A bus arrival information and advertisement publishing management system based on big data and its method
The bus arrival time is predicted through the big data management system and the LSTM network, and the dynamic loading algorithm is combined to match and publish advertisements, which solves the problems of bus arrival time and advertising quality, and improves passenger experience and advertising effectiveness.
Patent Information
- Application Number
- CN202311241171.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-09-25
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2043-09-25
AI Technical Summary
Accurate prediction of bus arrival time and reasonable advertising placement have become important issues in the utilization of waiting gaps on bus stops. The existing technology is difficult to meet passengers' accurate demand for arrival time and the improvement of advertising quality.
The bus arrival information and advertising release management system based on big data is adopted to predict the bus arrival time through the LSTM network, and the advertising information is matched and published in combination with the dynamic loading algorithm to ensure that the bus arrival time and advertising time match and release.
It improves the accuracy and advertising quality of bus arrival prediction, improves passenger experience and advertising release effect, and meets passengers' precise needs for arrival time and effective advertising.
Smart Images

Figure CN117314526B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of public transportation technology, and in particular to a bus arrival information and advertisement publishing management system based on big data and a method thereof. Background Art
[0002] The pace of electronic informationization and intelligentization of the public transportation industry is accelerating. Unmanned ticket sales have been popularized in many cities in China, and IC card charging methods have become popular in large and medium-sized cities in China. With the development of the public transportation industry, bus stops have higher and higher requirements for the accuracy of predicted arrival times of buses that are about to arrive. Passengers do not care how many vehicles are on this route, but only care that the vehicle he (she) is going to take will arrive in a few minutes. In addition, given that buses have become the main mode of transportation in urban transportation, bus stop advertisements have gradually shifted from traditional page advertisements to electronic advertisements. Placing electronic advertisements during waiting time has become an important channel for companies to increase their revenue. Passengers watch advertisements while waiting. If the advertisements are too long, passengers do not finish watching them, or the quality of the advertisements is too poor, it will lead to a decrease in passengers' consumption desire. Therefore, accurate prediction of bus arrival times and reasonable advertising placement have become issues that need to be urgently addressed. Summary of the invention
[0003] The purpose of the present invention is to solve the defects mentioned in the above background technology by proposing a bus arrival information and advertising publishing management system based on big data and a method thereof.
[0004] The technical solution adopted by the present invention is as follows:
[0005] A method for managing bus arrival information and advertising based on big data, the steps are as follows:
[0006] S1: Collect bus operation information and request advertising information respectively;
[0007] S2: pre-processing the collected bus operation information and advertising information;
[0008] S3: Load and obtain bus operation information and advertising information;
[0009] S4: Publish the bus arrival information and advertising information on split screens.
[0010] As a preferred technical solution of the present invention: in S2, the collected bus operation information and advertising information are respectively cleaned and sorted.
[0011] As a preferred technical solution of the present invention: the bus operation information includes the time information of the bus arriving at each station, the distance information between each station, the operation mileage information, the operation speed information, the operation acceleration information, the operation deceleration information and the station stop time information.
[0012] As a preferred technical solution of the present invention: in S3, the bus arrival time is predicted based on the collected bus operation information and the arrival prediction algorithm.
[0013] As a preferred technical solution of the present invention: the arrival prediction algorithm is as follows:
[0014] Input the bus operation information matrix x into the LSTM network to predict the time when the vehicle arrives at each station:
[0015] e t =δ(ω i *[x t h t-1 ]+b i )
[0016] f t =δ(ω f *[x t h t-1 ]+b f )
[0017] o t =δ(ω o *[x t h t-1 ]+b o )
[0018]
[0019]
[0020] h t =o t ⊙tanh(C t )
[0021] Among them, e t 、f t , o t are the input gate, forget gate, and output gate at time t, δ is the sigmoid activation function, tanh(·) is the activation layer, and x t is the bus operation information matrix input at time t, h t-1 is the bus operation status at the previous moment, is the memory unit before updating at time t, C t is the memory unit updated at time t, C t-1 is the memory unit updated at the last moment, ω i ,ω f ,ω o ,ω C are the weights of the input gate, forget gate, output gate and memory unit, respectively, and b i、b f 、b o 、b C are the bias matrices of the input gate, forget gate, output gate and memory unit respectively, ⊙ is the Hadamard product, h t is the state information at time t;
[0022] Based on the LSTM network, the predicted time T for the bus to arrive at the jth station from the current location along the driving direction is obtained. j :
[0023]
[0024] Among them, t i+1 is the time it takes for the vehicle to arrive at the station i closest to its current location, is the travel time between the vehicle's current location and the jth stop, t mn is the travel time of the vehicle between the mth station and the nth station, is the stop time of the vehicle arriving at the jth station among a total of ji-1 stations along the way, d k is the stop time of the vehicle at the kth station.
[0025] As a preferred technical solution of the present invention: in S3, the advertising information data is loaded based on the collected advertising information data based on a dynamic loading algorithm.
[0026] As a preferred technical solution of the present invention: S3 also preloads the advertisement information and stores it in a database.
[0027] As a preferred technical solution of the present invention: the dynamic loading algorithm is as follows:
[0028] For pixel point q in the advertising information image, for the image grid w where pixel point q is located q The mapping is as follows:
[0029]
[0030] Among them, σ q The original grid projection point of pixel point q Distance error, is the unit vector of the z-axis;
[0031] w q The projection error with the screen height is the screen space error μ q , as follows:
[0032]
[0033]
[0034] ΔL=L r -L r-1
[0035] Among them, I l is the unit vector of the transformed viewpoint direction, l is the transformed viewpoint position, θ is the cone angle, L is the viewpoint position before transformation, ΔL is the change in viewpoint position, R is the current image frame, and L R is the viewpoint position before the current image frame is transformed, L R-1 is the viewpoint position before the previous image frame is transformed;
[0036]
[0037] Among them, E(L R+1 ) is the predicted position L R+1 The expected value, p(L r ) is the viewpoint position L of the rth image frame r The probability set, ΔL r is the viewpoint position change.
[0038] As a preferred technical solution of the present invention: S4 matches and publishes the advertisements to be published based on the predicted bus arrival time and the advertisement duration.
[0039] A method for managing bus arrival information and advertising based on big data, the steps are as follows:
[0040] S1: Information collection module: used to collect bus operation information and request advertising information respectively;
[0041] S2: Information processing module: used to pre-process the collected bus operation information and advertising information;
[0042] S3: Information loading module: used to load and obtain bus operation information and advertising information;
[0043] S4: Information release module: used to release bus arrival information and advertising information on split screens.
[0044] The present invention provides a bus arrival information and advertising information management system based on big data and a method thereof, which has the following beneficial effects compared with the prior art:
[0045] The present invention calculates the time when the bus arrives at the current station based on the prediction output result of the LSTM network. The LSTM network can selectively memorize valid information and delete invalid information, which can effectively overcome the RNN gradient disappearance or explosion problem, output a more accurate bus arrival prediction time, and improve the passenger experience; the received advertising information is matched with the station screen loading through the dynamic loading algorithm to improve the picture quality and the advertising quality, and the advertisements are preloaded at the same time to facilitate the matching and publishing of the advertisements in the later stage, thereby improving the passenger viewing effect and the advertising publishing effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 A method flow chart of a preferred embodiment of the present invention;
[0047] Figure 2 A system block diagram of a preferred embodiment of the present invention;
[0048] The meanings of the marks in the figure are as follows: 100, information acquisition module; 200, information processing module; 300, information loading module; 400, information publishing module. DETAILED DESCRIPTION
[0049] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of this embodiment can be combined with each other. The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0050] Reference Figure 1 The preferred embodiment of the present invention provides a method for managing bus arrival information and advertising based on big data, and the steps are as follows:
[0051] S1: Collect bus operation information and request advertising information respectively;
[0052] S2: pre-processing the collected bus operation information and advertising information;
[0053] S3: Load and obtain bus operation information and advertising information;
[0054] S4: Publish the bus arrival information and advertising information on split screens.
[0055] In S2, the collected bus operation information and advertisement information are cleaned and sorted respectively.
[0056] The bus operation information includes the time information of the bus arriving at each station, the distance information between each station, the operation mileage information, the operation speed information, the operation acceleration information, the operation deceleration information and the bus station stop time information.
[0057] The bus operation information collected in S3 is used to predict the bus arrival time based on the arrival prediction algorithm.
[0058] The arrival prediction algorithm is as follows:
[0059] Input the bus operation information matrix x into the LSTM network to predict the time when the vehicle arrives at each station:
[0060] e t =δ(ω i *[x t h t-1 ]+b i )
[0061] f t =δ(ω f *[x t h t-1 ]+b f )
[0062] o t =δ(ω o *[x t h t-1 ]+b o )
[0063]
[0064]
[0065] h t =o t ⊙tanh(C t )
[0066] Among them, e t 、f t , o t are the input gate, forget gate, and output gate at time t, δ is the sigmoid activation function, tanh(·) is the activation layer, and x t is the bus operation information matrix input at time t, h t-1 is the bus operation status at the previous moment, is the memory unit before updating at time t, C t is the memory unit updated at time t, C t-1 is the memory unit updated at the last moment, ω i ,ω f ,ωo ,ω C are the weights of the input gate, forget gate, output gate and memory unit, respectively, and b i 、b f 、b o 、b C are the bias matrices of the input gate, forget gate, output gate and memory unit respectively, ⊙ is the Hadamard product, h t is the state information at time t;
[0067] Based on the LSTM network, the predicted time T for the bus to arrive at the jth station from the current location along the driving direction is obtained. j :
[0068]
[0069] Among them, t i+1 is the time it takes for the vehicle to arrive at the station i closest to its current location, is the travel time between the vehicle's current location and the jth stop, t mn is the travel time of the vehicle between the mth station and the nth station, is the stop time of the vehicle arriving at the jth station among a total of ji-1 stations along the way, d k is the stop time of the vehicle at the kth station.
[0070] In the S3, the advertisement information data is loaded based on the collected advertisement information data based on a dynamic loading algorithm.
[0071] The S3 also preloads the advertisement information and stores it in the database.
[0072] The dynamic loading algorithm is as follows:
[0073] For pixel point q in the advertising information image, for the image grid w where pixel point q is located q The mapping is as follows:
[0074]
[0075] Among them, σ q The original grid projection point of pixel point q Distance error, is the unit vector of the z-axis;
[0076] w q The projection error with the screen height is the screen space error μ q , as follows:
[0077]
[0078]
[0079] ΔL=L r -L r-1
[0080] Among them, I l is the unit vector of the transformed viewpoint direction, l is the transformed viewpoint position, θ is the cone angle, L is the viewpoint position before transformation, ΔL is the change in viewpoint position, R is the current image frame, and L R is the viewpoint position before the current image frame is transformed, L R-1 is the viewpoint position before the previous image frame is transformed;
[0081]
[0082] Among them, E(L R+1 ) is the predicted position L R+1 The expected value, p(L r ) is the viewpoint position L of the rth image frame r The probability set, ΔL r is the viewpoint position change.
[0083] The S4 matches and publishes the advertisement to be published based on the predicted bus arrival time and the advertisement duration.
[0084] Reference Figure 2 , a bus arrival information and advertising publishing management system based on big data and a method thereof, comprising:
[0085] Information collection module 100: used to collect public transportation operation information and request advertisement information respectively;
[0086] Information processing module 200: used to pre-process the collected bus operation information and advertising information;
[0087] Information loading module 300: used to load and obtain bus operation information and advertising information;
[0088] Information publishing module 400: used to publish bus arrival information and advertising information on split screens.
[0089] In this embodiment, the information collection module 100 is used to collect bus operation information and request advertising information respectively, and the information processing module 200 performs data cleaning and data sorting on the collected bus operation information and advertising information respectively to ensure the accuracy of the published bus arrival information and advertising release information.
[0090] The information loading module 300 predicts the bus arrival time based on the collected bus operation information and the arrival prediction algorithm:
[0091] Input the bus operation information matrix x into the LSTM network to predict the time when the vehicle arrives at each station:
[0092] e t =δ(ω i *[x t h t-1 ]+b i )
[0093] f t =δ(ω f *[x t h t-1 ]+b f )
[0094] o t =δ(ω o *[x t h t-1 ]+b o )
[0095]
[0096]
[0097]
[0098] Among them, e t 、f t , o t are the input gate, forget gate, and output gate at time t, δ is the sigmoid activation function, tanh(·) is the activation layer, and x t is the bus operation information matrix input at time t, h t-1 is the bus operation status at the previous moment, is the memory unit before updating at time t, C t is the memory unit updated at time t, C t-1 is the memory unit updated at the last moment, ω i ,ω f ,ω o ,ω C are the weights of the input gate, forget gate, output gate and memory unit, respectively, and b i 、b f 、b o 、b C are the bias matrices of the input gate, forget gate, output gate and memory unit respectively, ⊙ is the Hadamard product, h t is the state information at time t;
[0099] Let j = 5, and obtain the predicted time T of the bus from the current location to the jth station along the travel direction based on the LSTM network. j:
[0100]
[0101] Among them, t i+1 is the time it takes for the vehicle to arrive at the station i closest to its current location, is the travel time between the vehicle's current location and the fifth stop, t mn is the travel time of the vehicle between the mth station and the nth station, is the stop time of the vehicle when it arrives at the 5th station among the 4-i stations along the way, d k is the stop time of the vehicle at the kth station.
[0102] The information loading module 300 calculates the time when the bus arrives at the current station based on the prediction output results of the LSTM network. The LSTM network can selectively memorize valid information and delete invalid information, which can effectively overcome the RNN gradient disappearance or explosion problem, output a more accurate bus arrival prediction time, and improve the passenger experience.
[0103] The information loading module 300 also loads the advertising information data based on the collected advertising information data based on a dynamic loading algorithm:
[0104] For pixel point q in the advertising information image, for the image grid w where pixel point q is located q The mapping is as follows:
[0105]
[0106] Among them, σ q The original grid projection point of pixel point q Distance error, is the unit vector of the z-axis;
[0107] w q The projection error with the screen height is the screen space error μ q , as follows:
[0108]
[0109]
[0110] ΔL=L r -L r-1
[0111] Among them, I l is the unit vector of the transformed viewpoint direction, l is the transformed viewpoint position, θ is the cone angle, L is the viewpoint position before transformation, ΔL is the change in viewpoint position, R is the current image frame, and L Ris the viewpoint position before the current image frame is transformed, L R-1 is the viewpoint position before the previous image frame is transformed;
[0112]
[0113] Among them, E(L R+1 ) is the predicted position L R+1 The expected value, p(L r ) is the viewpoint position L of the rth image frame r The probability set, ΔL r is the viewpoint position change amount; the dynamic loading algorithm can also preload the advertising information and store it in the database, and the information publishing module 400 also matches and publishes the advertisements to be published based on the predicted bus arrival time and the advertisement duration.
[0114] The dynamic loading algorithm matches the received advertising information to the site screen for loading, which can improve the picture quality, improve the quality of advertising, preload the advertisements, facilitate the matching and publishing of the subsequent advertisements, and improve the advertising publishing effect.
[0115] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.
[0116] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.
Claims
1. A method for managing bus arrival information and advertising based on big data, the steps are as follows: S1: Collect bus operation information and request advertising information respectively; S2: Preprocessing the collected bus operation information and advertising information, and performing data cleaning and data sorting on the collected bus operation information and advertising information respectively. The bus operation information includes: The time information of bus arrival at each station, the distance between stations, the running mileage information, the running speed information, the running acceleration information, the running deceleration information and the stop time information of the station; S3: Load and obtain bus operation information and advertising information. The advertising information data is loaded based on a dynamic loading algorithm. The bus operation information is used to predict the bus arrival time based on an arrival prediction algorithm. The arrival prediction algorithm is as follows: Input the bus operation information matrix x into the LSTM network to predict the arrival time of the vehicle at each station: e t =δ(ω i *[x t h t-1 ]+b i ) f t =δ(ω f *[x t h t-1 ]+b f ) the t =δ(ω o *[x t h t-1 ]+b o ) h t =o t ☉tanh(C t ) Among them, e t 、f t , o t are the input gate, forget gate, and output gate at time t, δ is the sigmoid activation function, tanh(·) is the activation layer, and x t is the bus operation information matrix input at time t, h t-1 is the bus operation status at the previous moment, is the memory unit before updating at time t, c t is the memory unit updated at time t, C t-1 is the memory unit updated at the last moment, ω i ,ω f ,ω o ,ω C are the weights of the input gate, forget gate, output gate and memory unit, respectively, and b i 、b f 、b o 、b C are the bias matrices of the input gate, forget gate, output gate and memory unit respectively, ⊙ is the Hadamard product, h t is the state information at time t; Based on the LSTM network, the predicted time T for the bus to arrive at the jth station from the current location along the driving direction is obtained. j : Among them, t i+1 is the time it takes for the vehicle to arrive at the station i closest to its current location, is the travel time between the vehicle's current location and the jth stop, t mn is the travel time of the vehicle between the mth station and the nth station, is the stop time of the vehicle arriving at the jth station among a total of ji-1 stations along the way, d k is the stop time of the vehicle at the kth station; The dynamic loading algorithm is as follows: For pixel point q in the advertising information image, for the image grid w where pixel point q is located q The mapping is as follows: Among them, σ q The original grid projection point of pixel point q Distance error, is the unit vector of the z axis; w q The projection error with the screen height is the screen space error μ q , as follows: <h2 style=";text-align:left;direction:ltr">AL=L<h2 style=";text-align:left;direction:ltr"> r <h2 style=";text-align:left;direction:ltr"> -L<h2 style=";text-align:left;direction:ltr"> r-1 Among them, I l is the unit vector of the transformed viewpoint direction, l is the transformed viewpoint position, θ is the cone angle, L is the viewpoint position before transformation, ΔL is the change in viewpoint position, R is the current image frame, and L R is the viewpoint position before the current image frame is transformed, L R-1 is the viewpoint position before the previous image frame is transformed; Among them, E(L R+1 ) is the predicted position L R+1 The expected value, p(L r ) is the viewpoint position L of the rth image frame r The probability set, ΔL r is the viewpoint position change amount; specifically, after the received advertising information is matched and loaded on the station screen through a dynamic loading algorithm, the advertisement to be released is matched and released based on the predicted bus arrival time and the advertisement duration; S4: Publish the bus arrival information and advertising information on split screens.
2. According to the method for managing bus arrival information and advertising based on big data in claim 1, it is characterized by: The S3 also preloads the advertisement information and stores it in the database.
3. A bus arrival information and advertising publishing management system based on big data, based on the bus arrival information and advertising publishing management method based on big data according to claim 1, characterized in that: Specifically include: Information collection module (100): used to collect public transport operation information and request advertisement information respectively; Information processing module (200): used for pre-processing the collected bus operation information and advertising information; Information loading module (300): used to load and obtain bus operation information and advertising information; Information publishing module (400): used to publish bus arrival information and advertising information on split screens.
