CN112686457A - Route arrival time estimation method and device, electronic equipment and storage medium - Google Patents
Route arrival time estimation method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN112686457A CN112686457A CN202110004507.4A CN202110004507A CN112686457A CN 112686457 A CN112686457 A CN 112686457A CN 202110004507 A CN202110004507 A CN 202110004507A CN 112686457 A CN112686457 A CN 112686457A
- Authority
- CN
- China
- Prior art keywords
- route
- scene
- arrival time
- estimated
- historical
- 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.)
- Granted
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Navigation (AREA)
Abstract
The embodiment of the application discloses a method and a device for estimating route arrival time, electronic equipment and a storage medium, wherein the method comprises the following steps: the method comprises the steps of obtaining overall route characteristics of a route to be predicted and at least one route scene label to which the route to be predicted belongs; acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels; and performing characteristic combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall characteristics of the route to obtain the route combination characteristics of the route to be estimated, and predicting the estimated arrival time of the route to be estimated according to the route combination characteristics. The technical scheme of the embodiment of the application can improve the estimation accuracy of the route arrival time.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for estimating route arrival time, an electronic device, and a computer-readable storage medium.
Background
The navigation software provides a function of planning a route for the user and performs voice guidance along the road for the user during the driving process of the user. Estimation of Arrival Time (ETA) is a basic function in map software, and specifically, for a route and departure Time determined on a map, the Time required for the route to be taken is given.
At present, the ETA is estimated generally by inputting the characteristic value of the whole route into a machine learning algorithm for training and prediction by utilizing the machine learning algorithm, but the accuracy of estimation of the arrival time is low due to the fact that only the overall characteristic of the route is considered in the method and only the overall characteristic of the route is considered.
Disclosure of Invention
In order to solve the above technical problem, embodiments of the present application provide a method and an apparatus for estimating route arrival time, an electronic device, and a computer-readable storage medium, which can improve accuracy of estimating the arrival time.
Other features and advantages of the present application will be apparent from the following detailed description, or may be learned by practice of the application.
According to an aspect of the embodiments of the present application, there is provided a route arrival time estimation method, including: the method comprises the steps of obtaining overall route characteristics of a route to be predicted and at least one route scene label to which the route to be predicted belongs; acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels; and performing characteristic combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall characteristics of the route to obtain the route combination characteristics of the route to be estimated, and predicting the estimated arrival time of the route to be estimated according to the route combination characteristics.
According to an aspect of the embodiments of the present application, there is provided a route arrival time estimation apparatus, including: the system comprises an acquisition module, a judgment module and a display module, wherein the acquisition module is used for acquiring the overall route characteristics of a route to be predicted and at least one route scene label to which the route to be predicted belongs; the estimation module is used for acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels; and the confirmation module is used for performing characteristic combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall characteristics of the route to obtain the route combination characteristics of the route to be estimated, and predicting the estimated arrival time of the route to be estimated according to the route combination characteristics.
According to an aspect of the embodiments of the present application, there is provided an electronic device, including a processor and a memory, where the memory stores computer-readable instructions, and the computer-readable instructions, when executed by the processor, implement the route arrival time estimation method as described above.
According to an aspect of embodiments of the present application, there is provided a computer-readable storage medium having stored thereon computer-readable instructions, which, when executed by a processor of a computer, cause the computer to execute a route arrival time estimation method as described above.
According to an aspect of embodiments herein, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the route arrival time estimation method provided in the various alternative embodiments described above.
According to the technical scheme, after the route to be predicted is determined, the route overall characteristics of the route to be predicted and at least one route scene label to which the route to be predicted belongs are obtained, then the initial predicted arrival time of the route to be predicted in each route scene is obtained according to the route overall characteristics and the route scene labels, finally, the route joint characteristics of the route to be predicted are obtained by combining the initial predicted arrival time of the route to be predicted in each route scene with the route overall characteristics, and the predicted arrival time of the route to be predicted is predicted based on the route joint characteristics. When the to-be-estimated route has various route scene labels, namely the to-be-estimated route has various route scenes, firstly, the initial estimated arrival time of the to-be-estimated route under each route scene is combined with the overall characteristics of the route to realize the integration of various route scenes of the to-be-estimated route, then, the estimated arrival time of the to-be-estimated route is predicted based on the route combined characteristics, and in the calculation process of the estimated arrival time of the to-be-estimated route, the influence of various route scenes on the calculation result of the estimated arrival time of the to-be-estimated route is fully considered, thereby effectively reducing the influence of various route scenes on the estimated arrival time of the to-be-estimated route, the accuracy of the estimated route arrival time is improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application. It is obvious that the drawings in the following description are only some embodiments of the application, and that for a person skilled in the art, other drawings can be derived from them without inventive effort. In the drawings:
fig. 1 is a schematic diagram of a route arrival time estimation system according to an embodiment of the present invention;
FIG. 2 illustrates a schematic diagram of an estimated time reachable circle according to an embodiment of the application;
FIG. 3 shows a flow diagram of a route arrival time estimation method according to an embodiment of the present application;
fig. 4 shows another flowchart illustrating the description of step S20 according to the corresponding embodiment shown in fig. 3;
FIG. 5 shows a model schematic of a single-route scenario according to one embodiment of the present application;
FIG. 6 shows a schematic diagram of a time of arrival prediction model according to an embodiment of the present application;
FIG. 7 illustrates a model diagram of a multi-route scene according to one embodiment of the present application;
FIG. 8 shows a block diagram of a route arrival time estimation apparatus according to an embodiment of the present application;
FIG. 9 illustrates a schematic structural diagram of a computer system suitable for use in implementing the electronic device of an embodiment of the present application.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.
The block diagrams shown in the figures are functional entities only and do not necessarily correspond to physically separate entities. I.e. these functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor means and/or microcontroller means.
The flow charts shown in the drawings are merely illustrative and do not necessarily include all of the contents and operations/steps, nor do they necessarily have to be performed in the order described. For example, some operations/steps may be decomposed, and some operations/steps may be combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.
Some concepts related to the embodiments of the present application are described below.
1. Artificial Intelligence (AI)
Artificial intelligence is a theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is the research of the design principle and the realization method of various intelligent machines, so that the machines have the functions of perception, reasoning and decision making. The artificial intelligence technology mainly comprises a computer vision technology, a voice processing technology, machine learning/deep learning and other directions.
2. Machine learning
Machine learning is a multi-field cross discipline, and relates to a plurality of disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and the like. The special research on how a computer simulates or realizes the learning behavior of human beings so as to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve the performance of the computer. Machine learning is the core of artificial intelligence, is the fundamental approach for computers to have intelligence, and is applied to all fields of artificial intelligence. Machine learning generally includes techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and the like.
3. Time of arrival estimation
Arrival time estimation is a basic function in map software, and the function of the arrival time estimation is to give the time required for walking a route according to the route and the departure time on the map.
4. Actual Time of Arrival (ATA)
Actual arrival time the actual arrival time of a route can be extracted from the historical data of the map service, so the machine learning algorithm can be trained using this data as a true value to estimate the arrival time of the route.
5. Route and road segment (link)
In map applications, a route is a complete line connecting a starting point and an end point, and in an actual scene, the length of a route is usually in a range of one kilometer to several tens of kilometers.
The route is expressed using a sequence of road segments. In the map data, a road is divided into line segments connected in sequence, the length of the line segments varies from several tens of meters to several kilometers, and each line segment becomes a link and is given a globally unique ID. Thus, a route in the map is a sequence of all segments in the route.
Currently, in the field of ETA estimation, the most common algorithms include a rule-based road segment-by-road accumulation method, a tree model-based method, a depth model-based method, and the like.
The rule-based road segment-by-segment accumulation method depends on manual experience, estimates the passing time of each road segment according to the conditions of the length, the speed, the traffic lights and the like of each road segment, and accumulates the passing practices of each intersection to form the total time of the whole route. However, the method is very dependent on manual experience, and the actual road conditions are very complex, so that the set rule cannot cover various situations, the given time is often inaccurate, and more importantly, the method accumulates the estimated time of each road section, so that the error of the estimated time of each road section is also accumulated, and the final result is difficult to satisfy.
The method based on the tree model is an ETA estimation method which is started in recent years, the estimation is not carried out on the whole route section by section, the characteristics of the whole route, such as the total distance in the whole route, the average speed in the whole route at the starting time, the total number of traffic lights in the whole route, the total congestion mileage and the like in the whole route, are extracted, and then the characteristics are input into a machine learning algorithm based on the tree model for training. Currently, the most common algorithm in this class is the Gradient Boosting Decision Tree (GBDT) algorithm. However, this type of algorithm only considers the overall characteristics of the route and ignores the characteristics of the individual sections. However, in real life, extreme congestion at a certain intersection can have a great influence on the arrival time of the whole route, and the ETA obtained in the case is inaccurate only by considering the whole characteristics.
The method based on the depth model is characterized in that the characteristics of the whole route are input into the depth neural network, the depth model is trained end to end through a back propagation algorithm, and then the ETA is predicted by the trained depth model.
However, the above methods all have a common problem, ETA often has different laws in different scenes, but the three common models do not fully consider the problems of the scenes, for example, the laws of early peak and property are very different, the laws of working day and holiday are different, even if the same early peak is provided, the early peak of beijing and the early peak of charles have different characteristics, so that the consideration of the scenes needs to be added into the models.
Referring to fig. 1, fig. 1 is a schematic diagram of a route arrival time estimation system according to an embodiment of the present invention, where the system includes a terminal device 100 and a server 200, and the terminal device 100 and the server 200 may communicate via a wired or wireless network, which is not limited herein.
The terminal device 100 is configured to send a request to a server, and the terminal device 100 may be an electronic device such as a smart phone, a tablet, a notebook, a computer, and the like, but is not limited thereto. The terminal device 100 is installed with a client, which may be a web page client, a client installed in the terminal device 100, or a light application embedded in a third-party application, and the like, and the type of the client is not limited in the present application.
The server 200 is a server corresponding to a client in the terminal device 100, and the server 200 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, middleware service, a domain name service, a security service, a CDN (Content Delivery Network) and a big data and artificial intelligence platform, and is applied to a cloud electronic map product to meet a processing requirement for a large data volume of an electronic map.
The following describes a scenario in which the estimation process of the route arrival time is applicable.
First usage scenario:
when a user needs to plan a route using navigation software, a start point and an end point may be input in the terminal device 100, or the start point and the end point may be directly selected, or the like. The terminal device 100, in response to the user's operation, may determine several candidate routes between the start point and the end point based on these information, calculate the estimated arrival time of each candidate route, and then select one route from the candidate routes, which has the shortest estimated arrival time, for display to the user.
Second usage scenario:
after the client in the terminal device 100 enters the navigation state, that is, the route and the end point are determined, and every predetermined time, for example, every 10 minutes, 20 minutes, or 30 minutes, the predetermined time is set, the terminal device 100 sends the positioning position to the server 200, the server 200 calculates the expected arrival time corresponding to the current time according to the route and the end point, and further calculates the travel time required by the distance and feeds the travel time back to the terminal device 100, so that the user can know the remaining travel time, and the user can conveniently schedule the travel.
A third usage scenario:
the terminal device 100 sends the starting point input by the user to the server 200, and the server 200 may calculate the end point position in each route after the estimated time according to the received position as the starting point, so as to form an estimated time reachable circle, fig. 2 shows a schematic diagram of an estimated time reachable circle, where, taking the point O as the starting point, after half an hour of driving along 4 different routes, the reachable positions a, b, c, and d are respectively, and a, b, c, and d are connected in sequence, so that the half-hour reachable circle shown in fig. 2 can be obtained. The server 200 may send the estimated time length reachable circle to the terminal device 100, or send information in the circle to the terminal device 100 based on the estimated time length reachable circle, so that the user can know the situation around the location O conveniently.
A fourth usage scenario:
when the take-out background server 200 dispatches the take-out orders, the route time consumption corresponding to each take-out order can be calculated by taking the position of the last family as a starting point and the position of the client as an end point, so that the orders can be dispatched to the deliverers better according to the route time consumption, the result of dispatching the orders is optimized, and the dispatching efficiency is improved.
Fifth usage scenario:
when a user uses a client to taxi, the terminal device 100 sends the current position to the server 200, the server 200 takes the current position of each taxi driver as a starting point and the current position of the user as a terminal point, and the time length of each taxi driver driving to the position of the user is calculated, so that taxi drivers can be better arranged to take orders, and the passenger transport efficiency is improved.
Sixth usage scenario:
the server 200 receives the information of each route input by the client 100, and then outputs an ETA value corresponding to each route for the upstream service to evaluate the quality of each route and push the optimal route to the user.
A seventh usage scenario:
the server 200 receives the information of each route input by the client 100, and then outputs the influence weight of each route on ETA for the upstream service to avoid congestion or explain the estimated time.
The method for estimating the route arrival time provided by the embodiment of the present application is described below with reference to the foregoing application scenarios.
Referring to fig. 3, an embodiment of the present application provides a method for estimating route arrival time, as shown in fig. 3, the method includes
Step S10, obtaining the overall route characteristics of the route to be estimated and at least one route scene label to which the route to be estimated belongs;
step S20, acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels;
step S30, performing characteristic combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall characteristics of the route to obtain the route combination characteristics of the route to be estimated, and predicting the estimated arrival time of the route to be estimated according to the route combination characteristics.
These steps are described in detail below.
In step S10, the route global characteristics of the route to be predicted and at least one route scene tag to which the route to be predicted belongs are obtained.
Understandably, the overall characteristic is a characteristic related to the whole route, and is used for representing the overall traffic condition of the route, and in the embodiment, the overall characteristic of the route to be estimated comprises the following steps: the System comprises road information such as total distance of the whole journey, average speed limit of the whole journey, average free flow speed of the whole journey, total number of traffic lights of the whole journey, and occupied mileage of the whole journey, and the whole journey average speed of a departure time calculated according to real-time collected Global Positioning System (GPS) data, and the whole journey average vehicle speed of 5 minutes or 10 minutes around the same time mined according to historical GPS data collected in the past months.
The route scene label of the route to be predicted comprises the following steps: the time zone labels such as morning peak, evening peak, midnight, daytime, etc., or the date labels such as working day, holiday, etc., or the city labels such as beijing, shanghai, shenzhen, guangzhou, etc., or some other scene labels, and the content of the scene label is not limited in the present application.
It should be understood that multiple route scene tags may exist in the same road segment, each route scene tag corresponds to a route scene feature, for example, beijing, early peak, and congestion, one route scene tag indicates that the city is beijing, one route scene tag indicates that the time period is early peak, and the other route scene tag indicates that the current state of the road segment is congestion, and different route scene tags may exist in different road segments.
Understandably, in this embodiment, the route scene features may be encoded by a one-hot code, which is a method for representing the category features, for example, gender may be encoded, male may correspond to (0, 1), female may correspond to (1, 0), and a one-hot code of a gender may be obtained, or season may be encoded by a one-hot code, spring may correspond to (1, 0, 0, 0), summer may correspond to (0, 1, 0, 0), autumn may correspond to (0, 0, 1, 0), winter may correspond to (0, 0, 0, 1), and a one-hot code of a season may be obtained. In this embodiment, the morning peak corresponds to (1, 0, 0, 0), the evening peak corresponds to (0, 1, 0, 0), the midnight corresponds to (0, 0, 1, 0), and the daytime corresponds to (0, 0, 0, 1).
In this embodiment, when a user inputs a start point and an end point of a route at a client, or specifies the start point and the end point on a map of the client, the completed route can be acquired, and then, according to the acquired road information and GPS data information, the corresponding overall route characteristics and at least one route scene label to which the route to be estimated belongs can be acquired.
In step S20, an initial estimated arrival time of the route to be estimated in each route scene is obtained according to the route global characteristics and the route scene tags.
In this embodiment, the initial estimated arrival time of the route to be estimated in each route scene is obtained according to the obtained route overall characteristics and the route scene tags, specifically, referring to fig. 4, fig. 4 shows another flowchart for describing step S20 according to the embodiment corresponding to fig. 3, where the flowchart includes:
step S21, acquiring a first real value vector corresponding to the overall route characteristics;
step S22, according to the route scene characteristics of the route to be estimated in the route scene identified by the route scene label, acquiring a second real value vector corresponding to the route scene label;
step S23, performing inner product on the first real value vector and the second real value vector to obtain the initial estimated arrival time of the route to be estimated in the route scene identified by the route scene tag.
It can be understood that the embedded representation technology is a feature extraction method which has emerged in recent years, and by the technology, words, users, merchants, network nodes and the like can be converted into a real-value vector, the real-value vector is correspondingly called as a word vector, a user vector and the like, and after the real-value vector is obtained, subsequent tasks such as classification, regression, recognition and the like can be completed by using a machine learning method. In this embodiment, a first real-valued vector corresponding to the global feature of the route and a second real-valued vector corresponding to the route scene tag are obtained by an embedded representation technique, where the embedded representation technique is not limited to a general neural network or a dimension reduction method, and it should be noted that, in order to perform inner product calculation on the first real-valued vector and the second real-valued vector in the following, vector dimensions of the first real-valued vector and the second real-valued vector need to be kept the same, and an initial estimated arrival time can be obtained by performing inner product on the first real-valued vector and the second real-valued vector.
In the embodiment, the integration of various route scenes of the estimated route can be realized by acquiring the initial estimated arrival time of the to-be-estimated route in different route scenes, and when the estimated arrival time of the to-be-estimated route is calculated, the influence of the various route scenes of the to-be-estimated route on the estimated arrival time can be fully considered, so that the estimated arrival time based on various route scenes and the overall characteristics of the route is obtained, and the accuracy of the estimated arrival time of the route is effectively improved.
In an exemplary embodiment, the step of obtaining a first real-valued vector corresponding to the overall route feature is performed by a first embedded representation model, and the step of obtaining a second real-valued vector corresponding to a route scene tag is performed by a second embedded representation model matching the route scene tag according to the route scene feature of the route to be predicted in the route scene identified by the route scene tag.
Further, the method further comprises:
acquiring training data, wherein the training data comprises historical route overall characteristics of historical routes and historical route scene characteristics of the historical routes in a route scene identified by a route scene label;
inputting the historical route overall characteristics to the first embedded representation model and inputting the historical route scene characteristics to the second embedded representation model;
carrying out inner product operation on the output signal of the first embedded representation model and the output signal of the second embedded representation model to obtain initial estimated arrival time of the historical route in the route scene identified by the route scene label;
and calculating a training loss value based on the initial estimated arrival time of the historical route in the route scene identified by the route scene label and the actual arrival time of the historical route, so as to update the second embedded representation model according to the training loss value, wherein the updated second embedded representation model is matched with the route scene label.
Referring to fig. 5, fig. 5 shows a model diagram of a single-route scenario according to an embodiment of the present application, in the embodiment, a first embedded representation model and a second embedded representation model need to be trained, the training data includes historical route overall characteristics of a historical route and historical route scene characteristics of the historical route in a route scenario identified by a route scenario tag, where the historical route overall characteristics of the historical route are input into the first embedded representation model, an output signal of the first embedded representation model is obtained, the historical route scene characteristics are input into the second embedded representation model, an output signal of the second embedded representation model is obtained, and an initial estimated arrival time of the historical route in the route scenario identified by the route scenario tag is obtained by performing an inner product of the output signal of the first embedded representation model and the output signal of the second embedded representation model, a training loss value may be calculated based on the obtained initial estimated arrival time and the actual arrival time of the historical route.
It should be noted that in this embodiment, scene extension may be performed on the route scene labels, for example, if a city B is newly added and supported, in fig. 5, the branches of the overall features of the route on the left side may be maintained, and only the branches of the route scene labels on the right side may be trained. In the embodiment, the scene expansion of the route scene label can be realized, so that the method for estimating the route arrival time has wide applicability and can adapt to different application scenes.
In this embodiment, the training loss value can be calculated by using a commonly used loss function, for example, a loss of square (ETA, ATA) ═ ETA-ATA2Note that an absolute value loss function loss (ETA, ATA) ═ ETA-ATA | may also be used.
In this embodiment, the first embedded representation model and the second embedded representation model are respectively updated according to the training loss value, and the commonly used optimization updating method for the embedded representation model includes: the random gradient descent method, the Newton method, the conjugate gradient method and the like adopt different embedded representation models and need to adopt different optimization methods for optimization updating.
Further, when the number of route scene labels is multiple, the training data further comprises historical route scene characteristics of historical routes under other route scenes; the method further comprises the following steps:
inputting historical route scene characteristics of the historical route in other route scenes into a third embedded model;
and calculating initial estimated arrival times of the historical route in other route scenes according to the output signals of the third embedded model and the output signals of the first embedded representation model, and calculating a training loss value based on the initial estimated arrival times of the historical route in all route scenes and the actual arrival time of the historical route.
In this embodiment, please refer to fig. 6, fig. 6 shows a schematic diagram of an arrival time estimation model according to an embodiment of the present application, in which a route scene tag a and a route scene tag B may be respectively input into corresponding embedded representation models from two branches, and then signals of the corresponding embedded representation models are output, and initial estimated arrival times in different route scenes may be obtained by performing inner products on the signals of the embedded representation models.
Further, calculating a training loss value based on the initial estimated arrival time of the historical route in all route scenarios and the actual arrival time of the historical route, including:
calculating the difference between the initial estimated arrival time and the actual arrival time of the historical route in each route scene; and summing the squares of the differences to obtain the training loss value.
For example, the loss of square function (ETA) may be used0，ETA1，ATA)＝(ETA0-ATA)2+(ETA1-ATA)2Training by calculationLoss value, wherein ETA0For initial estimated arrival time, ETA, under a route scenario label A1The initial estimated arrival time under the route scene label B, ATA is the actual arrival time of the historical route.
In step S30, the initial estimated arrival time of the route to be estimated in each route scene is feature-combined with the overall route features to obtain the route combination features of the route to be estimated, so as to predict the estimated arrival time of the route to be estimated according to the route combination features.
In this embodiment, the initial estimated arrival time of the route to be estimated in each route scene is spliced with the overall route characteristics to obtain the route joint characteristics of the route to be estimated, and then, as shown in fig. 7, fig. 7 shows a model diagram of a multipath scene according to an embodiment of the present application, and the route joint characteristics of the route to be estimated are input into a machine learning algorithm to obtain the estimated arrival time of the route to be estimated. The machine learning algorithm may be a GBDT method, a neural network method, or other methods such as linear regression.
According to the analysis, different route scene characteristics of the route to be predicted are represented as different real-valued vectors through an embedded representation technology, meanwhile, the overall characteristics of the route are also represented as the real-valued vectors, the inner product of the real-valued vectors corresponding to the route scene characteristics and the real-valued vectors corresponding to the overall characteristics of the route is used as initial predicted arrival time, and finally predicted arrival time is obtained by integrating the initial predicted arrival time under the different route scene characteristics; on the other hand, the route scene can be conveniently expanded, and when a new route scene is added, only the embedded representation model of the corresponding route scene needs to be trained; furthermore, a certain route scene can be optimized independently on the premise of not modifying other route scenes.
The following describes embodiments of an apparatus of the present application, which may be used to perform the database access control method in the above embodiments of the present application. For details not disclosed in the embodiments of the apparatus of the present application, please refer to the embodiments of the route arrival time estimation method described above in the present application.
Fig. 8 shows a block diagram of a route arrival time estimation apparatus according to an embodiment of the present application, and referring to fig. 8, a route arrival time estimation apparatus 800 according to an embodiment of the present application includes:
the obtaining module 810 is configured to obtain route overall characteristics of a route to be predicted, and at least one route scene tag to which the route to be predicted belongs; the estimation module 820 is used for acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels; the confirmation module 830 performs feature combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall features of the route to obtain the route combination features of the route to be estimated, and predicts the estimated arrival time of the route to be estimated according to the route combination features.
In one embodiment of the present application, the obtaining module 810 includes: the route overall characteristic acquiring unit is used for acquiring the overall characteristics of the route to be estimated; and the route scene label unit is used for acquiring at least one route scene label to which the route to be estimated belongs.
In one embodiment of the present application, predictor module 820 includes: the first real-valued vector acquisition unit is used for acquiring a first real-valued vector corresponding to the overall route characteristic; the second real value vector acquisition unit is used for acquiring a second real value vector corresponding to the route scene label according to the route scene characteristics of the route to be predicted in the route scene identified by the route scene label; and the initial estimated arrival time obtaining unit is used for carrying out inner product on the first real value vector and the second real value vector to obtain the initial estimated arrival time of the route to be estimated in the route scene identified by the route scene label.
In one embodiment of the present application, the validation module 830 comprises: the characteristic combination unit is used for carrying out characteristic combination on the initial estimated arrival time of the route to be estimated and the overall characteristics of the route in each route scene to obtain the route combination characteristics of the route to be estimated; and the confirmation unit is used for predicting the estimated arrival time of the route to be estimated according to the route joint characteristics.
It should be noted that the apparatus provided in the foregoing embodiment and the method provided in the foregoing embodiment belong to the same concept, and the specific manner in which each module and unit execute operations has been described in detail in the method embodiment, and is not described again here.
Embodiments of the present application further provide an electronic device, including a processor and a memory, where the memory stores computer readable instructions, and the computer readable instructions, when executed by the processor, implement the route arrival time estimation method as described above.
FIG. 9 illustrates a schematic structural diagram of a computer system suitable for use in implementing the electronic device of an embodiment of the present application.
It should be noted that the computer system 1600 of the electronic device shown in fig. 9 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.
As shown in fig. 9, computer system 1600 includes a Central Processing Unit (CPU)1601, which can perform various appropriate actions and processes, such as executing the methods described in the above embodiments, according to a program stored in a Read-Only Memory (ROM) 1602 or a program loaded from a storage portion 1608 into a Random Access Memory (RAM) 1603. In the RAM 1603, various programs and data necessary for system operation are also stored. The CPU 1601, ROM 1602, and RAM 1603 are connected to each other via a bus 1604. An Input/Output (I/O) interface 1605 is also connected to the bus 1604.
The following components are connected to the I/O interface 1605: an input portion 1606 including a keyboard, a mouse, and the like; an output section 1607 including a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, a speaker, and the like; a storage portion 1608 including a hard disk and the like; and a communication section 1609 including a Network interface card such as a LAN (Local Area Network) card, a modem, or the like. The communication section 1609 performs communication processing via a network such as the internet. The driver 1610 is also connected to the I/O interface 1605 as needed. A removable medium 1611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 1610 as necessary, so that a computer program read out therefrom is mounted in the storage portion 1608 as necessary.
In particular, according to embodiments of the application, the processes described above with reference to the flow diagrams may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising a computer program for performing the method illustrated by the flow chart. In such embodiments, the computer program may be downloaded and installed from a network via the communication portion 1609, and/or installed from the removable media 1611. When the computer program is executed by a Central Processing Unit (CPU)1601, various functions defined in the system of the present application are executed.
It should be noted that the computer readable medium shown in the embodiments of the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may be, for example, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a Read-Only Memory (ROM), an Erasable Programmable Read-Only Memory (EPROM), a flash Memory, an optical fiber, a portable Compact Disc Read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with a computer program embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in the embodiments of the present application may be implemented by software, or may be implemented by hardware, and the described units may also be disposed in a processor. Wherein the names of the elements do not in some way constitute a limitation on the elements themselves.
Another aspect of the present application also provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the route arrival time estimation method as described above. The computer-readable storage medium may be included in the electronic device described in the above embodiment, or may exist separately without being incorporated in the electronic device.
Another aspect of the application also provides a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, so that the computer device executes the route arrival time estimation method provided in the above embodiments.
The above description is only a preferred exemplary embodiment of the present application, and is not intended to limit the embodiments of the present application, and those skilled in the art can easily make various changes and modifications according to the main concept and spirit of the present application, so that the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for estimating route arrival time, comprising:
the method comprises the steps of obtaining overall route characteristics of a route to be predicted and at least one route scene label to which the route to be predicted belongs;
acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels;
and performing characteristic combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall characteristics of the route to obtain the route combination characteristics of the route to be estimated, and predicting the estimated arrival time of the route to be estimated according to the route combination characteristics.
2. The method according to claim 1, wherein the obtaining of the initial estimated arrival time of the route to be estimated in each route scene according to the route global feature and the route scene tag comprises:
acquiring a first real-valued vector corresponding to the overall route characteristic;
according to the route scene characteristics of the route to be predicted under the route scene identified by the route scene label, acquiring a second real value vector corresponding to the route scene label;
and performing inner product on the first real-value vector and the second real-value vector to obtain initial estimated arrival time of the route to be estimated in the route scene identified by the route scene label.
3. The method of claim 2, wherein the vector dimensions of the first real-valued vector and the second real-valued vector are the same.
4. The method according to claim 2, wherein the step of obtaining a first real-valued vector corresponding to the overall route feature is performed by a first embedded representation model, and the step of obtaining a second real-valued vector corresponding to the route scene tag is performed by a second embedded representation model matching with the route scene tag according to the route scene feature of the route to be predicted in the route scene identified by the route scene tag; the method further comprises the following steps:
acquiring training data, wherein the training data comprises historical route overall characteristics of historical routes and historical route scene characteristics of the historical routes under the route scenes identified by the route scene labels;
inputting the historical route overall characteristics to a first embedded representation model and inputting the historical route scene characteristics to a second embedded representation model;
performing inner product operation on the output signal of the first embedded representation model and the output signal of the second embedded representation model to obtain initial estimated arrival time of the historical route in the route scene identified by the route scene label;
calculating a training loss value based on the initial estimated arrival time of the historical route in the route scene identified by the route scene tag and the actual arrival time of the historical route, so as to update the second embedded representation model according to the training loss value, wherein the updated second embedded representation model is matched with the route scene tag.
5. The method of claim 4, further comprising historical route scenario features of the historical route under other route scenarios; the method further comprises the following steps:
inputting historical route scene characteristics of the historical route in other route scenes into a third embedded model;
calculating initial estimated arrival times of the historical route in the other route scenes according to the output signals of the third embedded model and the output signals of the first embedded representation model, and calculating the training loss value based on the initial estimated arrival times of the historical route in all route scenes and the actual arrival time of the historical route.
6. The method of claim 5, wherein calculating the training loss value based on initial estimated arrival times of the historical route in all route scenarios and actual arrival times of the historical route comprises:
calculating the difference between the initial estimated arrival time and the actual arrival time of the historical route in each route scene;
and performing summation operation on the square of the difference value to obtain the training loss value.
7. The method of claim 1, wherein the characterizing the initial estimated arrival time of the route to be estimated in each route scenario with the overall route characteristics comprises:
and splicing the initial estimated arrival time of the route to be estimated in each route scene with the overall route characteristics to obtain the route joint characteristics of the route to be estimated.
8. A route arrival time estimation device, comprising:
the system comprises an acquisition module, a judgment module and a display module, wherein the acquisition module is used for acquiring the overall route characteristics of a route to be predicted and at least one route scene label to which the route to be predicted belongs;
the estimation module is used for acquiring initial estimated arrival time of the route to be estimated in each route scene according to the overall route characteristics and the route scene labels;
and the confirmation module is used for performing characteristic combination on the initial estimated arrival time of the route to be estimated in each route scene and the overall characteristics of the route to obtain the route combination characteristics of the route to be estimated, and predicting the estimated arrival time of the route to be estimated according to the route combination characteristics.
9. An electronic device, comprising:
a memory storing computer readable instructions;
a processor to read computer readable instructions stored by the memory to perform the method of any of claims 1-7.
10. A computer-readable storage medium having computer-readable instructions stored thereon, which, when executed by a processor of a computer, cause the computer to perform the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110004507.4A CN112686457B (en) | 2021-01-04 | 2021-01-04 | Route arrival time estimation method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110004507.4A CN112686457B (en) | 2021-01-04 | 2021-01-04 | Route arrival time estimation method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112686457A true CN112686457A (en) | 2021-04-20 |
| CN112686457B CN112686457B (en) | 2022-06-03 |
Family
ID=75457095
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110004507.4A Active CN112686457B (en) | 2021-01-04 | 2021-01-04 | Route arrival time estimation method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112686457B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113902427A (en) * | 2021-12-09 | 2022-01-07 | 腾讯科技（深圳）有限公司 | Method, device and equipment for determining estimated arrival time and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU5415386A (en) * | 1985-02-28 | 1986-09-04 | Mitsubishi Denki Kabushiki Kaisha | Interframe adaptive vector quantization encoding apparatus |
| JPH07294270A (en) * | 1994-04-22 | 1995-11-10 | Sharp Corp | Route searching system |
| CN102262648A (en) * | 2010-05-31 | 2011-11-30 | 索尼公司 | Evaluation predicting device, evaluation predicting method, and program |
| CN107944375A (en) * | 2017-11-20 | 2018-04-20 | 北京奇虎科技有限公司 | Automatic Pilot processing method and processing device based on scene cut, computing device |
| CN111141302A (en) * | 2019-12-26 | 2020-05-12 | 斑马网络技术有限公司 | Estimation method and device for vehicle driving arrival time and electronic equipment |
| CN111582559A (en) * | 2020-04-21 | 2020-08-25 | 腾讯科技（深圳）有限公司 | Method and device for estimating arrival time |
| CN111667693A (en) * | 2020-06-15 | 2020-09-15 | 腾讯科技（深圳）有限公司 | Method, apparatus, device and medium for determining estimated time of arrival |
| CN111815046A (en) * | 2020-07-06 | 2020-10-23 | 北京交通大学 | Traffic flow prediction method based on deep learning |
-
2021
- 2021-01-04 CN CN202110004507.4A patent/CN112686457B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU5415386A (en) * | 1985-02-28 | 1986-09-04 | Mitsubishi Denki Kabushiki Kaisha | Interframe adaptive vector quantization encoding apparatus |
| JPH07294270A (en) * | 1994-04-22 | 1995-11-10 | Sharp Corp | Route searching system |
| CN102262648A (en) * | 2010-05-31 | 2011-11-30 | 索尼公司 | Evaluation predicting device, evaluation predicting method, and program |
| CN107944375A (en) * | 2017-11-20 | 2018-04-20 | 北京奇虎科技有限公司 | Automatic Pilot processing method and processing device based on scene cut, computing device |
| CN111141302A (en) * | 2019-12-26 | 2020-05-12 | 斑马网络技术有限公司 | Estimation method and device for vehicle driving arrival time and electronic equipment |
| CN111582559A (en) * | 2020-04-21 | 2020-08-25 | 腾讯科技（深圳）有限公司 | Method and device for estimating arrival time |
| CN111667693A (en) * | 2020-06-15 | 2020-09-15 | 腾讯科技（深圳）有限公司 | Method, apparatus, device and medium for determining estimated time of arrival |
| CN111815046A (en) * | 2020-07-06 | 2020-10-23 | 北京交通大学 | Traffic flow prediction method based on deep learning |
Non-Patent Citations (2)
| Title |
|---|
| YUAN YUAN: "Scene Recognition by Manifold Regularized Deep Learning Architecture", 《IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS》 * |
| 时朝阳: "城市道路网络中行程时间分布的估计及预测方法研究", 《中国博士学位论文全文数据库》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113902427A (en) * | 2021-12-09 | 2022-01-07 | 腾讯科技（深圳）有限公司 | Method, device and equipment for determining estimated arrival time and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112686457B (en) | 2022-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110400015B (en) | Time estimation method, device and equipment thereof | |
| CN110570678B (en) | Method and device for predicting total travel time of bus from starting point to end point | |
| CN111582559B (en) | Arrival time estimation method and device | |
| CN110942211A (en) | Prediction arrival time prediction method and device based on deep neural network | |
| CN109724611A (en) | Paths planning method, device, electronic equipment and storage medium | |
| CN107392389A (en) | Taxi dispatching processing method based on ARIMA models | |
| CN113762595B (en) | Traffic time prediction model training method, traffic time prediction method and equipment | |
| CN112071062B (en) | Driving time estimation method based on graph convolution network and graph attention network | |
| CN111667693B (en) | Method, apparatus, device and medium for determining estimated time of arrival | |
| CN112884235B (en) | Travel recommendation method, and training method and device of travel recommendation model | |
| CN115311860B (en) | Online federal learning method of traffic flow prediction model | |
| CN114881315A (en) | Travel arrival time determining method and device, electronic equipment and storage medium | |
| Xu et al. | A taxi dispatch system based on prediction of demand and destination | |
| CN112686457B (en) | Route arrival time estimation method and device, electronic equipment and storage medium | |
| Kumar et al. | Integration of exponential smoothing with state space formulation for bus travel time and arrival time prediction | |
| Zhang et al. | Gof-tte: Generative online federated learning framework for travel time estimation | |
| CN113570096A (en) | Intelligent trip method and device based on big data | |
| CN116662815B (en) | Training method of time prediction model and related equipment | |
| CN115994726B (en) | Dispatch path adjustment method, dispatch path adjustment device, electronic equipment and computer readable medium | |
| CN116629425A (en) | Method and device for calculating vehicle energy consumption, computer readable medium and electronic equipment | |
| CN116295409A (en) | Route processing method, route processing device, computer readable medium and electronic equipment | |
| US20240102813A1 (en) | Route design system, cost function learning device, designed route output device, method, and program | |
| CN113741459B (en) | Method for determining training sample and training method and device for automatic driving model | |
| CN113918802B (en) | Navigation method, device, equipment, medium and product | |
| CN110363320A (en) | A kind of method, apparatus of predictive estimation arrival time and calculate 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 | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40041941 Country of ref document: HK |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |