CN111222381A - User travel mode identification method and device, electronic equipment and storage medium - Google Patents
User travel mode identification method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN111222381A CN111222381A CN201811424962.4A CN201811424962A CN111222381A CN 111222381 A CN111222381 A CN 111222381A CN 201811424962 A CN201811424962 A CN 201811424962A CN 111222381 A CN111222381 A CN 111222381A
- Authority
- CN
- China
- Prior art keywords
- target user
- travel
- user
- track
- grid
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the invention provides a user travel mode identification method, a user travel mode identification device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring a travel track of a target user, and dividing an area containing the travel track into grids; calculating the residence time of the target user in each grid, and combining the residence time of the target user in each grid to obtain a trajectory feature matrix of the target user; and identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user. The embodiment of the invention realizes the automatic identification of the user travel mode and improves the accuracy of the identification of the user travel mode.
Description
Technical Field
The embodiment of the invention belongs to the technical field of intelligent transportation, and particularly relates to a user travel mode identification method and device, electronic equipment and a storage medium.
Background
At present, two methods are mainly used for acquiring the resident trip mode, namely, the traditional resident trip mode investigation and the identification model construction through the position information and the time information of the smart phone.
Most of the resident trip mode surveys adopt a paper questionnaire or an online survey mode, and residents fill in own preference trip modes by themselves, so that the resident trip mode surveys subjectively. Due to the rapid popularization of a Global Positioning System (GPS), a method for constructing an identification model by using position information and time information of a smart phone together becomes an effective new way for obtaining a user trip mode. The method mainly uses the idea of detecting the change of displacement within a certain time interval as a main body, or superposes various rules, or applies various data mining models, and comprehensively judges the travel traffic mode according to displacement indexes such as residence point information, speed and the like. In the aspect of inputting a data source, the prior art mostly uses GPS track data, and GPS equipment is adopted to collect the GPS track data. In the aspect of displacement indexes, the prior art scheme includes primary indexes such as average speed, maximum speed, mode of speed, travel distance and the like, and also includes composite indexes with high complexity such as 95 quantiles of speed, median of positive acceleration, signal quality, low speed point proportion, mean square direction change and the like, and all displacement indexes formed after track data are quantized are numerical variables. In the aspect of identification results, the conventional machine learning model is established by using displacement index features in the prior art, and is identified aiming at various traffic travel modes, in particular to various traffic modes such as walking, bicycles, buses, automobiles, subways and the like.
The resident trip mode survey is influenced by subjective consciousness of a surveyed person in practice, phenomena of missed report and wrong report are prone to occurring, quality of survey data is affected, and meanwhile the problems of high cost, large workload, low recovery rate, long processing period and the like exist. The traditional machine learning algorithm, such as a decision tree classifier, a Bayes classifier, a random forest, a support vector machine and the like, is used, the selection of characteristic variables is relied on to a great extent, and the variables can only extract characteristic information of displacement, so that the travel mode of a user is not accurately identified.
Disclosure of Invention
In order to overcome the problems of time and labor waste and inaccurate identification result of the existing user travel mode identification method or at least partially solve the problems, embodiments of the present invention provide a user travel mode identification method, an apparatus, an electronic device and a storage medium.
According to a first aspect of the embodiments of the present invention, there is provided a user travel mode identification method, including:
acquiring a travel track of a target user, and dividing an area containing the travel track into grids;
calculating the residence time of the target user in each grid, and combining the residence time of the target user in each grid to obtain a trajectory feature matrix of the target user;
and identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
According to a second aspect of the embodiments of the present invention, there is provided a user travel mode identification apparatus, including:
the dividing module is used for acquiring a travel track of a target user and dividing an area containing the travel track into grids;
the calculation module is used for calculating the residence time of the target user in each grid, combining the residence time of the target user in each grid and acquiring a track characteristic matrix of the target user;
and the identification module is used for identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
According to a third aspect of the embodiments of the present invention, there is also provided an electronic apparatus, including:
at least one processor; and
at least one memory communicatively coupled to the processor, wherein:
the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the user travel pattern recognition method provided by any one of the various possible implementations of the first aspect.
According to a fourth aspect of the embodiments of the present invention, there is further provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the user travel pattern recognition method provided in any one of the various possible implementation manners of the first aspect.
The embodiment of the invention provides a user travel mode identification method, a device, electronic equipment and a storage medium, wherein the method comprises the steps of dividing an area containing a travel track of a target user into grids, calculating the residence time of the target user in each grid, generating a track image with the residence time as a gray value, namely a track characteristic matrix of the target user, converting the travel track characteristic of the user into a track characteristic matrix containing positions and residence time under the positions to form a track image containing rich information, then extracting image characteristics and space-time characteristics from the track image by using an image identification technology convolutional neural network, more completely describing the integral travel track change of the user, automatically identifying the travel mode of the user according to the extracted characteristics, and improving the accuracy of the user travel mode identification.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic overall flow chart of a user travel mode identification method according to an embodiment of the present invention;
fig. 2 is a schematic flow chart illustrating a process of identifying a user travel mode by using a convolutional neural network in the user travel mode identification method according to the embodiment of the present invention;
fig. 3 is a schematic view of an overall structure of a user travel mode identification apparatus according to an embodiment of the present invention;
fig. 4 is a schematic view of an overall structure of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In an embodiment of the present invention, a user travel mode identification method is provided, and fig. 1 is a schematic overall flow chart of the user travel mode identification method provided in the embodiment of the present invention, where the method includes: s101, obtaining a travel track of a target user, and dividing an area containing the travel track into grids;
the target user is a user needing to identify the travel mode of the target user. The travel track of the target user can be obtained according to GPS positioning, and this embodiment is not limited to the method for obtaining the travel track of the target user. The area containing the travel track of the target user is a rectangular area containing the geographic range of the travel track of the target user, and the rectangular area is preferably a circumscribed rectangular area of the travel track of the target user. And dividing the area containing the travel track of the target user into two-dimensional grids. For example, the area containing the travel trajectory of the target user is divided into two-dimensional grids of 136 × 136 with 0.01 degree longitude and 0.01 degree latitude as scales. In addition, the travel track of the target user is matched with various traffic route maps, and the travel route of the target user can be identified.
S102, calculating the residence time of the target user in each grid, combining the residence time of the target user in each grid, and acquiring a track characteristic matrix of the target user;
the residence time of the target user in each grid is the time required for the target user to experience the travel trajectory in each grid, and this embodiment is not limited to the method for calculating the residence time of the target user in each grid. Each grid in the area is regarded as a pixel, and the residence time of the user in each grid is regarded as the gray value of each pixel. And taking an image formed by each pixel as a track image, namely a track characteristic matrix of the target user, wherein each element in the track characteristic matrix corresponds to the grid one by one. When the divided two-dimensional grid is 136 × 136, the size of the trajectory feature matrix is also 136 × 136.
S103, identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
And taking the track characteristic matrix of the target user as the input of the convolutional neural network, and taking the output of the convolutional neural network as the travel mode of the target user. Specifically, the probability that the travel mode of the target user belongs to each preset travel mode is obtained, and the preset travel mode with the maximum probability is used as the travel mode of the target user.
According to the embodiment of the invention, the area containing the travel track of the target user is divided into grids, the residence time of the target user in each grid is calculated, and the track image taking the residence time as a gray value, namely the track characteristic matrix of the target user is generated, so that the travel track characteristics of the user are converted into the track characteristic matrix containing positions and residence times under the positions, the track image containing rich information is formed, then the image characteristics and the space-time characteristics are extracted from the track image by using the convolutional neural network of the image recognition technology, the integral travel track change of the user is described more completely, the travel mode of the user is automatically recognized according to the extracted characteristics, and the accuracy of the travel mode recognition of the user is improved.
On the basis of the foregoing embodiment, the step of acquiring the travel track of the target user in this embodiment specifically includes: when the GPS data of the target user does not exist or is wrong, the signaling data of the client of the target user is obtained, and the travel track of the target user is determined according to the position of the base station generating the signaling data.
In particular, GPS data acquisition has instability and is not highly resilient to travel trajectories. Currently, most of GPS data is obtained based on a mobile phone APP. However, the user of the mobile phone does not always turn on the positioning function, and therefore, the local loss of the track is likely to be caused. In addition, the positioning deviation of the GPS indoors or underground is large, which may affect the identification accuracy. Therefore, in the embodiment, when the GPS data of the target user does not exist or has an error, the travel track of the target user is determined based on the signaling data, and the defects of the GPS data are made up by the characteristics of wide distribution and high completeness of the signaling data, so that the acquired travel track of the user is more complete. The signaling data of the target user's client, such as a mobile phone, includes short messages, calls, and records of LAC (location area Code) area switching or communication with the base station at preset time intervals. The LAC area typically contains multiple base station cells. The location of the base station that generated the signaling data is taken as the location experienced by the target user. And determining the travel track of the target user according to the experienced position of the target user. And when the GPS data of the target user exists and is correct, determining the travel track of the target user by using the GPS data.
Since the travel trajectory of the target user is related to privacy, desensitization is required before analysis according to the user travel trajectory. Specifically, through technologies such as encryption and fuzzy, the IMSI (International Mobile subscriber identity Number), the Mobile phone Number and the user identifier in the signaling data are encrypted, so that the corresponding relationship between the user accurate information and the travel track is removed, and data desensitization is realized.
On the basis of the foregoing embodiment, in this embodiment, when the GPS data of the target user does not exist or has an error, the step of acquiring the signaling data of the client of the target user, and determining the travel trajectory of the target user according to the position of the base station that generates the signaling data specifically includes: when GPS data of a user does not exist or has errors, if a base station interacting with a client of a target user is switched, acquiring first signaling data for generating the base station switching; and determining the travel track of the target user according to the position of the base station generating each first signaling data.
Specifically, if the target user does not move or moves a small distance in a certain period of time, the signaling data will be generated only under the same base station in the certain period of time. In the embodiment of the invention, the travel track of the target user is determined according to the position of the base station generating the signaling data, so that only one piece of signaling data under the same base station needs to be reserved, and other signaling data under the same base station are removed, thereby greatly reducing the calculated amount. In order to calculate the residence time of the target user in each grid more accurately, only the first signaling data of base station switching generated under the same base station is reserved. And determining the travel track of the target user according to the position of the base station generating each first signaling data.
On the basis of the foregoing embodiment, in this embodiment, the step of calculating the residence time of the target user in each grid specifically includes: acquiring time for generating first signaling data of any two adjacent base station switches; taking the time interval between the two times as the residence time of the target user in the coverage area of the base station which generates the first signaling data in the previous time of the two adjacent times; and acquiring the residence time of the target user in each grid according to the residence time corresponding to the base station in each grid.
Specifically, a time interval between times of generating first signaling data of any two adjacent base station handovers is obtained, and the signaling data generated first in the two signaling data is taken as the previous signaling data, and the time interval is taken as the residence time of the target user in the coverage area of the base station generating the previous signaling data. And acquiring the base stations in each grid according to the positions of the base stations generating the signaling data. And acquiring the residence time of the target user in each grid according to the residence time of the target user in the coverage area of the base station in each grid. Some base stations have large coverage areas and some base stations have small coverage areas. For any grid containing the travel track of the target user, if the grid is covered by base stations in other grids except the grid, the number of the grids covered by the base stations and containing the travel track is obtained, and the average value obtained by dividing the residence time corresponding to the base station by the value of the number of the grids is used as the residence time of the target user in the grid. For any grid containing the travel track of the target user, if the grid contains a plurality of base stations and the base stations do not cover grids containing the travel track except the grid, adding the residence time lengths corresponding to the base stations to be used as the residence time length of the target user in the grid.
On the basis of the foregoing embodiment, in this embodiment, after the step of calculating the residence time of the target user in each grid, the method further includes: and if the residence time of the target user in each grid is greater than 0, taking the logarithm of the residence time of the target user in each grid.
Specifically, in order to obtain a better application effect, if the residence time of the target user in each grid is greater than 0, the residence time of the target user in each grid is logarithmized and then used as the gray value corresponding to each grid. And if the logarithm result is more than 0 and less than 1, considering the logarithm result as 1. And if the residence time of the target user in each grid is equal to 0, not taking the logarithm.
On the basis of the foregoing embodiments, in this embodiment, before the step of identifying the travel mode of the target user by using the trained convolutional neural network according to the trajectory feature matrix of the target user, the method further includes: obtaining a motor vehicle travel track sample according to the displacement track of the on-board network card of the motor vehicle; acquiring a connecting line between any adjacent subway base stations, and taking the intersection of the travel track of the user using the subway APP and the connecting line as a subway travel track sample; the subway APP is used for entering a subway station in a code scanning mode; taking the travel track of the user sample using the sharing bicycle APP as a non-motor vehicle travel track sample; acquiring a working place and a residence of a user sample according to a resident base station of a user in a first preset time period and a second preset time period, and if the distance between the working place and the residence is smaller than a first preset threshold value and the average speed of the target user sample in the commuting period is smaller than a second preset threshold value, taking the travel track of the user sample as a walking travel track sample; and training a convolutional neural network according to the motor vehicle travel track sample, the subway travel track sample, the non-motor vehicle travel track sample and the walking travel track sample.
Specifically, the present embodiment considers the travel modes of all users into four categories of subway, motor vehicle, non-motor vehicle, and walking. In order to obtain reliable training samples, the training samples and the testing samples with higher confidence coefficient are determined in a mode of combining position and APP analysis. For example, taking the travel track of 6-10 points of the users in the internet of vehicles as a motor vehicle travel track sample. The car networking card is generally installed on a motor vehicle, the moving track of the car networking card represents the moving route of the motor vehicle, and the reliability is high. Through URL analysis, 6-10 mobile phone users who use subway APP are obtained, if the mobile phone users of the APP such as the metro metropolitan meeting are matched with the subway base station in behavior track, the intersection of the behavior track and the subway base station is taken as the subway trip track, and the reliability is high. Wherein the position of the subway base station is marked in advance. And through URL analysis, 6-10 users who use the sharing bicycle APP are obtained, the users judged to be subway travel are removed, and the travel track of the users is taken as the non-motor vehicle travel track. The first preset time period is a time period of working time in the day, the second preset time period is a time period of rest time at night, and the working place and the living place of the user are judged according to the base station of which the interaction time with the user is greater than a preset threshold value in the two time periods. For example, if the distance between the work place and the residence place is not more than 2 km and the average speed during the commute is not more than 10km/h, the users are determined to be walking users, users who are determined to be traveling on the subway and the non-motor vehicle are removed, and the traveling track is taken as the walking traveling track. And (3) labeling the travel track samples in the four travel modes with the travel modes, randomly extracting a plurality of travel track samples as training samples, and randomly extracting a plurality of travel track samples as test samples. For example, 40000 samples are randomly drawn as training samples and 20000 samples are drawn as test samples. The convolutional neural network model can be trained with the TensorFlow platform. In the actual training process, all samples are divided into a plurality of parts, each part comprises a certain number of samples, and then a part of training samples are updated each time to participate in operation, so that the convergence speed is accelerated. The TensorFlow can store the trained convolutional neural network model, thereby realizing quick multiplexing and strengthened training.
On the basis of the foregoing embodiments, in this embodiment, the step of identifying the travel mode of the target user by using the trained convolutional neural network according to the trajectory feature matrix of the target user specifically includes: using the convolution layer and the pooling layer of the convolutional neural network to extract the characteristics of the track characteristic matrix of the target user; converting the extracted features into vectors using a first fully-connected layer of the convolutional neural network; acquiring the prediction probability of each preset travel mode according to the vector by using a second full-connection layer of the convolutional neural network; and determining the travel mode of the target user according to the prediction probability of each preset travel mode.
In particular, the present embodiments are not limited to the number and size of convolutional and pooling layers in convolutional neural networks. For example, a travel track of the target user is obtained, and an area containing the travel track is divided into 136 × 136 grids. And acquiring a 136 x 136 track feature matrix according to the residence time of the target user in each grid. As shown in fig. 2, the trace feature matrix size of 136 × 136 was input to the first convolutional layer of the convolutional neural network, the filter size of the first convolutional layer was 5 × 5 and the depth was 32, and the edge of the convolution result was filled with 0 so that the output size of the first convolutional layer was still 136 × 136 and the depth was 32. The matrix of 136 x 32 output from the first convolutional layer is reduced in dimension using the first pooling layer of the convolutional neural network, the size of the first pooling layer filter is set to 2 x 2, and the length and width steps are also set to 2, so the first pooling layer outputs the matrix of 68 x 32 as input to the second convolutional layer of the convolutional neural network. The second convolution layer has a filter size of 5 x 5 and a depth of 64 and fills the edges of the convolution result with 0 to make this layer output a matrix of 68 x 64. The output of the second convolutional layer was reduced in dimension using the second pooling layer of the convolutional neural network, the filter size of the second pooling layer was set to 2 x 2, the length and width steps were also set to 2, and thus a matrix of outputs 34 x 64. The elements in the 34 x 64 matrix output by the second pooled layer are straightened into vectors, i.e., the matrix is vectorized, using the first fully-connected layer of the convolutional neural network. The first fully connected layer has a total of 34 × 64 — 73984 input nodes, and 1024 output nodes. The input nodes of the second right connecting layer are 1024, the output nodes are 4, and the prediction probabilities of the four travel modes correspond to the input nodes and the output nodes. And the output layer of the convolutional neural network outputs the travel mode with the maximum prediction probability in the second full-connection layer as the recognition result.
In addition, the longitude and latitude of part of the overhead access points are obtained through a Baidu map API, and the base station which is closest to each access point and is less than a third preset threshold value is calculated to be the overhead base station. The method can identify the traveling track of the motor vehicle approaching the overhead road.
In another embodiment of the present invention, a user travel mode identification device is provided, which is used to implement the methods in the foregoing embodiments. Therefore, the description and definition in the embodiments of the user travel mode identification method may be used for understanding each execution module in the embodiments of the present invention. Fig. 3 is a schematic view of an overall structure of a user travel mode identification apparatus according to an embodiment of the present invention, where the apparatus includes a dividing module 301, a calculating module 302, and an identifying module 303; wherein:
the dividing module 301 is configured to obtain a travel track of a target user, and divide an area including the travel track into grids;
the target user is a user needing to identify the travel mode of the target user. The travel track of the target user can be obtained according to GPS positioning, and this embodiment is not limited to the method for obtaining the travel track of the target user. The area containing the travel track of the target user is a rectangular area of the geographic range containing the travel track of the target user. The dividing module 301 divides an area containing a travel trajectory of a target user into two-dimensional grids.
The calculation module 302 is configured to calculate a residence time of the target user in each grid, combine the residence times of the target user in each grid, and obtain a trajectory feature matrix of the target user;
the residence time of the target user in each grid is the time required for the target user to experience the travel trajectory in each grid, and this embodiment is not limited to the method for calculating the residence time of the target user in each grid. The calculation module 302 calculates the residence time of the target user in each grid, regarding each grid in the area as a pixel, and regarding the residence time of the user in each grid as the gray value of each pixel. And taking an image formed by each pixel as a track image, namely a track characteristic matrix of the target user, wherein each element in the track characteristic matrix corresponds to the grid one by one.
The identification module 303 is configured to identify a travel mode of the target user by using a trained convolutional neural network according to the trajectory feature matrix of the target user.
The recognition module 303 uses the trajectory feature matrix of the target user as an input of the convolutional neural network, and outputs the convolutional neural network as a travel mode of the target user. Specifically, the probability that the travel mode of the target user belongs to each preset travel mode is obtained, and the preset travel mode with the maximum probability is used as the travel mode of the target user.
According to the embodiment of the invention, the area containing the travel track of the target user is divided into grids, the residence time of the target user in each grid is calculated, and the track image taking the residence time as a gray value, namely the track characteristic matrix of the target user is generated, so that the travel track characteristics of the user are converted into the track characteristic matrix containing positions and residence times under the positions, the track image containing rich information is formed, then the image characteristics and the space-time characteristics are extracted from the track image by using the convolutional neural network of the image recognition technology, the integral travel track change of the user is described more completely, the travel mode of the user is automatically recognized according to the extracted characteristics, and the accuracy of the travel mode recognition of the user is improved.
On the basis of the foregoing embodiment, the dividing module in this embodiment is specifically configured to: when the GPS data of the target user does not exist or is wrong, the signaling data of the client of the target user is obtained, and the travel track of the target user is determined according to the position of the base station generating the signaling data.
On the basis of the foregoing embodiment, the dividing module in this embodiment is further configured to: when GPS data of a user does not exist or has errors, if a base station interacting with a client of a target user is switched, acquiring first signaling data for generating the base station switching; and determining the travel track of the target user according to the position of the base station generating each first signaling data.
On the basis of the foregoing embodiment, the calculating module in this embodiment is specifically configured to: acquiring time for generating first signaling data of any two adjacent base station switches; taking the time interval between the two times as the residence time of the target user in the coverage area of the base station which generates the first signaling data in the previous time of the two adjacent times; and acquiring the residence time of the target user in each grid according to the residence time corresponding to the base station in each grid.
On the basis of the foregoing embodiment, in this embodiment, the method further includes a processing module, configured to log the residence time of the target user in each grid if the residence time of the target user in each grid is greater than 0.
On the basis of the above embodiments, the embodiment further includes a training module, configured to obtain a motor vehicle travel track sample according to a displacement track of the vehicle networking card on the motor vehicle; acquiring a connecting line between any adjacent subway base stations, and taking the intersection of the travel track of the user using the subway APP and the connecting line as a subway travel track sample; the subway APP is used for entering a subway station in a code scanning mode; taking the travel track of the user sample using the sharing bicycle APP as a non-motor vehicle travel track sample; acquiring a working place and a residence of a user sample according to a resident base station of a user in a first preset time period and a second preset time period, and if the distance between the working place and the residence is smaller than a first preset threshold value and the average speed of the target user sample in the commuting period is smaller than a second preset threshold value, taking the travel track of the user sample as a walking travel track sample; and training a convolutional neural network according to the motor vehicle travel track sample, the subway travel track sample, the non-motor vehicle travel track sample and the walking travel track sample.
On the basis of the foregoing embodiments, the identification module in this embodiment is specifically configured to: using the convolution layer and the pooling layer of the convolutional neural network to extract the characteristics of the track characteristic matrix of the target user; converting the extracted features into vectors using a first fully-connected layer of the convolutional neural network; acquiring the prediction probability of each preset travel mode according to the vector by using a second full-connection layer of the convolutional neural network; and determining the travel mode of the target user according to the prediction probability of each preset travel mode.
The embodiment provides an electronic device, and fig. 4 is a schematic view of an overall structure of the electronic device according to the embodiment of the present invention, where the electronic device includes: at least one processor 401, at least one memory 402, and a bus 403; wherein the content of the first and second substances,
the processor 401 and the memory 402 communicate with each other via a bus 403;
the memory 402 stores program instructions executable by the processor 401, and the processor calls the program instructions to perform the methods provided by the above method embodiments, for example, the methods include: acquiring a travel track of a target user, and dividing an area containing the travel track into grids; calculating the residence time of the target user in each grid, and combining the residence time of the target user in each grid to obtain a trajectory feature matrix of the target user; and identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
The present embodiments provide a non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the methods provided by the above method embodiments, for example, including: acquiring a travel track of a target user, and dividing an area containing the travel track into grids; calculating the residence time of the target user in each grid, and combining the residence time of the target user in each grid to obtain a trajectory feature matrix of the target user; and identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A user travel mode identification method is characterized by comprising the following steps:
acquiring a travel track of a target user, and dividing an area containing the travel track into grids;
calculating the residence time of the target user in each grid, and combining the residence time of the target user in each grid to obtain a trajectory feature matrix of the target user;
and identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
2. The method according to claim 1, wherein the step of obtaining the travel track of the target user specifically comprises:
when the GPS data of the target user does not exist or is wrong, the signaling data of the client of the target user is obtained, and the travel track of the target user is determined according to the position of the base station generating the signaling data.
3. The method according to claim 2, wherein when the GPS data of the target user does not exist or is in error, the step of obtaining the signaling data of the client of the target user, and determining the travel trajectory of the target user according to the location of the base station generating the signaling data specifically includes:
when GPS data of a user does not exist or has errors, if a base station interacting with a client of a target user is switched, acquiring first signaling data for generating the base station switching;
and determining the travel track of the target user according to the position of the base station generating each first signaling data.
4. The method according to claim 3, wherein the step of calculating the residence time of the target user in each of the grids specifically comprises:
acquiring time for generating first signaling data of any two adjacent base station switches;
taking the time interval between the two times as the residence time of the target user in the coverage area of the base station which generates the first signaling data in the previous time of the two adjacent times;
and acquiring the residence time of the target user in each grid according to the residence time corresponding to the base station in each grid.
5. The method of claim 1, wherein the step of calculating the residence time of the target user in each of the grids further comprises:
and if the residence time of the target user in each grid is greater than 0, taking the logarithm of the residence time of the target user in each grid.
6. The method according to claim 1, wherein the step of identifying the travel mode of the target user by using the trained convolutional neural network according to the trajectory feature matrix of the target user further comprises:
obtaining a motor vehicle travel track sample according to the displacement track of the on-board network card of the motor vehicle;
acquiring a connecting line between any adjacent subway base stations, and taking the intersection of the travel track of the user using the subway APP and the connecting line as a subway travel track sample; the subway APP is used for entering a subway station in a code scanning mode;
taking the travel track of the user sample using the sharing bicycle APP as a non-motor vehicle travel track sample;
the method comprises the steps that a base station is stationed according to a first preset time period and a second preset time period of a user, a working place and a residence place of a user sample are obtained, and if the distance between the working place and the residence place is smaller than a first preset threshold value and the average speed of a target user sample during a commuting period is smaller than a second preset threshold value, a travel track of the user sample is used as a walking travel track sample;
and training a convolutional neural network according to the motor vehicle travel track sample, the subway travel track sample, the non-motor vehicle travel track sample and the walking travel track sample.
7. The method according to any one of claims 1 to 6, wherein the step of identifying the travel mode of the target user by using the trained convolutional neural network according to the trajectory feature matrix of the target user specifically comprises:
using the convolution layer and the pooling layer of the convolutional neural network to extract the characteristics of the track characteristic matrix of the target user;
converting the extracted features into vectors using a first fully-connected layer of the convolutional neural network;
acquiring the prediction probability of each preset travel mode according to the vector by using a second full-connection layer of the convolutional neural network;
and determining the travel mode of the target user according to the prediction probability of each preset travel mode.
8. A user travel mode recognition apparatus, comprising:
the dividing module is used for acquiring a travel track of a target user and dividing an area containing the travel track into grids;
the calculation module is used for calculating the residence time of the target user in each grid, combining the residence time of the target user in each grid and acquiring a track characteristic matrix of the target user;
and the identification module is used for identifying the travel mode of the target user by using the trained convolutional neural network according to the track characteristic matrix of the target user.
9. An electronic device, comprising:
at least one processor, at least one memory, and a bus; wherein the content of the first and second substances,
the processor and the memory complete mutual communication through the bus;
the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 7.
10. A non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811424962.4A CN111222381A (en) | 2018-11-27 | 2018-11-27 | User travel mode identification method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811424962.4A CN111222381A (en) | 2018-11-27 | 2018-11-27 | User travel mode identification method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111222381A true CN111222381A (en) | 2020-06-02 |
Family
ID=70828834
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811424962.4A Pending CN111222381A (en) | 2018-11-27 | 2018-11-27 | User travel mode identification method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111222381A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113052047A (en) * | 2021-03-18 | 2021-06-29 | 北京百度网讯科技有限公司 | Traffic incident detection method, road side equipment, cloud control platform and system |
| CN115100858A (en) * | 2022-06-20 | 2022-09-23 | 上海海事大学 | Method and system for identifying travel modes of shared bicycle and private bicycle |
| CN116887257A (en) * | 2023-08-29 | 2023-10-13 | 北京友坤科技有限责任公司 | Abuse identification method and device for vehicle-to-vehicle network card, electronic equipment and storage medium |
| CN116916295A (en) * | 2023-09-11 | 2023-10-20 | 中移(苏州)软件技术有限公司 | Resident base station determining method and device and electronic equipment |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102708680A (en) * | 2012-06-06 | 2012-10-03 | 北京交通大学 | Commute travel mode identification method based on AGPS technology |
| CN102799897A (en) * | 2012-07-02 | 2012-11-28 | 杨飞 | Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling |
| CN103810851A (en) * | 2014-01-23 | 2014-05-21 | 广州地理研究所 | Mobile phone location based traffic mode identification method |
| CN105101092A (en) * | 2015-09-01 | 2015-11-25 | 上海美慧软件有限公司 | Mobile phone user travel mode recognition method based on C4.5 decision tree |
| CN105608505A (en) * | 2015-12-22 | 2016-05-25 | 重庆邮电大学 | Cellular signaling data based track traffic travel mode identification method for resident |
| CN105682025A (en) * | 2016-01-05 | 2016-06-15 | 重庆邮电大学 | User residing location identification method based on mobile signaling data |
| CN106412835A (en) * | 2016-11-18 | 2017-02-15 | 中国联合网络通信集团有限公司 | Method for identifying user travel mode and apparatus thereof |
| WO2017140175A1 (en) * | 2016-02-18 | 2017-08-24 | 华南理工大学 | Toll road network traffic information collection and guidance system based on route identification system |
| US20170330074A1 (en) * | 2016-05-11 | 2017-11-16 | The Board Of Regents Of The Nevada System Of Higher Education On Behalf Of The University Of NV | Methods And Systems For Providing Travel Recommendations |
| CN108566618A (en) * | 2018-04-04 | 2018-09-21 | 广州杰赛科技股份有限公司 | Obtain method, apparatus, equipment and storage medium that user is resident rule |
-
2018
- 2018-11-27 CN CN201811424962.4A patent/CN111222381A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102708680A (en) * | 2012-06-06 | 2012-10-03 | 北京交通大学 | Commute travel mode identification method based on AGPS technology |
| CN102799897A (en) * | 2012-07-02 | 2012-11-28 | 杨飞 | Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling |
| CN103810851A (en) * | 2014-01-23 | 2014-05-21 | 广州地理研究所 | Mobile phone location based traffic mode identification method |
| CN105101092A (en) * | 2015-09-01 | 2015-11-25 | 上海美慧软件有限公司 | Mobile phone user travel mode recognition method based on C4.5 decision tree |
| CN105608505A (en) * | 2015-12-22 | 2016-05-25 | 重庆邮电大学 | Cellular signaling data based track traffic travel mode identification method for resident |
| CN105682025A (en) * | 2016-01-05 | 2016-06-15 | 重庆邮电大学 | User residing location identification method based on mobile signaling data |
| WO2017140175A1 (en) * | 2016-02-18 | 2017-08-24 | 华南理工大学 | Toll road network traffic information collection and guidance system based on route identification system |
| US20170330074A1 (en) * | 2016-05-11 | 2017-11-16 | The Board Of Regents Of The Nevada System Of Higher Education On Behalf Of The University Of NV | Methods And Systems For Providing Travel Recommendations |
| CN106412835A (en) * | 2016-11-18 | 2017-02-15 | 中国联合网络通信集团有限公司 | Method for identifying user travel mode and apparatus thereof |
| CN108566618A (en) * | 2018-04-04 | 2018-09-21 | 广州杰赛科技股份有限公司 | Obtain method, apparatus, equipment and storage medium that user is resident rule |
Non-Patent Citations (1)
| Title |
|---|
| 袁冠 等: "移动对象轨迹数据挖掘技术", 中国矿业大学出版社, pages: 26 - 35 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113052047A (en) * | 2021-03-18 | 2021-06-29 | 北京百度网讯科技有限公司 | Traffic incident detection method, road side equipment, cloud control platform and system |
| CN113052047B (en) * | 2021-03-18 | 2023-12-29 | 阿波罗智联(北京)科技有限公司 | Traffic event detection method, road side equipment, cloud control platform and system |
| CN115100858A (en) * | 2022-06-20 | 2022-09-23 | 上海海事大学 | Method and system for identifying travel modes of shared bicycle and private bicycle |
| CN115100858B (en) * | 2022-06-20 | 2024-05-07 | 上海海事大学 | Method and system for identifying travel modes of shared bicycle and private bicycle |
| CN116887257A (en) * | 2023-08-29 | 2023-10-13 | 北京友坤科技有限责任公司 | Abuse identification method and device for vehicle-to-vehicle network card, electronic equipment and storage medium |
| CN116887257B (en) * | 2023-08-29 | 2023-11-28 | 北京友坤科技有限责任公司 | Abuse identification method and device for vehicle-to-vehicle network card, electronic equipment and storage medium |
| CN116916295A (en) * | 2023-09-11 | 2023-10-20 | 中移(苏州)软件技术有限公司 | Resident base station determining method and device and electronic equipment |
| CN116916295B (en) * | 2023-09-11 | 2024-01-26 | 中移(苏州)软件技术有限公司 | Resident base station determining method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111222381A (en) | User travel mode identification method and device, electronic equipment and storage medium | |
| Alexander et al. | Assessing the impact of real-time ridesharing on urban traffic using mobile phone data | |
| US20210302585A1 (en) | Smart navigation method and system based on topological map | |
| CN109688532B (en) | Method and device for dividing city functional area | |
| CN108372856A (en) | Effective context-aware of perception stream in autonomous driving system | |
| CN109510851A (en) | The construction method and equipment of map datum | |
| CN108170793A (en) | Dwell point analysis method and its system based on vehicle semanteme track data | |
| CN110598917B (en) | Destination prediction method, system and storage medium based on path track | |
| CN110460962B (en) | Rail transit user identification method and device | |
| CN112556717B (en) | Travel mode screening method and travel route recommending method and device | |
| CN109859505B (en) | Early warning processing method, device, server and medium for high-speed station | |
| CN109410576A (en) | Road condition analyzing method, apparatus, storage medium and the system of multisource data fusion | |
| CN112163446A (en) | Obstacle detection method and device, electronic equipment and storage medium | |
| CN112163063A (en) | Method, apparatus and computer-readable storage medium for generating high-precision map | |
| CN110830915B (en) | Method and device for determining starting point position | |
| CN114596709A (en) | Data processing method, device, equipment and storage medium | |
| CN111582527A (en) | Travel time estimation method and device, electronic equipment and storage medium | |
| CN110708664B (en) | Traffic flow sensing method and device, computer storage medium and electronic equipment | |
| CN112351394A (en) | Traffic travel model construction method based on mobile phone signaling data | |
| CN115565376B (en) | Vehicle journey time prediction method and system integrating graph2vec and double-layer LSTM | |
| CN116129643A (en) | Bus travel characteristic identification method, device, equipment and medium | |
| CN112732446B (en) | Task processing method and device and storage medium | |
| CN114245329B (en) | Traffic mode identification method, device, equipment and storage medium | |
| CN116957005A (en) | Training method, device, equipment and medium of corresponding relation prediction network | |
| CN112566043B (en) | MAC address identification method and device, storage medium and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |