Training method of passenger flow inference model and method and device for inferring passenger flow

By using deep learning technology in the passenger flow inference model and training the model based on the thermal value and real passenger flow data, the problem of insufficient passenger flow prediction in the existing technology is solved, and accurate prediction of tourist flow at scenic spots is achieved.

CN113643066BActive Publication Date: 2024-11-19JINGDONG CITY BEIJING DIGITS TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110938943.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-08-16
Publication Date
2024-11-19
Estimated Expiration
2041-08-16

AI Technical Summary

Technical Problem

The existing passenger flow inference model is difficult to accurately predict the passenger flow of scenic spots during training, especially during peak hours such as holidays, which leads to inaccurate prediction results.

Method used

By determining the location boundaries and the minimum external rectangle in the area, performing grid processing, determining the location characteristics based on the thermal value, and using these features and corresponding real passenger flow data to train a deep learning model to obtain a more accurate passenger flow inference model.

Benefits of technology

More accurate predictions of tourist traffic at scenic spots are achieved, especially during peak hours, improving the accuracy and reliability of the inference model.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113643066B_ABST
    Figure CN113643066B_ABST
Patent Text Reader

Abstract

The present disclosure provides a training method for a passenger flow inference model and a method and device for inferring passenger flow, and relates to the field of artificial intelligence technology including deep learning. Specific implementation methods include: determining the boundaries of places in a region, and determining the minimum circumscribed rectangle of the boundary; gridding the circumscribed rectangle, and determining the grid within the boundary in each obtained grid as the location grid; determining the characteristics of the location based on the thermal value of each location grid; obtaining the real passenger flow data corresponding to the characteristics of the place in the region; using the characteristics of the place in the region as input, and using the real passenger flow data corresponding to the characteristics as the target output, training the passenger flow prediction model to be trained, and obtaining the trained passenger flow prediction model. The present disclosure can better train the passenger flow inference model of the scenic spot through thermal values ​​to obtain a more accurate passenger flow inference model, so that accurate passenger flow inference can be achieved through the trained passenger flow inference model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, specifically to the field of artificial intelligence technology including deep learning, and more particularly to a training method for a passenger flow inference model and a method and device for inferring passenger flow. Background Art

[0002] As transportation becomes more and more convenient, people's travel frequency has increased significantly, such as traveling in scenic spots, shopping in commercial areas, etc.

[0003] In a region, there may be various locations, such as scenic spots in a scenic area or shops in a commercial area. The passenger flow in these areas often has regularities. For example, the passenger flow during holidays is significantly higher than that during normal times. In related technologies, passenger flow can be inferred based on these regularities. Summary of the invention

[0004] Provided are a training method and device for a passenger flow inference model, an electronic device, and a storage medium, as well as a processing method and device for inferring passenger flow, an electronic device, and a storage medium.

[0005] According to a first aspect, a method for training a passenger flow inference model is provided, comprising: determining the boundaries of places in a region, and determining the minimum circumscribed rectangle of the boundaries; gridding the circumscribed rectangles, and determining the grids within the boundaries among the obtained grids as the place grids; determining the characteristics of the places based on the thermal values ​​of the grids of each place; obtaining real passenger flow data corresponding to the characteristics of the places in the region; using the characteristics of the places in the region as input, and using the real passenger flow data corresponding to the characteristics as target output, to train the passenger flow inference model to be trained, and obtain the trained passenger flow inference model.

[0006] According to a second aspect, a method for inferring passenger flow is provided, including: the method adopts the passenger flow inference model obtained in the first aspect.

[0007] According to a third aspect, a training device for a passenger flow inference model is provided, comprising: a rectangle determination unit, configured to determine the boundary of a place in a region, and determine the minimum circumscribed rectangle of the boundary; a grid unit, configured to grid the circumscribed rectangle, and determine the grid within the boundary as the place grid among the obtained grids; a feature determination unit, configured to determine the feature of the place based on the thermal value of each place grid; an acquisition unit, configured to acquire real passenger flow data corresponding to the feature of the place in the region; a training unit, configured to use the feature of the place in the region as input, and use the real passenger flow data corresponding to the feature as target output, to train the passenger flow inference model to be trained, and obtain the trained passenger flow inference model.

[0008] According to a fourth aspect, a device for inferring passenger flow is provided, comprising: the device adopts a passenger flow inference model obtained in any one of the third aspects.

[0009] According to the fifth aspect, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method of any embodiment of any of the above methods.

[0010] According to a sixth aspect, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute a method of any embodiment of any of the above methods.

[0011] According to a seventh aspect, a computer program product is provided, comprising a computer program, which implements the method of any embodiment of any of the above methods when executed by a processor.

[0012] According to the scheme disclosed in the present invention, the passenger flow inference model of the scenic spot can be better trained through thermal values ​​to obtain a more accurate passenger flow inference model, so that accurate passenger flow inference can be achieved through the trained passenger flow inference model. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Other features, objects and advantages of the present disclosure will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0014] Figure 1 is an exemplary system architecture diagram in which some embodiments of the present disclosure may be applied;

[0015] Figure 2 is a flow chart of an embodiment of a method for training a passenger flow inference model according to the present disclosure;

[0016] Figure 3 is a schematic diagram of an application scenario of the training method of the passenger flow inference model disclosed in the present invention;

[0017] Figure 4A is a flow chart of an embodiment of a method for inferring passenger flow according to the present disclosure;

[0018] Figure 4B is a schematic diagram of a regional grid in the method for inferring passenger flow according to the present disclosure;

[0019] Figure 4C is a schematic diagram of the training process of the passenger flow inference model according to the present disclosure;

[0020] Figure 5is a structural schematic diagram of an embodiment of a training device for a passenger flow inference model according to the present disclosure;

[0021] Figure 6 It is a block diagram of an electronic device used to implement the training method of the passenger flow inference model of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0022] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0023] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved all comply with the provisions of relevant laws and regulations, take necessary confidentiality measures, and do not violate public order and good morals.

[0024] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0025] Figure 1 An exemplary system architecture 100 is shown to which an embodiment of a method for training a passenger flow inference model or a device for training a passenger flow inference model of the present disclosure can be applied.

[0026] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0027] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as video applications, live broadcast applications, instant messaging tools, email clients, social platform software, etc.

[0028] The terminal devices 101, 102, 103 here can be hardware or software. When the terminal devices 101, 102, 103 are hardware, they can be various electronic devices with display screens, including but not limited to smart phones, tablet computers, e-book readers, laptop portable computers and desktop computers, etc. When the terminal devices 101, 102, 103 are software, they can be installed in the electronic devices listed above. It can be implemented as multiple software or software modules (for example, multiple software or software modules for providing distributed services), or it can be implemented as a single software or software module. No specific limitation is made here.

[0029] The server 105 may be a server that provides various services, such as a background server that provides support for the terminal devices 101, 102, and 103. The background server may analyze and process the received data such as the boundary of the location, and feed back the processing results (such as the trained passenger flow inference model) to the terminal device.

[0030] It should be noted that the method for training the passenger flow inference model and inferring the passenger flow provided in the embodiments of the present disclosure can be executed by the server 105 or the terminal devices 101, 102, 103. Accordingly, the device for training the passenger flow inference model and inferring the passenger flow can be set in the server 105 or the terminal devices 101, 102, 103.

[0031] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0032] Continue to refer Figure 2 , shows a process 200 of an embodiment of a method for training a passenger flow inference model according to the present disclosure. The method for training a passenger flow inference model comprises the following steps:

[0033] Step 201, determine the boundary of a location in a region, and determine the circumscribed rectangle of the boundary.

[0034] In this embodiment, the training method of the passenger flow inference model runs on the execution subject (e.g. Figure 1 The server or terminal device shown in the figure can determine the boundary, i.e., edge, of the place in the region, and determine the bounding rectangle of the boundary. The boundary can be a polygon. The bounding rectangle can be a minimum bounding rectangle. In some implementations, the region can include only the above-mentioned place, or can include other places.

[0035] Specifically, the boundary can be represented by a series of longitude and latitude points. For example, when the longitude and latitude points include the i-th position on the boundary, that is, the 1st, 2nd, ... nth positions, the longitude and latitude points in the boundary can be represented as [(lng1, lat1), (lng2, lat2), ..., (lng n ,lat n )].

[0036] Step 202 , gridding the circumscribed rectangle, and determining the grid within the boundary among the obtained grids as the location grid.

[0037] In this embodiment, the execution entity may grid the circumscribed rectangle and determine the grid within the boundary among the grids obtained by gridding. The execution entity may use the determined grid as the location grid. In practice, the grid may be represented by longitude and latitude points, and specifically, the longitude and latitude points may be the intersection of the diagonals of the grid.

[0038] Specifically, the execution subject may enumerate the grids within the circumscribed rectangle, and determine the positional relationship between each enumerated grid and the boundary to determine whether the enumerated grid is within the boundary, thereby finding the grid within the boundary as the location grid.

[0039] Step 203, determining the characteristics of the location based on the thermal values ​​of each location grid.

[0040] In this embodiment, the execution entity may determine the characteristics of the location based on the thermal values ​​of the location grids. In practice, the execution entity may determine the characteristics of the location based on the thermal values ​​of each location grid in various ways. For example, the execution entity may directly use the sum of the thermal values ​​of each location grid as the characteristic of the location. Alternatively, the execution entity may input the thermal values ​​of each location grid into a preset model and obtain the characteristics of the location output from the preset model. The preset model may calculate the characteristics of the location through the thermal values ​​of each location grid in the location. The characteristics are model inputs and may reflect the characteristics of the location.

[0041] The thermal value of a location grid may have a time attribute, indicating the thermal value within the location grid at a certain time.

[0042] Step 204, obtaining real passenger flow data corresponding to the characteristics of the locations in the region.

[0043] In this embodiment, the execution subject may obtain the real passenger flow data corresponding to the characteristics of the location in the region. The real passenger flow data corresponding to the characteristics may refer to the real passenger flow data at the location and time of the characteristics.

[0044] The real data of passenger flow can be obtained in many ways, such as ticket checking data, which is the secondary ticket checking data required to enter the scenic spot.

[0045] Step 205 , using the features of the locations in the region as input, and using the real passenger flow data corresponding to the features as target output, to train the passenger flow inference model to be trained, and obtain a trained passenger flow inference model.

[0046] In this embodiment, the above-mentioned execution entity can use the characteristics of the locations in the area as input, and use the real passenger flow data corresponding to the input characteristics as the target output to train the passenger flow inference model to be trained, thereby obtaining the trained passenger flow inference model.

[0047] The passenger flow inference model is a deep neural network, such as a feedforward neural network, a convolutional neural network, a residual neural network, etc.

[0048] The passenger flow inference model can be used to infer passenger flow through the characteristics of a location, where the location can be a scenic spot, a commercial center, etc. During training, the characteristics of a single location in the above-mentioned area and the corresponding real passenger flow data are generally used for training. In some cases, the characteristics of at least two scenic spots and the corresponding real passenger flow data can also be used for training. Specifically, the obtained passenger flow inference model can be used to infer passenger flow for a location, a region, a local area in a location, and / or at least two locations in a region.

[0049] The method provided by the above-mentioned embodiment of the present disclosure can better train the passenger flow inference model of the scenic spot through thermal values ​​to obtain a more accurate passenger flow inference model, so that accurate passenger flow inference can be achieved through the trained passenger flow inference model.

[0050] In some optional implementations of this embodiment, the above-mentioned step 203 may include: determining the characteristics of the location based on the thermal value of each location grid, including: in response to the relationship between the thermal value of the location and the passenger flow in the area being a linear relationship, determining the sum of the thermal values ​​of each location grid of the location; determining the characteristics of the location based on the sum of the thermal values.

[0051] In these optional implementations, the execution entity may assume that the relationship between the thermal value of the location and the passenger flow is a linear relationship. In this case, the execution entity may determine the sum of the thermal values ​​of each location grid of the location. Furthermore, the execution entity may determine the characteristics of the location based on the sum of the thermal values.

[0052] In practice, the execution entity may determine the characteristics of the place according to the sum of thermal values ​​in various ways. For example, the execution entity may directly determine the sum of thermal values ​​as the characteristics of the place. Alternatively, the execution entity may input the sum of thermal values ​​into a specified model and obtain the characteristics of the place output from the specified model. The specified model may be used to infer the characteristics of the place through the sum of thermal values.

[0053] Specifically, the linear mapping relationship (linear function relationship) between the thermal value heat and the passenger flow cnt can be expressed as f: f(heat k )→cnt k , where k represents the kth location in the region, i.e. location k, heat k Indicates the thermal value of location k, cnt k represents the passenger flow of location k. For location grid i and location grid j in the location, there exists f(heat i +heat j )=f(heat i )+f(heat j ).

[0054] The boundary of location k can be expressed as boundary k , the sum of the thermal values ​​of the location can be expressed as The sum of the thermal values ​​can be used as the feature feat of location k k .

[0055] These optional implementations can more accurately determine the characteristics of the location based on the sum of the thermal values ​​when there is a linear relationship between the thermal value of the location and the passenger flow.

[0056] Continue to see Figure 3 , Figure 3 FIG. 1 is a schematic diagram of an application scenario of the method for training a passenger flow inference model according to this embodiment. Figure 3 In the application scenario, the execution subject 301 determines the boundary 302 of the place in the region, and determines the circumscribed rectangle 303 of the boundary. The execution subject 301 grids the circumscribed rectangle, and determines the grid within the boundary as the place grid 304 among the obtained grids. The execution subject 301 determines the feature 306 of the place based on the thermal value 305 of each place grid 304. The execution subject 301 obtains the real passenger flow data 307 corresponding to the feature 306 of the place in the region. The execution subject 301 uses the feature 306 of the place in the region as input, and uses the real passenger flow data 307 corresponding to the feature 306 as the target output, trains the passenger flow inference model to be trained, and obtains the trained passenger flow inference model 308.

[0057] Further references Figure 4A , which shows a process 400 of an embodiment of a method for training a passenger flow inference model. The process 400 includes the following steps:

[0058] Step 401, determine the boundary of a location in a region, and determine the minimum bounding rectangle of the boundary.

[0059] Step 402, determine the grids in the circumscribed rectangle, and determine the grids within the boundaries among the grids as the location grids.

[0060] Step 403: In response to the relationship between the thermal value of a location in a region and the passenger flow being a nonlinear relationship, the passenger flow of the location is expressed by using the sum of the nonlinear functions of the thermal values ​​of each location grid of the location.

[0061] In this embodiment, the method for training the passenger flow inference model runs on an execution subject (e.g. Figure 1 The server or terminal device shown in the figure may assume that the relationship between the thermal value of the location and the passenger flow is a nonlinear relationship. In this case, the above execution subject may determine the nonlinear functional relationship between the thermal value of the location and the passenger flow of the location. For location grid i and location grid j in the location, there exists f(heat i +heat j )≠f(heat i )+f(heat j ).

[0062] In practice, the passenger flow can be represented by the sum of the nonlinear functions of the thermal values ​​of each location grid of the location.

[0063] The nonlinear function of the location grid i can be g(heat i ), the sum of the nonlinear functions of each location grid in the location can be Therefore, the passenger flow cnt at location k k It can be expressed as the sum of the nonlinear functions.

[0064] Step 404: Perform Taylor expansion processing on the sum of the nonlinear functions to obtain a polynomial.

[0065] In this embodiment, the execution entity may perform Taylor expansion processing on the sum of the nonlinear functions to obtain a polynomial.

[0066] In practice, Performing k-order Taylor expansion, the obtained polynomial is:

[0067] In practice, k in the above k-order may be 2, that is, greater than or equal to 2. In order to improve the accuracy of the polynomial and thus more accurately determine the characteristics of the location, the above k may be 3, that is, greater than or equal to 3.

[0068] Step 405, determining the characteristics of the location based on the non-coefficient content of each term in the polynomial, wherein the non-coefficient content is expressed using the thermal value of the location grid.

[0069] In this embodiment, the execution subject may determine the characteristics of the place according to the non-coefficient content of each term in the polynomial in various ways. For example, the execution subject may directly determine the non-coefficient content as the characteristics of the place. Alternatively, the execution subject may perform preset processing on the non-coefficient content, such as inputting a pre-trained model to obtain the characteristics of the place output from the model. The model may determine the characteristics of the place through the non-coefficient content. The non-coefficient content in the polynomial is expressed in thermal values.

[0070] In practice, the polynomial The coefficients in are: g(0),g′(0),…,g k (0). The above execution entity can perform o(x k ) is ignored, so the non-coefficient content of each term in the above polynomial is:

[0071] This non-coefficient content can be used as a feature feat k .

[0072] Step 406, obtaining real passenger flow data corresponding to the characteristics of the locations in the region.

[0073] Step 407 , using the features of the locations in the region as input, and using the real passenger flow data corresponding to the features as target output, to train the passenger flow inference model to be trained, and obtain a trained passenger flow inference model.

[0074] The above steps 401, 402, 406 and 407 are respectively implemented in the same or similar manner as steps 201, 202, 204 and 205, and are not described in detail here.

[0075] In this embodiment, when there is a nonlinear relationship between the thermal value of a place and the passenger flow, the nonlinear function can be processed to more accurately determine the characteristics of the place based on the thermal value.

[0076] In some optional implementations of any embodiment of the present disclosure, the method may further include: obtaining a grid size of the region, wherein the grid size is determined based on a latitude and longitude step set for the grid; gridding the region according to the grid size to obtain a region grid; and determining the grids in the circumscribed rectangle and the grids within the boundaries of each grid as location grids may include: determining the region grid within the circumscribed rectangle as a candidate region grid, and determining the candidate region grid within the boundary as a location grid.

[0077] In these optional implementations, the above-mentioned execution entity can obtain the grid size of the above-mentioned area. The above-mentioned execution entity or other electronic devices can first obtain the longitude and latitude step length, which refers to the longitude difference and latitude difference, and can correspond to a geographical length. For example, the longitude and latitude step length refers to the longitude difference and latitude difference are both 0.0001 degrees, and the corresponding geographical length is about 10 meters, which can be used as the length and width of the grid size. That is, the above-mentioned execution entity or other electronic devices can directly convert the longitude and latitude step length into the grid size.

[0078] The execution subject may mesh the region in various ways, such as directly obtaining the bounding rectangle of the region boundary from the device or other electronic devices, or determining the bounding rectangle of the region boundary on the device. Afterwards, the execution subject may mesh the bounding rectangle of the region boundary according to the mesh size to mesh the region and obtain the region mesh. Alternatively, the execution subject may input the polygon and mesh size of the region into a preset meshing model to obtain the region mesh output from the model.

[0079] The execution subject may first determine the area grids within the bounding rectangle of the location as candidate area grids, and then determine the candidate area grids within the boundary of the location and use the candidate area grids as location grids.

[0080] like Figure 4B As shown, the figure shows the regional grid obtained by gridding the region.

[0081] like Figure 4C As shown in the figure, a schematic diagram of the training process of the passenger flow inference model is shown. Specifically, the above-mentioned execution subject can first determine the boundary of the region. Then, the region is gridded through the boundary to obtain the regional grid, that is, the regional longitude and latitude data. The gridding here can be to first determine the outer rectangle of the regional boundary, and then grid the outer rectangle.

[0082] Afterwards, the above execution entity can obtain the positioning data within the regional scope. The format of the positioning data is a four-tuple (id, time, lng, lat), where id represents the device that generates the current positioning data and time represents the timestamp of generating the current positioning data. Through the positioning data and the regional grid, grid positioning thermal calculation is performed. The grid positioning thermal calculation can locate each grid and perform thermal calculation on the grid. The calculated regional grid positioning thermal has a time attribute.

[0083] The above-mentioned execution subject can aggregate the heat of each grid in the location through the regional grid positioning heat and the location boundary to obtain the location positioning heat. The location positioning heat is a feature with a time attribute. The location ticket inspection data corresponding to the feature, that is, the real passenger flow data, is obtained, and the deep learning model, that is, the passenger flow inference model to be trained, is trained to obtain the passenger flow inference model of the trained location.

[0084] These implementations can determine the grid of a location through the grid of a region, thereby improving the accuracy of determining the grid of the location and enhancing the consistency of the grids of various locations in the region.

[0085] In some optional implementations of any one of the embodiments of the present disclosure, the passenger flow inference model is a multi-layer feedforward neural network.

[0086] In these optional implementations, the passenger flow inference model may be a multi-layer feedforward neural network, namely a multi-layer perceptron (MLP).

[0087] In practice, the above execution entity or other electronic devices can obtain a training data set. The data set D includes multiple pairs of x k and k . Where x k =feat k ,y k =cnt k Among them, the processing of each processing layer can be expressed as σ(Wfeat k +b). Where W and b are the parameters that need to be modified during the training process, and σ is the activation function. For example, if two of the three layers in the passenger flow inference model are processing layers, that is, each of the two processing layers processes the output of the previous layer, thereby changing the output, then the final output of the model is y k =σ(W2σ(W1feat k +b1)+b2), where W1 and b1 are the parameters of the first processing layer, and W2 and b2 are the parameters of the second processing layer.

[0088] It should be noted that the number of processing layers of the above multi-layer perceptron is greater than or equal to 3, and the specific number of processing layers can be determined by cross-validation.

[0089] These implementations can utilize feedforward neural network training to obtain accurate passenger flow inference models.

[0090] In some optional implementations of any one of the embodiments of the present disclosure, the region is a scenic area, and the places in the region are scenic spots.

[0091] In these optional implementations, the above-mentioned execution entity can infer the passenger flow of scenic spots in the scenic area, thereby increasing the universality of the solution in the present disclosure. In addition, the present disclosure can also limit the area of ​​the region, and the area of ​​the region does not exceed the target geographical area, that is, the area of ​​the region is less than or equal to the target geographical area.

[0092] Further references Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a training device for a passenger flow inference model. Figure 2 Corresponding to the method embodiment shown in FIG. 1 , in addition to the features described below, the device embodiment may also include Figure 2 The device can be applied to various electronic devices.

[0093] like Figure 5 As shown, the training device 500 of the passenger flow inference model of this embodiment includes: a rectangle determination unit 501, a grid unit 502, a feature determination unit 503, an acquisition unit 504 and a training unit 505. Among them, the rectangle determination unit 501 is configured to determine the boundary of the place in the area and determine the circumscribed rectangle of the boundary; the grid unit 502 is configured to grid the circumscribed rectangle and determine the grid within the boundary as the place grid among the obtained grids; the feature determination unit 503 is configured to determine the characteristics of the place based on the thermal value of each place grid; the acquisition unit 504 is configured to obtain the real passenger flow data corresponding to the characteristics of the place in the area; the training unit 505 is configured to use the characteristics of the place in the area as input, and use the real passenger flow data corresponding to the characteristics as target output, to train the passenger flow inference model to be trained, and obtain the trained passenger flow inference model.

[0094] In this embodiment, the specific processing of the rectangle determination unit 501, the grid unit 502, the feature determination unit 503, the acquisition unit 504 and the training unit 505 of the training device 500 for inferring the passenger flow model and the technical effects thereof can be referred to respectively. Figure 2The relevant descriptions of step 201, step 202, step 203, step 204 and step 205 in the corresponding embodiment are not repeated here.

[0095] In some optional implementations of this embodiment, the feature determination unit is further configured to perform determination of the features of the location based on the thermal values ​​of each location grid in the following manner: in response to the relationship between the thermal value of the location and the passenger flow in the area being a linear relationship, determine the sum of the thermal values ​​of each location grid of the location; and determine the features of the location based on the sum of the thermal values.

[0096] In some optional implementations of the present embodiment, the feature determination unit is further configured to perform determination of the features of the location based on the thermal values ​​of each location grid in the following manner: in response to the fact that the relationship between the thermal value of a location in a region and the passenger flow is a nonlinear relationship, the passenger flow of the location is expressed by the sum of the nonlinear functions of the thermal values ​​of each location grid of the location; Taylor expansion is performed on the sum of the nonlinear functions to obtain a polynomial; the features of the location are determined based on the non-coefficient content of each term in the polynomial, wherein the non-coefficient content is expressed by the thermal values ​​of the location grid.

[0097] In some optional implementations of this embodiment, the device also includes: an acquisition unit, configured to acquire a grid size of the area, wherein the grid size is determined based on a latitude and longitude step set for the grid; a division unit, configured to divide the area into grids according to the grid size to obtain area grids; and a grid unit, further configured to perform gridding on the circumscribed rectangle in the following manner, and determine the grid within the boundary among the obtained grids as the location grid: determine the area grid within the circumscribed rectangle as a candidate area grid, and determine the candidate area grid within the boundary as the location grid.

[0098] In some optional implementations of this embodiment, the passenger flow inference model is a multi-layer feedforward neural network.

[0099] The present disclosure also shows a process of an embodiment of a method for inferring passenger flow. The process includes the following steps: the method adopts a passenger flow inference model obtained in any of the above embodiments.

[0100] The method provided in this embodiment can accurately infer the visitor flow of a scenic spot through thermal value.

[0101] In some optional implementations of this embodiment, the above method may also include: inputting the characteristics of the place to be inferred into the passenger flow inference model to obtain the passenger flow output from the passenger flow inference model, wherein the place to be inferred is the same as or different from the target place corresponding to the characteristics used in training, and the distance between the target place and the place to be inferred does not exceed a preset distance threshold.

[0102] In these optional implementations, the location corresponding to the features used in the training can be used as the target location. The location to be inferred using the passenger flow inference model can be the same location as the target location or a different location. The above-mentioned execution subject can achieve more accurate passenger flow inference when the distance between the target location and the location to be predicted using the passenger flow inference model is short. For example, the target location and the location to be inferred are scenic spots in the same scenic area.

[0103] Further, as an implementation of the method shown in the above embodiment, the present disclosure provides an embodiment of a device for inferring passenger flow, which corresponds to the above method embodiment for inferring passenger flow. In addition to the features described below, the device embodiment may also include features or effects that are the same or corresponding to the method embodiment. The device may be specifically applied to various electronic devices.

[0104] The device for inferring a passenger flow model in this embodiment includes: a passenger flow inference model obtained by using any one of the training methods for the passenger flow inference model.

[0105] In some optional implementations of this embodiment, the method may also include: inputting the characteristics of the place to be inferred into the passenger flow inference model to obtain the passenger flow output from the passenger flow inference model, wherein the place to be inferred is the same as or different from the target place corresponding to the characteristics used in training, and the distance between the target place and the place to be inferred does not exceed a preset distance threshold.

[0106] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0107] like Figure 6 , is a block diagram of an electronic device for a method of training a passenger flow inference model and an electronic device for inferring passenger flow according to an embodiment of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0108] like Figure 6As shown, the electronic device includes: one or more processors 601, a memory 602, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 6 A processor 601 is taken as an example.

[0109] The memory 602 is a non-transitory computer-readable storage medium provided by the present disclosure. The memory stores instructions executable by at least one processor, so that at least one processor executes the training method of the passenger flow inference model and the method of inferring passenger flow provided by the present disclosure. The non-transitory computer-readable storage medium of the present disclosure stores computer instructions, which are used to cause a computer to execute the training method of the passenger flow inference model and the method of inferring passenger flow provided by the present disclosure.

[0110] The memory 602 is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, such as the program instructions / modules corresponding to the training passenger flow inference model and the method for inferring passenger flow in the embodiments of the present disclosure (for example, the attached Figure 5 The processor 601 executes various functional applications and data processing of the server by running the non-transient software programs, instructions and modules stored in the memory 602, that is, the training method of the passenger flow inference model and the method of inferring passenger flow in the above method embodiment are implemented.

[0111] The memory 602 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of electronic devices for training passenger flow inference models and inferring passenger flow, etc. In addition, the memory 602 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage devices. In some embodiments, the memory 602 may optionally include a memory remotely arranged relative to the processor 601, and these remote memories may be connected to the electronic devices for training passenger flow inference models and inferring passenger flow through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0112] The electronic device of the method for training a passenger flow inference model and the electronic device of the method for inferring passenger flow may also include: an input device 603 and an output device 604. The processor 601, the memory 602, the input device 603 and the output device 604 may be connected via a bus or other means. Figure 6 The example of connecting through bus is taken in the following.

[0113] The input device 603 can receive input digital or character information, and generate key signal input related to user settings and function control of electronic devices for training passenger flow inference models and inferring passenger flow, such as input devices such as touch screen, keypad, mouse, track pad, touch pad, indicator rod, one or more mouse buttons, trackball, joystick, etc. The output device 604 may include a display device, an auxiliary lighting device (e.g., LED) and a tactile feedback device (e.g., a vibration motor), etc. The display device may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0114] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0115] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for programmable processors and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or means (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0116] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0117] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0118] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0119] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0120] The units involved in the embodiments described in the present disclosure may be implemented by software or by hardware. The units described may also be set in a processor, for example, may be described as: a processor includes a rectangle determination unit, a grid unit, a feature determination unit, an acquisition unit, and a training unit. The names of these units do not constitute a limitation on the unit itself in some cases, for example, the rectangle determination unit may also be described as "a unit that determines the boundaries of places in an area and determines the circumscribed rectangle of the boundaries".

[0121] As another aspect, the present disclosure also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by the device, the device: determines the boundary of the place in the area, and determines the circumscribed rectangle of the boundary; grids the circumscribed rectangle, and determines the grid within the boundary in each obtained grid as the place grid; determines the characteristics of the place based on the thermal value of each place grid; obtains the real passenger flow data corresponding to the characteristics of the place in the area; uses the characteristics of the place in the area as input, and uses the real passenger flow data corresponding to the characteristics as the target output, trains the passenger flow inference model to be trained, and obtains the trained passenger flow inference model.

[0122] As another aspect, the present disclosure further provides a computer-readable medium, which may be included in the device described in the above embodiment; or may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by the device, the device: adopts any one of the training methods for the passenger flow inference model to obtain a passenger flow inference model.

[0123] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present disclosure is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above invention concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other.

Claims

1. A method for training a passenger flow inference model, the method comprising: Determining the boundaries of the locations in the region and determining a bounding rectangle of the boundaries; Gridding the circumscribed rectangle, and determining, among the obtained grids, the grids within the boundary as location grids; Determining the characteristics of the location based on the thermal values ​​of each of the location grids according to a linear or nonlinear relationship between the thermal values ​​of the location and the passenger flow in the region; In response to the relationship between the thermal value of the location in the region and the passenger flow being a nonlinear relationship, performing Taylor expansion processing on the sum of the nonlinear functions of the thermal values ​​of each location grid to determine the characteristics of the location; Acquiring real data of passenger flow corresponding to the characteristics of places in the region; The characteristics of the places in the area are used as input, and the real passenger flow data corresponding to the characteristics are used as target output to train the passenger flow inference model to be trained, so as to obtain the trained passenger flow inference model.

2. The method according to claim 1, wherein: The determining of the characteristics of the location based on the thermal values ​​of each of the location grids according to the linear or nonlinear relationship between the thermal values ​​of the location in the region and the passenger flow includes: In response to the relationship between the thermal value of the location in the region and the passenger flow being a linear relationship, determining the sum of the thermal values ​​of each location grid of the location; Based on the sum of the thermal values, the characteristics of the location are determined.

3. The method according to claim 1, wherein: The determining of the characteristics of the location based on the thermal values ​​of each of the location grids according to the linear or nonlinear relationship between the thermal values ​​of the location in the region and the passenger flow includes: In response to the relationship between the thermal value and the passenger flow of the location in the region being a nonlinear relationship, the passenger flow of the location is expressed by using the sum of the nonlinear functions of the thermal values ​​of each location grid of the location; Performing Taylor expansion processing on the sum of the nonlinear functions to obtain a polynomial; The characteristics of the location are determined based on the non-coefficient content of each term in the polynomial, wherein the non-coefficient content is expressed using thermal values ​​of the location grid.

4. The method according to claim 1, wherein: The method further comprises: Acquire a grid size of the region, wherein the grid size is determined based on a latitude and longitude step set for the grid; Dividing the region into grids according to the grid size to obtain a regional grid; and The step of gridding the circumscribed rectangle and determining a grid within the boundary as a location grid among the obtained grids includes: The region grids within the circumscribed rectangle are determined as candidate region grids, and the candidate region grids within the boundary are determined as location grids.

5. The method according to any one of claims 1 to 4, wherein: The passenger flow inference model is a multi-layer feedforward neural network.

6. A method for inferring passenger flow, wherein: The method adopts the passenger flow inference model obtained in any one of claims 1-5.

7. The method according to claim 6, wherein: The method comprises: The features of the location to be inferred are input into the passenger flow inference model to obtain the passenger flow output from the passenger flow inference model, wherein the location to be inferred is the same as or different from the target location corresponding to the features used in training, and the distance between the target location and the location to be inferred does not exceed a preset distance threshold.

8. A training device for a passenger flow inference model, the device comprising: a rectangle determination unit configured to determine a boundary of a location in a region and determine a circumscribed rectangle of the boundary; A grid unit configured to grid the circumscribed rectangle and determine, among the obtained grids, a grid within the boundary as a location grid; a feature determination unit configured to determine the feature of the location based on the thermal value of each of the location grids according to a linear or nonlinear relationship between the thermal value of the location in the region and the passenger flow; The feature determination unit is further configured to, in response to the relationship between the thermal value of the location in the region and the passenger flow being a nonlinear relationship, perform Taylor expansion processing on the sum of the nonlinear functions of the thermal values ​​of each location grid to determine the feature of the location; an acquisition unit configured to acquire real passenger flow data corresponding to the characteristics of the places in the area; The training unit is configured to use the characteristics of the locations in the area as input and the real passenger flow data corresponding to the characteristics as target output to train the passenger flow inference model to be trained to obtain the trained passenger flow inference model.

9. The device according to claim 8, wherein: The feature determination unit is further configured to perform the step of determining the feature of the location based on the thermal value of each of the location grids according to the linear relationship or nonlinear relationship between the thermal value of the location in the region and the passenger flow in the region in the following manner: In response to the relationship between the thermal value of the location in the region and the passenger flow being a linear relationship, determining the sum of the thermal values ​​of each location grid of the location; Based on the sum of the thermal values, the characteristics of the location are determined.

10. The device according to claim 8, wherein: The feature determination unit is further configured to perform the step of determining the feature of the location based on the thermal value of each of the location grids according to the linear relationship or nonlinear relationship between the thermal value of the location in the region and the passenger flow in the region in the following manner: In response to the relationship between the thermal value and the passenger flow of the location in the region being a nonlinear relationship, the passenger flow of the location is expressed by using the sum of the nonlinear functions of the thermal values ​​of each location grid of the location; Performing Taylor expansion processing on the sum of the nonlinear functions to obtain a polynomial; The characteristics of the location are determined based on the non-coefficient content of each term in the polynomial, wherein the non-coefficient content is expressed using thermal values ​​of the location grid.

11. The device according to claim 8, wherein: The device also includes: an acquisition unit configured to acquire a grid size of the region, wherein the grid size is determined based on a latitude and longitude step set for the grid; a division unit configured to divide the region into grids according to the grid size to obtain a regional grid; and The grid unit is further configured to perform the gridding of the circumscribed rectangle in the following manner, and determine the grid within the boundary as the location grid among the obtained grids: The region grids within the circumscribed rectangle are determined as candidate region grids, and the candidate region grids within the boundary are determined as location grids.

12. A device for inferring passenger flow, wherein: The device adopts the passenger flow inference model obtained by any one of claims 8-11.

13. The device according to claim 12, wherein: The device comprises: The features of the location to be inferred are input into the passenger flow inference model to obtain the passenger flow output from the passenger flow inference model, wherein the location to be inferred is the same as or different from the target location corresponding to the features used in training, and the distance between the target location and the location to be inferred does not exceed a preset distance threshold.

14. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

15. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.