An Internet of Things-based customized bus travel method for bus digital transformation

Abstract

The present invention discloses an Internet of Things-based customized bus travel method for bus digital transformation, including: S1. Providing an input interface for bus customization services to passengers through a mobile terminal, and collecting the travel demand information sent by the passengers to the server; S2. The server obtains a corresponding travel target interest set and route creation data set for each passenger according to the travel demand information; S3. Generating multiple travel routes based on the travel target interest set and the route creation data set; S4. Judging whether a certain travel route meets the route opening conditions, generating a new customized bus route and notifying the corresponding passengers that they can make a reservation. After the corresponding passengers successfully make a reservation, generating boarding identification information and sending it to the passengers; S5. During the travel route, pushing vehicle arrival information to the passengers. The customization method of the present invention improves timeliness and shortens the route opening cycle; it can better reflect the real travel needs of passengers, facilitate passengers to plan time, and save the time cost of passengers.

Description

Technical Field

[0001] The present invention relates to the field of customized buses, and particularly to an Internet of Things bus travel customization method based on bus digital transformation. Background Art

[0002] Customized buses belong to direct buses between stations and can guide passengers to choose the mode of collective travel by customized buses, providing an effective solution to alleviate urban traffic congestion. However, in the customized bus technology, the main customization method is to designate the stations with large passenger flows found during the operation of ordinary bus lines as customized bus lines, or to customize the bus lines after on-site verification of relevant data such as passenger flow information and passenger flow patterns, and then announce the customized bus lines. Passengers select the customized bus lines they need and make reservations through the announced customized bus lines. This customization method has disadvantages such as low timeliness, long line opening cycle, and inability to well reflect the real travel needs of passengers. Summary of the Invention

[0003] The present invention provides an Internet of Things bus travel customization method based on bus digital transformation to overcome the above problems.

[0004] The present invention includes the following steps:

[0005] S1. Provide an input interface for bus customization services to passengers through a mobile terminal, and collect the travel demand information sent by the passengers to the server. The travel demand information includes the user's departure location, identity identifier, and a destination attention list. The destination attention list includes multiple tourist destinations that the user wants to pass by;

[0006] S2. The server obtains the corresponding tourist target interest set and route creation data set for each passenger according to the travel demand information; the tourist target interest set includes a set composed of the two tourist destinations that the passenger in the destination attention list most wants to pass by; the route creation data set includes the first interest tourist destination and the second interest tourist destination; the first interest tourist destination is a tourist destination whose number of times of being recognized as a tourist destination exceeds a first threshold, or a tourist destination whose departure location is the tourist destination of other passengers and the number of times exceeds a second threshold; the remaining tourist destinations other than the departure location are the second interest tourist destinations;

[0007] S3. Create a dataset based on the travel target interest set and the route to generate multiple travel routes. Each travel route takes two adjacent travel destinations as the positioning centers, selects the user departure locations that meet the selection criteria within a specified distance range to form a positioning data set, and forms a virtual route trajectory with the positioning data set. The starting point of each route trajectory is a different user departure location. The selection criteria mean that two adjacent travel destinations serving as the positioning centers belong to the travel target interest set of the passenger.

[0008] S4. Determine whether a certain travel route meets the route opening conditions. If so, generate a new customized bus route and notify the corresponding passengers that they can make a reservation. After the corresponding passengers successfully reserve the customized bus route, generate boarding identification information and send it to the passengers.

[0009] S5. Push vehicle arrival information to passengers synchronously during the travel route journey.

[0010] Further, the route opening conditions mean meeting the minimum total cost of passengers - operators, which is calculated through a route operation cost calculation model. The calculation formula for determining whether a certain travel route meets the route opening conditions is:

[0011] d i =fun(f n (c i ,h i ,o i ,p i )) (1)

[0012] Among them, d i is the travel route of the i-th route. If d i meets the minimum total cost of passengers - operators, then return d i ; p is the travel target interest set, o is the route creation dataset, c is the positioning center, h is the cost of each route, i represents the i-th route, fun represents a residual network trained using cross-entropy loss, and f m (*) represents the features extracted using a convolutional network.

[0013] Further, the virtual route trajectory is determined through a path traversal tree model.

[0014] Further, after the corresponding passengers successfully reserve the customized bus route, boarding time and driving route information are also generated synchronously based on the user departure location. The driving route information includes: customized service time period and driving stops. The calculation formula for the driving route information is:

[0015] m i =f b (u i ,a i ,yi , v i ) ∩ f c (u i , a i , y i , v i ) (2)

[0016] Among them, m is the driving route information, the user's departure location y, the vehicle spacing is a, the customized bus line u, and the distance v from the user to the nearest stop of the customized line. i represents the i-th stop, 1 <= i <= w, and w is the total number of stops. f c (*) represents a prediction network trained using perceptual loss, f b (*) represents a prediction network trained using cross-entropy loss.

[0017] Furthermore, during the process of the travel route, S5 synchronously pushes the vehicle arrival information to passengers, including: automatically calculating the spatial distance between the passenger and each bus stop according to the current GPS positioning data of the bus vehicle, so as to automatically calculate the estimated arrival time of the vehicle.

[0018] While the present invention realizes guiding passengers to choose the customized bus collective travel mode and provides an effective solution for alleviating urban traffic congestion; the timeliness of the customization method is improved; the formula is used to calculate whether to open a line, shortening the opening cycle; it can better reflect the real travel needs of passengers; provides the positioning of the bus and predicts the arrival time of the bus, facilitating passengers to plan time and saving the time cost of passengers. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. Figure 1 It is the flowchart of the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0021] Figure 1 It is the flowchart of the method of the present invention, as Figure 1As shown in the figure, the method of this embodiment may include:

[0022] S1. Provide an input interface for passengers to customize bus services through a mobile terminal, and collect the travel demand information sent by the passengers to the server. The travel demand information includes: the user's departure location, identity identifier, and a destination interest list. The destination interest list includes multiple tourist destinations that the user wants to pass by;

[0023] Specifically, on the input page provided for the user, the departure location entered by the passenger, the passenger's identity, and the tourist destinations concerned by the user will be collected, which can collect more comprehensive tourist information expected by the user.

[0024] S2. The server obtains the corresponding tourist target interest set and route creation data set for each passenger according to the travel demand information; the tourist target interest set includes a set composed of the two tourist destinations that the passenger most wants to pass by in the destination interest list; the route creation data set includes the first interest tourist destination and the second interest tourist destination; the first interest tourist destination is a tourist destination whose number of times of being recognized as a tourist destination exceeds a first threshold, or a tourist destination whose departure location is the tourist destination of other passengers and the number of times exceeds a second threshold; the remaining tourist destinations other than the departure location are the second interest tourist destinations;

[0025] Specifically, after the server obtains the user's identity, departure location, and destination, the server will obtain the set of the two tourist destinations that the passenger most wants to pass by, that is, the first tourist destination and the second tourist destination. When the number of times of being recognized as a tourist destination exceeds a certain set threshold, it becomes the first interest tourist destination; or the departure location is the destination of other passengers, and the number of times this situation occurs exceeds the second threshold. The remaining ones are the second tourist destinations; this can identify the places that the passengers most want to go to, so as to facilitate more accurate push and route planning.

[0026] S3. Generate multiple tourist routes based on the tourist target interest set and the route creation data set; the tourist routes are centered on two adjacent tourist destinations, and select the user departure locations that meet the selection conditions within a specified distance range to form a positioning data set, and form a virtual route trajectory with the positioning data set; the starting point of each route trajectory is a different user departure location; the selection condition means that two adjacent tourist destinations as the positioning center belong to the tourist target interest set of the passenger;

[0027] Specifically, plan the route according to the tourist destinations that the passengers want to visit, and plan the route centered on two adjacent destinations, which can save the passengers' time and energy while meeting the passengers' needs.

[0028] S4. Determine whether a certain tourist route meets the route opening conditions. If so, generate a new customized bus route and notify the corresponding passengers that they can make a reservation. After the corresponding passengers successfully reserve the customized bus route, generate boarding identification information and send it to the passengers.

[0029] Specifically, when the passenger confirms the reserved route, feedback the boarding information to the passenger, so that the passenger can know the bus information they are about to take and can also know that their reservation has been successful.

[0030] The route opening conditions refer to meeting the minimum total cost of passengers - operators, which is calculated through a route operation cost calculation model. The calculation formula for determining whether a certain tourist route meets the route opening conditions is:

[0031] d i =fun(f m (c i ,h i ,o i ,p i )) (1)

[0032] Among them, d i is the tourist route of the i-th route. If d i meets the minimum total cost of passengers - operators, then return d i ; p is the set of tourist target interests, o is the route creation data set, c is the positioning center, h is the cost of each route, i represents the i-th route, fun represents the residual network trained using cross-entropy loss, and f m (*) represents the features extracted using a convolutional network.

[0033] For the input p, o, c, h, use a convolutional network trained with cross-entropy to extract features from the input. The input is a set of data p, o, c, h, and then it is fed into the convolutional layer, pooling layer, and fully connected layer in sequence. The specific convolutional network is VGG-19, and finally, a 1*512-dimensional feature vector is output. fun represents the residual network trained using cross-entropy loss; during training, the input is the matching pair of the feature vector and the tourist route, and then it is fed into the convolutional layer, pooling layer, and fully connected layer in sequence. The specific convolutional network is ResNet-100, and the loss function is cross-entropy loss. Therefore, when the input is a 1*512-dimensional feature vector, the corresponding output is the tourist route d. Based on the corresponding tourist route, the cost can be calculated, and thus it can be determined whether the total cost is the minimum.

[0034] Specifically, only when the route opening conditions are met can passengers be allowed to make a reservation. The established calculation formula can minimize the cost while meeting the needs of passengers, ensure the needs of passengers for service quality, save costs, and improve operating profits.

[0035] S5. During the process of the travel itinerary, push the vehicle arrival information to passengers synchronously.

[0036] Specifically, passengers can understand in real time the bus information they are going to take, which is convenient for passengers to plan their time, arrive at the appointed place in time to wait for the bus, and save time costs.

[0037] Preferably, the virtual route trajectory is determined by a path traversal tree model.

[0038] Specifically, using the path traversal tree model, that is, taking each location between the departure place and the destination as a node of the traversal tree, calculating all the paths between the departure place and the destination and the time and price used, and finally determining the optimal route. This can ensure that each passing location and each route trajectory are confirmed, and ensure the accuracy of the calculated route planning and line opening conditions.

[0039] Preferably, after the corresponding passengers successfully reserve the customized bus line, generate the boarding time and the driving route information based on the user's departure location synchronously. The driving route information includes: the customized service period and the bus stop sites; the calculation formula of the driving route information is:

[0040] m i =f b (u i ,a i ,y i ,v i )∩f c (u i ,a i ,y i ,v i ) (2)

[0041] Wherein, m is the driving route information, the user's departure location is y, the distance between buses is a, the customized bus line is u, and the distance from the user to the nearest bus stop of the customized line is v. i represents the i-th stop, 1 <= i <= w, w is the total number of stops, f c (*) represents the prediction network trained using perceptual loss, and f b (*) represents the prediction network trained using cross-entropy loss.

[0042] Specifically, for the input u, a, y, v, the prediction network trained using perceptual loss means that during training, the input is the matching pairs of possible u, a, y, v values and the driving route, and then fed into the convolutional layer, pooling layer and fully connected layer in sequence. The specific convolutional network is IResnet-100, and the loss function is perceptual loss; when the input is a set of data u, a, y, v, the output is the predicted driving route. f b(*) indicates a prediction network trained using cross - entropy loss. For the input u, a, y, v, a prediction network trained using cross - entropy loss means that during training, the input is a pair of possible values of u, a, y, v and the matching bus routes. Then, they are fed into a convolutional layer, a pooling layer, and a fully - connected layer in sequence. Specifically, the convolutional network is VGG - 19, and the loss function is cross - entropy loss. When the input is a set of data u, a, y, v, the output is the predicted bus route. Taking the intersection of the predicted bus routes obtained by the two prediction networks gives the bus route information for the corresponding stops.

[0043] Preferably, S5 pushing vehicle arrival information to passengers synchronously during the travel route of the tour includes: automatically calculating the spatial distance between the passenger and each bus stop according to the current GPS positioning data of the bus vehicle, so as to automatically calculate the estimated arrival time of the vehicle;

[0044] Specifically, providing the estimated arrival time of the vehicle can enable passengers to better plan their time. Without wasting the passengers' time, it increases the probability of passengers arriving on time and saves time costs.

[0045] Beneficial effects:

[0046] The present invention realizes guiding passengers to choose the customized bus collective travel mode, provides an effective solution for alleviating urban traffic congestion; improves the timeliness of the customization method; uses a formula to calculate whether to open a line, shortening the opening cycle; can better reflect the real travel needs of passengers; provides the positioning of the bus and predicts the bus arrival time, facilitating passengers to plan their time and saving the time costs of passengers.

[0047] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An Internet of Things bus travel customization method based on bus digital transformation, characterized in that, it includes the following steps: S1. Provide an input interface for bus customization services to passengers through a mobile terminal, and collect the travel demand information sent by passengers to the server. The travel demand information includes the user's departure location, identity identification, and a destination attention list. The destination attention list includes multiple tourist destinations that the user wants to pass by; S2. The server obtains the corresponding tourist target interest set and route creation data set for each passenger according to the travel demand information; the tourist target interest set includes a set composed of the two tourist destinations that the passenger in the destination attention list most wants to pass by; the route creation data set includes the first interest tourist destination and the second interest tourist destination; the first interest tourist destination is a tourist destination whose number of times identified as a tourist destination exceeds a first threshold, or a tourist destination whose departure location is the tourist destination of other passengers and the number of times exceeds a second threshold; the remaining tourist destinations other than the departure location are the second interest tourist destinations; S3. Based on the tourist target interest set and the route creation data set, generate multiple tourist routes; the tourist routes use adjacent two tourist destinations as the positioning centers, select the user departure locations that meet the selection conditions within a specified distance range to form a positioning data set, and form a virtual route trajectory with the positioning data set; the starting point of each route trajectory is a different user departure location; the selection condition means that the adjacent two tourist destinations as the positioning centers belong to the tourist target interest set of the passenger; S4. Judge whether a certain tourist route meets the line opening conditions. If so, generate a new customized bus line and notify the corresponding passengers that they can book, and generate a boarding identification information and send it to the passengers after the corresponding passengers successfully book the customized bus line; S5. Push the vehicle arrival information to the passengers synchronously during the travel process of the tourist route; wherein, the line opening condition means that the total cost of passengers - operators is minimized, and it is calculated through a line operation cost calculation model; the calculation formula for judging whether a certain tourist route meets the line opening condition is: d i = fun(f m (c i , h i , o i , p i )) (1) where d i is the travel route of the i-th route. If d i satisfies the minimum total cost of passengers - operators, then return d i ; p is the set of tourist target interests, o is the route creation data set, c is the positioning center, h is the cost of each route, i represents the i-th route, fun represents the residual network trained using cross-entropy loss, and f m (*) represents the features extracted using the convolutional network.

2. The Internet of Things bus travel customization method based on bus digital transformation according to claim 1, characterized in that, the virtual route trajectory is determined through a path traversal tree model.

3. The Internet of Things bus travel customization method based on bus digital transformation according to claim 1, characterized in that, after the corresponding passengers successfully book the customized bus line, the boarding time and driving route information are also generated synchronously based on the user's departure location. The driving route information includes: the customized service period and the driving stops; the calculation formula for the driving route information is: m i = f b (u i , a i , y i , v i ) ∩ f c (u i , a i , y i , v i ) (2) Among them, m is the driving route information, including the user's departure location y, the vehicle spacing a, the customized bus line u, and the distance v from the user to the nearest stop of the customized line. i represents the i-th stop, where 1 <= i <= w and w is the total number of stops, and f c (*) represents the prediction network trained using the perceptual loss, f b (*) represents the prediction network trained using the cross-entropy loss.

4. The Internet of Things bus travel customization method based on bus digital transformation according to claim 1, characterized in that, The S5 synchronously pushes vehicle arrival information to passengers during the travel route process, including: automatically calculating the spatial distance between the passenger and each bus stop based on the current GPS positioning data of the bus vehicle, so as to automatically calculate the estimated arrival time of the vehicle.

Images