the two prime numbers are recorded as P and Q respectively; N is set to be the product of the two prime numbers, where N=P*Q, and N is a part of the public key and the private key; M is set to be the product of the two prime numbers minus one, where M=(P-1)*(Q-1); E is set as a part of the public key, and E and M are mutually prime, where E and M are mutually prime odd numbers, and 1<E<M; D is set as a part of the private key, and D is the modular inverse element of E with respect to M, where E*Dmod M=1; P is set to be the training tourism associated data, and C is the target training tourism associated data, where C=P^e mod n.

[0074] Step 103: Establish a CNN-Transformer hybrid crowd flow data prediction model, use Next-V i T as a hybrid framework of the CNN-Transformer hybrid crowd flow data prediction model, and obtain a target CNN-Transformer hybrid crowd flow data prediction model;

[0075] Specifically, in this embodiment, a CNN-Transformer hybrid crowd flow data prediction model is established, and the CNN-Transformer hybrid crowd flow data prediction model includes at least an initial layer, an input layer, an output layer, a fully connected layer, and an average pooling layer; a patch linear embedding layer is placed before each layer of the CNN-Transformer hybrid crowd flow data prediction model, and the linear embedding layer is used for channel matching and extracting patch information; Next-Vi T is used as a hybrid framework of the CNN-Transformer hybrid crowd flow data prediction model; the NCB module and the NTB module in the CNN-Transformer hybrid crowd flow data prediction model are stacked in a mixed manner of (N+1)*L to obtain the target CNN-Transformer hybrid crowd flow data prediction model.

[0076] Step 104: input the target training tourism-related data into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction, use the CNN convolutional neural network to obtain the associated features of the training tourism-related data, and Transformer to obtain the global information of the training tourism-related data to obtain real-time tourist number prediction data;

[0077] Specifically, in this embodiment, the target training tourism associated data is input into the target CNN-Transformer mixed passenger flow data prediction model for identification and prediction; the CNN convolutional neural network is used to obtain the associated features of the training tourism associated data, and Transfome obtains the global information of the training tourism associated data; the MHCA multi-head convolutional attention mechanism is used as a token mixer for the target CNN-Transformer mixed passenger flow data prediction model; deep learning is performed on the data information of multiple subspaces in the target CNN-Transformer mixed passenger flow data prediction model, and data information is obtained from multiple parallel subspaces to obtain real-time tourist number prediction data; the real-time tourist number prediction data at least includes: tourist vehicle flow data and tourist flow data.

[0078] Step 105: judging whether the flow of people at the scenic spot is higher than the preset limit flow of people at the scenic spot based on the real-time tourist number forecast data; if it is higher than the limit flow of people at the scenic spot, generating an alternative scenic spot travel plan, and displaying the alternative scenic spot travel plan to the user's mobile terminal;

[0079] Specifically, in this embodiment, the number of parking spaces and the number of tickets sold at the scenic spot are obtained, and the maximum passenger flow data of the scenic spot is generated based on the number of parking spaces and the number of tickets sold at the scenic spot; the passenger flow data of the scenic spot is determined according to the real-time tourist number prediction data whether it is higher than the preset maximum passenger flow data of the scenic spot; if the real-time tourist number prediction data is higher than the maximum passenger flow data of the scenic spot, the target scenic spot similar to the real-time tourist attraction type is obtained from the database, and the passenger flow data of the target scenic spot is obtained; if the passenger flow data of the target scenic spot is less than the real-time tourist number prediction data, an alternative scenic spot travel plan is generated, and the alternative scenic spot travel plan is transmitted to the user's mobile terminal.

[0080] Step 106: obtain the user's real-time tourist number feedback data, compare the real-time tourist number feedback data with the real-time tourist number prediction data, and if the difference between the two data is greater than 20%, input the real-time tourist number feedback data into the target CNN-Transformer hybrid traffic data prediction model for training.

[0081] Specifically, in this embodiment, scenic spot survey feedback data is obtained based on the user's mobile terminal, and the scenic spot survey feedback data at least includes: real-time image data of the scenic spot, real-time survey data on the number of people at the scenic spot, and traffic data on the scenic spot; real-time tourist number feedback data is generated based on the scenic spot survey feedback data, and the real-time tourist number feedback data and the real-time tourist number prediction data are compared; if the difference between the two data is less than 20%, the parameters of the target CNN-Transformer mixed traffic data prediction model are adjusted; if the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer mixed traffic data prediction model for training.

[0082] The beneficial effect is that by obtaining the real-time tourism related data in the database, the real-time tourism related data is preprocessed to obtain the training tourism related data, and the real-time tourism related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket booking data, and real-time hot social media data; the training tourism related data is encrypted by using the RSA asymmetric encryption algorithm to obtain the target training tourism related data; a CNN-Transformer hybrid passenger flow data prediction model is established, and the Next-Vi T is used as a hybrid framework of the CNN-Transformer hybrid traffic data prediction model to obtain the target CNN-Transformer hybrid traffic data prediction model; the target training tourism associated data is input into the target CNN-Transformer hybrid traffic data prediction model for identification and prediction, and the CNN convolutional neural network is used to obtain the associated features of the training tourism associated data, and Transfome obtains the global information of the training tourism associated data to obtain the real-time tourist number prediction data; based on the real-time tourist number prediction data, it is judged whether the traffic data of the scenic spot is higher than the preset scenic spot limit traffic data. If it is higher than the scenic spot limit traffic data, an alternative scenic spot tourism plan is generated, and the alternative scenic spot tourism plan is displayed to the user's mobile terminal; the user's real-time tourist number feedback data is obtained, and the real-time tourist number feedback data is compared with the real-time tourist number prediction data. If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid traffic data prediction model for training. In the process of collecting, storing and analyzing tourism data, advanced encryption algorithms and technical means are used to ensure the security and privacy of the data. It not only protects consumers' personal information from being leaked, but also maintains the commercial secrets of tourism enterprises. And a comprehensive analysis of the tourism market is achieved. Whether it is search engine query volume, social media user comments, or online travel platform booking data, they are all included in the analysis scope, thus more accurately reflecting the real situation of the tourism market. This comprehensive data analysis makes the prediction results more accurate and provides strong decision-making support for tourism companies.

[0083] In this embodiment, please refer to Figure 2 In a second embodiment of a tourist number prediction method based on big data analysis in an embodiment of the present invention, target training tourist associated data is input into a target CNN-Transformer hybrid human flow data prediction model for identification and prediction, and a CNN convolutional neural network is used to obtain associated features of the training tourist associated data. Transformer obtains global information of the training tourist associated data, and obtains real-time tourist number prediction data, including the following steps:

[0084] Step 201: input the target training tourism-related data into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction;

[0085] Step 202: using the CNN convolutional neural network to obtain the associated features of the training tourism associated data, Transfome obtains the global information of the training tourism associated data;

[0086] Step 203, using the MHCA multi-head convolutional attention mechanism as a token mixer for the target CNN-Transformer hybrid traffic data prediction model;

[0087] Step 204: perform deep learning on the data information of multiple subspaces in the target CNN-Transformer hybrid passenger flow data prediction model, obtain data information from multiple parallel subspaces, and obtain real-time tourist number prediction data;

[0088] Step 205: The real-time tourist number forecast data at least includes: tourist vehicle flow data and tourist flow data.

[0089] In this embodiment, please refer to Figure 3 In a third embodiment of a tourist number prediction method and system based on big data analysis in an embodiment of the present invention, it is determined whether the flow data of the scenic spot is higher than the preset limit flow data of the scenic spot based on the real-time tourist number prediction data, and if it is higher than the limit flow data of the scenic spot, an alternative scenic spot travel plan is generated, and the alternative scenic spot travel plan is displayed to the user's mobile terminal, including the following steps:

[0090] Step 301: Obtain the number of parking spaces and the number of tickets sold at the scenic spot, and generate the maximum passenger flow data of the scenic spot based on the number of parking spaces and the number of tickets sold at the scenic spot;

[0091] Step 302: judging whether the flow of people at the scenic spot is higher than the preset limit flow of people at the scenic spot according to the real-time tourist number forecast data;

[0092] Step 303: If the real-time tourist number forecast data is higher than the tourist attraction limit flow data, then obtain a target attraction in the database that is similar to the real-time tourist attraction type, and obtain the target attraction flow data;

[0093] Step 304: If the passenger flow data of the target scenic spot is less than the real-time tourist number forecast data, an alternative scenic spot travel plan is generated and transmitted to the user's mobile terminal.

[0094] The above describes a tourist number prediction method based on big data analysis provided by an embodiment of the present invention. The following describes a tourist number prediction system based on big data analysis according to an embodiment of the present invention. Figure 4 In one embodiment of the present invention, a tourist number prediction system includes:

[0095] A data processing module is used to obtain real-time tourism-related data in a database, perform data preprocessing on the real-time tourism-related data, and obtain training tourism-related data. The real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data;

[0096] A data encryption module is used to encrypt the training tourism related data using the RSA asymmetric encryption algorithm to obtain the target training tourism related data;

[0097] The model building module is used to build a CNN-Transformer hybrid human flow data prediction model, using Next-Vi T as the hybrid framework of the CNN-Transformer hybrid human flow data prediction model to obtain the target CNN-Transformer hybrid human flow data prediction model;

[0098] The number of people prediction module is used to input the target training tourism-related data into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction, and use the CNN convolutional neural network to obtain the associated features of the training tourism-related data. Transformer obtains the global information of the training tourism-related data to obtain real-time tourist number prediction data;

[0099] A travel selection module is used to determine whether the flow of tourists at a scenic spot is higher than the preset limit flow of tourists at the scenic spot based on the real-time tourist number forecast data. If it is higher than the limit flow of tourists at the scenic spot, an alternative scenic spot travel plan is generated and the alternative scenic spot travel plan is displayed to the user's mobile terminal;

[0100] The prediction comparison module is used to obtain the user's real-time tourist number feedback data, compare the real-time tourist number feedback data with the real-time tourist number prediction data, and if the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid traffic data prediction model for training.

[0101] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A tourist number prediction method based on big data analysis, characterized in that: The tourist number prediction method comprises the following steps: Acquire real-time tourism-related data in a database, perform data preprocessing on the real-time tourism-related data, and obtain training tourism-related data, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data; Obtain training tourism association data, randomly select two prime numbers, and record the two prime numbers as P and Q respectively; Set N to the product of two prime numbers, where N=P*Q, and N is a part of the public key and the private key; Let M be the product of two prime numbers minus one, where M = (P-1)*(Q-1); Set E as part of the public key, and E and M are relatively prime, where E and M are relatively prime odd numbers, and 1<E<M; Let D be part of the private key, and D is the modular inverse element of E with respect to M, where E*Dmod M=1; Assume P as the training tourism associated data, C as the target training tourism associated data, where C = P^e mod n, to obtain the target training tourism associated data; Establish a CNN-Transformer hybrid human flow data prediction model, use Next-ViT as a hybrid framework of the CNN-Transformer hybrid human flow data prediction model, and obtain a target CNN-Transformer hybrid human flow data prediction model; The target training tourism associated data is input into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction, and the associated features of the training tourism associated data are obtained by using the CNN convolutional neural network. Transfome obtains the global information of the training tourism associated data to obtain real-time tourist number prediction data; Based on the real-time tourist number forecast data, determine whether the flow of people at the scenic spot is higher than the preset limit flow of people at the scenic spot; if it is higher than the limit flow of people at the scenic spot, generate an alternative scenic spot travel plan, and display the alternative scenic spot travel plan to the user's mobile terminal; The real-time tourist number feedback data of the user is obtained, and the real-time tourist number feedback data is compared with the real-time tourist number prediction data. If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid traffic data prediction model for training.

2. A tourist number prediction method based on big data analysis as claimed in claim 1, characterized in that: The real-time tourism-related data in the database is obtained, and data preprocessing is performed on the real-time tourism-related data to obtain training tourism-related data, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data, including: Acquire real-time tourism-related data in the database based on the user's real-time scenic spot query needs, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data; Segmenting the real-time tourism related data to obtain multiple layers of real-time tourism related data; Performing binary conversion on the multi-layer real-time tourism related data to obtain binary real-time tourism related data; The binary real-time tourism-related data is classified into data sets to obtain training tourism-related data.

3. A tourist number prediction method based on big data analysis as claimed in claim 1, characterized in that: The method of establishing a CNN-Transformer hybrid human flow data prediction model and using Next-Vi T as a hybrid framework of the CNN-Transformer hybrid human flow data prediction model to obtain a target CNN-Transformer hybrid human flow data prediction model includes: Establishing a CNN-Transformer hybrid human flow data prediction model, wherein the CNN-Transformer hybrid human flow data prediction model at least includes an initial layer, an input layer, an output layer, a fully connected layer, and an average pooling layer; A linear embedding layer of a patch is placed before each layer of the CNN-Transformer hybrid crowd flow data prediction model, and the linear embedding layer is used for channel matching and extracting patch information; Using Next-Vi T as a hybrid framework for the CNN-Transformer hybrid crowd flow data prediction model; The NCB module and the NTB module in the CNN-Transformer hybrid human flow data prediction model are stacked in a (N+1)*L hybrid manner to obtain a target CNN-Transformer hybrid human flow data prediction model.

4. A tourist number prediction method based on big data analysis as claimed in claim 1, characterized in that: The target training tourism associated data is input into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction, and the associated features of the training tourism associated data are obtained by using the CNN convolutional neural network. Transfome obtains the global information of the training tourism associated data to obtain real-time tourist number prediction data, including: Input the target training tourism-related data into the target CNN-Transformer hybrid passenger flow data prediction model for identification and prediction; Using a CNN convolutional neural network to obtain the associated features of the training tourism associated data, Transfome obtains the global information of the training tourism associated data; Based on the MHCA multi-head convolutional attention mechanism as a token mixer for the target CNN-Transformer hybrid traffic data prediction model; Performing deep learning on the data information of multiple subspaces in the target CNN-Transformer hybrid passenger flow data prediction model, acquiring data information from multiple parallel subspaces, and obtaining real-time tourist number prediction data; The real-time tourist number prediction data at least includes: tourist vehicle flow data and tourist flow data.

5. The tourist number prediction method based on big data analysis as claimed in claim 1, characterized in that: The method of judging whether the flow of people at the scenic spot is higher than the preset limit flow of people at the scenic spot based on the real-time tourist number prediction data, and generating an alternative scenic spot travel plan if it is higher than the limit flow of people at the scenic spot, and displaying the alternative scenic spot travel plan to the user's mobile terminal includes: Acquire the number of parking spaces and the number of tickets sold for the scenic spot, and generate the maximum passenger flow data of the scenic spot based on the number of parking spaces and the number of tickets sold for the scenic spot; Determine whether the crowd flow data of the scenic spot is higher than the preset limit crowd flow data of the scenic spot according to the real-time tourist number prediction data; If the real-time tourist number forecast data is higher than the maximum tourist flow data of the scenic spot, then obtaining a target scenic spot in the database that is similar to the real-time scenic spot type, and obtaining the tourist flow data of the target scenic spot; If the passenger flow data of the target scenic spot is less than the real-time tourist number forecast data, an alternative scenic spot travel plan is generated and transmitted to the user's mobile terminal.

6. A tourist number prediction method based on big data analysis as claimed in claim 1, characterized in that: The real-time tourist number feedback data of the user is obtained, and the real-time tourist number feedback data is compared with the real-time tourist number prediction data. If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid human flow data prediction model for training, including: Acquiring scenic spot survey feedback data based on the user's mobile terminal, wherein the scenic spot survey feedback data at least includes: scenic spot real-time image data, scenic spot number real-time survey data, and scenic spot traffic flow data; Generating real-time tourist number feedback data based on the scenic spot survey feedback data, and comparing the real-time tourist number feedback data with the real-time tourist number forecast data; If the difference between the two data is less than 20%, the parameters of the target CNN-Transformer hybrid human flow data prediction model are adjusted; If the difference between the two data is greater than 20%, the real-time tourist number feedback data is input into the target CNN-Transformer hybrid passenger flow data prediction model for training.

7. A tourist number prediction system based on big data analysis, characterized in that: The tourist number prediction system includes the following modules: A data processing module is used to obtain real-time tourism-related data in a database, perform data preprocessing on the real-time tourism-related data, and obtain training tourism-related data, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot social media data; A data encryption module, used for encrypting the training tourism related data by using an RSA asymmetric encryption algorithm to obtain target training tourism related data; A model building module is used to build a CNN-Transformer hybrid human flow data prediction model, using Next-ViT as a hybrid framework of the CNN-Transformer hybrid human flow data prediction model to obtain a target CNN-Transformer hybrid human flow data prediction model; The number of people prediction module is used to input the target training tourism associated data into the target CNN-Transformer hybrid human flow data prediction model for identification and prediction, use the CNN convolutional neural network to obtain the associated features of the training tourism associated data, and Transformer obtains the global information of the training tourism associated data to obtain real-time tourist number prediction data; A travel selection module, for determining whether the flow of people at a scenic spot is higher than a preset limit flow of people at the scenic spot based on the real-time tourist number forecast data, and generating an alternative scenic spot travel plan if it is higher than the limit flow of people at the scenic spot, and displaying the alternative scenic spot travel plan to a user's mobile terminal; The prediction comparison module is used to obtain the user's real-time tourist number feedback data, compare the real-time tourist number feedback data with the real-time tourist number prediction data, and if the difference between the two data is greater than 20%, input the real-time tourist number feedback data into the target CNN-Transformer hybrid traffic data prediction model for training.

8. A tourist number prediction system based on big data analysis as claimed in claim 8, characterized in that: The data processing module includes the following submodules: An acquisition submodule is used to acquire real-time tourism-related data in a database based on the user's real-time scenic spot query requirements, wherein the real-time tourism-related data at least includes: real-time weather data, real-time traffic flow data, real-time ticket reservation data, and real-time hot spot social media data; A segmentation module, used for segmenting the real-time tourism related data to obtain multiple layers of real-time tourism related data; A binary submodule, used for performing binary conversion on the multi-layer real-time tourism related data to obtain binary real-time tourism related data; A submodule is obtained, which is used to classify the binary real-time tourism related data into data sets to obtain training tourism related data.

9. A tourist number prediction system based on big data analysis as claimed in claim 8, characterized in that: The model building module includes the following submodules: Establish a submodule for establishing a CNN-Transformer hybrid human flow data prediction model, wherein the CNN-Transformer hybrid human flow data prediction model includes at least an initial layer, an input layer, an output layer, a fully connected layer, and an average pooling layer; An embedding submodule, used for placing a linear embedding layer of a patch before each layer of the CNN-Transformer hybrid pedestrian flow data prediction model, wherein the linear embedding layer performs user channel matching and extracts patch information; A hybrid submodule, for using Next-ViT as a hybrid framework of the CNN-Transformer hybrid crowd flow data prediction model; The stacking submodule is used to stack the NCB module and the NTB module in the CNN-Transformer hybrid human flow data prediction model in a (N+1)*L hybrid manner to obtain a target CNN-Transformer hybrid human flow data prediction model.

Images