CN111582527A - Travel time estimation method and device, electronic equipment and storage medium - Google Patents
Travel time estimation method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN111582527A CN111582527A CN201910117958.1A CN201910117958A CN111582527A CN 111582527 A CN111582527 A CN 111582527A CN 201910117958 A CN201910117958 A CN 201910117958A CN 111582527 A CN111582527 A CN 111582527A
- Authority
- CN
- China
- Prior art keywords
- road
- time
- travel
- distribution
- travel time
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Quality & Reliability (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the invention relates to the technical field of information processing, and discloses a method and a device for estimating travel time, electronic equipment and a nonvolatile storage medium. The method for estimating the travel time comprises the following steps: acquiring the passing time of each road on a road network; determining a travel route of target distribution resources according to the passing time of each road; predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance, and the speed and the accuracy of travel time estimation can be improved.
Description
Technical Field
The present invention relates to the field of information processing technologies, and in particular, to a method and an apparatus for estimating a travel time, an electronic device, and a non-volatile storage medium.
Background
With the rapid development of the mobile internet industry, the transportation industry is also developing, especially the takeaway industry which is emerging in recent years. With the continuous improvement of the requirement of the user on the timely delivery rate of the food delivery personnel, the related technical personnel are also prompted to continuously and accurately obtain the time required by the food delivery personnel from the time of acquiring a take-out order to the time of delivering the take-out order to the order personnel.
However, the inventor finds that, in the prior art, the way of estimating the time length required for delivery is generally to pre-process the travel distance and travel time of a point pair of a specified city and store the pre-processed travel distance and travel time to a server cluster by crawling a path planning service on a map, and when estimation is required, query is performed in data stored in the cluster, wherein two positions where any two places in the city are located represent two points, and the two points can be referred to as a point pair. This approach has a long crawl time, a slow speed when estimating the rider's travel time, and is not accurate enough.
Disclosure of Invention
The invention aims to provide a method and a device for estimating travel time, electronic equipment and a nonvolatile storage medium, which can improve the speed and accuracy of travel time estimation.
In order to solve the above technical problem, an embodiment of the present invention provides a method for estimating a travel time, including: acquiring the passing time of each road on a road network; determining a travel route of target distribution resources according to the passing time of each road; predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
The embodiment of the invention also provides a device for estimating travel time, which comprises: the acquisition module is used for acquiring the passing time of each road on the road network; the determining module is used for determining a travel route of the target distribution resources according to the passing time of each road; the estimation module estimates the travel time of the target delivery resources according to the travel path and an estimation model used for estimating the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
Embodiments of the present invention further provide an electronic device, including a memory and a processor, where the memory stores a computer program, and the processor executes the program to perform: acquiring the passing time of each road on a road network;
determining a travel route of target distribution resources according to the passing time of each road; predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
Embodiments of the present invention also provide a non-volatile storage medium for storing a computer-readable program for a computer to execute the method of estimating a travel time as described above.
Compared with the prior art, the implementation mode of the invention has the main differences and the effects that: the travel route of the target distribution resources is determined through the acquired passing time of each road on the road network, and the passing time of each road is considered, so that the travel route suitable for the target distribution resources is determined in each road. According to the determined travel path and the estimation model for estimating the travel time, the travel time of the target delivery resources can be accurately estimated. Moreover, the prediction model is obtained by training according to the collected historical data of the distribution resources in advance, namely the data for training the prediction model is derived from the real historical data of the distribution resources, so that the reference value is high, and the prediction result is more accurate and reliable. Meanwhile, the embodiment of the invention does not depend on crawler data, avoids the problem of low speed of estimated travel time caused by long crawler crawling time, and improves the estimated speed to a certain extent.
In addition, acquiring the passing time of each road on the road network specifically comprises: acquiring distribution resource track data on each road on the road network; and acquiring the passing time of the distribution resources of each road on the road network according to the acquired distribution resource track data. The method is favorable for obtaining the passing time of each road aiming at the distribution resources according to the distribution resource track data, thereby being favorable for determining the travel route adapting to the target distribution resources.
In addition, the determining a route of the target delivery resource according to the passing time of each road specifically includes: replacing the passing time of each road in the road network data by using the acquired distribution resource passing time of each road; the road network data is output by analyzing open source map OSM data according to an open source map service framework OSRM; and determining the travel path of the target delivery resource according to the replaced road network data. The method has the advantages that the OSM is favorable for obtaining free source-opening and editable map services, the transit time of each road in the road network data is replaced by the transit time of the distribution resource of each road, and the transit time in the road network is favorable for adapting the transit time of the distribution resource, so that the journey path adapted to the target distribution resource can be determined according to the replaced road network data, and the reliability and pertinence of the journey path for determining the target distribution resource are improved.
In addition, the determining a route path of the target delivery resource according to the replaced road network data specifically includes: acquiring the shortest path between the initial position and the target position of the target distribution resource according to the replaced road network data; wherein the shortest path is a path with shortest transit time required for reaching the destination position from the starting position; and taking the shortest path as the determined travel path of the target distribution resource. The transit time in the road network data after replacement is the transit time for adapting the distribution resources, so that the shortest transit time required between the starting position and the target position of the target distribution resources can be acquired conveniently according to the road network data after replacement. The distribution resource passing time on each road is considered, the most reasonable travel route is determined for the target distribution resources, and therefore the distribution efficiency can be improved.
Additionally, the historical data of the dispatch resources includes: delivery characteristics of different delivery resources; the predicting the travel time of the target delivery resource according to the travel path and the prediction model for predicting the travel time specifically comprises: and taking the travel path and the distribution characteristics of the target distribution resources as the input of the pre-estimation model, and outputting the pre-estimated travel time matched with the distribution characteristics of the target distribution resources. The delivery characteristics of the target delivery resources are considered, personalized estimation of the travel time of the delivery resources is facilitated, and personalized requirements of the delivery resources with different delivery characteristics are met.
In addition, the pre-estimated model is specifically a model trained by utilizing XGboost offline. The XGboost frame training is beneficial to obtaining an efficient prediction model for predicting the travel time, and the accuracy of prediction is improved.
Drawings
Fig. 1 is a flowchart of a method of estimating a travel time according to a first embodiment of the present invention;
FIG. 2 is a flowchart of an implementation process of step S102 according to a first embodiment of the present invention;
FIG. 3 is a flowchart of an implementation procedure of step S102 according to a second embodiment of the present invention;
FIG. 4 is a schematic structural diagram of an estimation apparatus of travel time according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it will be appreciated by those of ordinary skill in the art that numerous technical details are set forth in order to provide a better understanding of the present application in various embodiments of the present invention. However, the technical solution claimed in the present application can be implemented without these technical details and various changes and modifications based on the following embodiments. The following embodiments are divided for convenience of description, and should not constitute any limitation to the specific implementation manner of the present invention, and the embodiments may be mutually incorporated and referred to without contradiction.
The first embodiment of the invention relates to a method for predicting travel time, which can be applied to the prediction of travel time required for completing delivery when delivery resources are delivered. The delivery resources may be understood as a rider on the takeaway platform, a courier on the express platform, and the like, and for convenience of description, the delivery resources in this embodiment are exemplified by the rider on the takeaway platform, but not limited to this in practical application.
Fig. 1 shows a method for estimating a travel time in the present embodiment, which includes:
step S101, the passing time of each road on the road network is obtained.
Specifically, the road network can be understood as: the urban area is a network structure organized by roads with different functions, grades and locations in a certain density and in a proper form, and the transit time of each road can be understood as the average transit time on each road. Since the lengths of different roads are generally determined, the passing time of each road is obtained, which is equivalent to obtaining the passing speed of each road. In this embodiment, the obtaining of the passing time of each road may specifically be: the method comprises the steps of firstly obtaining distribution resource track data on each road on a road network, and then obtaining the passing time of distribution resources of each road on the road network according to the obtained distribution resource track data.
The delivery resources can be understood as resources used for completing delivery, the delivery can comprise takeout delivery, express delivery and the like, and the corresponding delivery resources can comprise riders completing the takeout delivery on the takeout platform, couriers completing the express delivery on the express platform and the like. The distribution resources in the present embodiment are taken as an example of a rider, but the present invention is not limited thereto in practical applications. Specifically, the takeaway platform may acquire rider trajectory data on each road, and acquire rider passage time on each road according to the rider trajectory data. The rider track data can be reported by a mobile phone of the rider, and a positioning device can also be arranged on the battery car driven by the rider and reported by the positioning device on the battery car. The skilled person can select to obtain the rider track data on each road in real time or periodically according to actual needs.
Further, the rider trajectory data may be the roads the rider travels during the delivery of the order, and the transit time the rider spends riding on each road. If there are a plurality of riders on the same road, the passing times of the plurality of riders may be averaged, and the average passing time obtained may be used as the passing time of the rider on the same road.
In addition, the rider passing time on the same road may not be the same under different conditions, such as the distribution time period and the weather condition during distribution may affect the rider passing time on the same road. Therefore, when rider track data on each road on a road network is obtained, current environment characteristics can be obtained firstly, the current environment characteristics can comprise weather characteristics, time characteristics and the like, the obtained rider tracks can be influenced by the difference of the weather characteristics and the time characteristics, for example, the weather characteristics can be rain, snow, wind power, temperature and the like, the time characteristics can be a peak-noon time period with high distribution pressure, a peak-early-late time period with traffic congestion and the like, the rider track data under different scenes can be obtained through subdivision according to the environment characteristics, the passing time under different scenes on each road can be obtained according to the rider track data under different scenes, and the obtained passing time can be more accurately matched with the current environment.
And step S102, determining a travel route of the target distribution resource according to the passing time of each road.
Specifically, the target delivery resource may be a delivery resource to be estimated, and the delivery resource to be estimated in this embodiment is the rider to be estimated. Since the acquired transit times of the respective roads are different, the route of the rider to be estimated can be determined based on the transit times of the roads that the delivery start place and the delivery destination of the rider need to pass through. For example, there are a plurality of roads that can reach the distribution location, and the transit times of different roads to the distribution location may be different, so that one of the plurality of roads that can reach the distribution location may be selected as the determined route of the rider to be predicted.
In one example, the implementation of step S102 can be as shown in FIG. 2, including
Step S201, analyzing the open source map OSM data according to the open source map service framework OSRM, and outputting the road network data.
Specifically, the open source map OSM data may be understood as basic geographic data, such as road names, road levels, and specific longitude and latitude information of locations in different areas. The open source map service framework OSRM can analyze the OSM data to obtain road network data. The road network data may include the resolved passing distance, passing time, passing speed, etc. of each road, and the resolved passing time is generally the walking passing time, bicycle passing time or the passing time of driving a car on the road. For example, the parsed road network data may be as shown in table 1 below:
TABLE 1
| Road 1 | Road 2 | Road 3 | Road 4 | Road 5 | |
| Road length | 2 kilometers | 3 kilometers | 4 kilometers | 5 kilometers | 3.5 km |
| Walk through time | 30 minutes | 45 minutes | 60 minutes | 75 minutes | 50 minutes |
| Passing time of bicycle | 15 minutes | 25 minutes | 35 minutes | 45 minutes | 30 minutes |
| Time of passage for driving a vehicle | 8 minutes | 10 minutes | 17 minutes | 20 minutes | 15 minutes |
It should be noted that the road network data in table 1 are only examples of road length, walking passing time, bicycle passing time, and vehicle driving passing time, and are not limited to these in practical applications. The road 1 to the road 5 are also simply named for conveniently distinguishing different roads, and in practical application, the names of the road 1 to the road 5 may be real and actually existing road names. The road network includes a large number of roads, and in the present embodiment, only 5 roads are taken as an example, but the present invention is not limited to this.
Step S202, the acquired distribution resource passing time of each road is used for replacing the passing time of each road in the road network data.
That is, the passing time of each road in the road network data is replaced by the rider passing time acquired in step 101, so that the passing time in the road network data after replacement is more adaptive to the characteristics of the rider-powered electric bicycle.
Specifically, the road network data analyzed according to the open source map service framework OSRM is fixed and not updated in real time, and the influences of road conditions, time intervals and the like are not considered, for example, the passing time of the same road on a heavy rainy day and the passing time of a fine rainy day may be different, the passing time of a peak time interval is inevitably different from the passing time of a non-peak time interval, but the passing time of the same road in the analyzed road network data is fixed and not changed. In addition, the present embodiment is mainly intended to estimate the travel time of a rider, but the travel time in the road network data is not suitable for the rider riding the battery car, although the walking travel time, the bicycle travel time, or the vehicle driving travel time in the road network data can represent the road traffic state to some extent. Taking the bicycle passing time most similar to that of the battery car as an example, when the bicycle passes through a certain road, the action of the platform bridge on the cart may exist, but the battery car is rarely ridden by a rider, so that the passing time of the battery car on the same road and the passing time of the bicycle have a large difference. Therefore, in order to make the estimated travel time more suitable for the rider riding the battery car, the passing time on each road in the road network data is replaced with the passing time of the rider obtained in step 101. The road network data after replacement can be shown in the following table 2:
TABLE 2
| Road 1 | Road 2 | Road 3 | Road 4 | Road 5 | |
| Road length | 2 kilometers | 3 kilometers | 4 kilometers | 5 kilometers | 3.5 km |
| Rider passing time | 10 minutes | 20 minutes | 30 minutes | 45 minutes | 25 minutes |
It should be noted that, since the lengths of the respective roads and the rider passage times have been listed in table 2, the rider passage speed on each road can be calculated from the lengths and times.
Step S203, determining the travel route of the target delivery resource according to the road network data after replacement.
Specifically, the road network data after replacement may include the passing distance of each road, i.e., the road length, the rider passing time, the rider passing speed, and the like. And inquiring roads which can reach the delivery address in the road network data after replacement according to the delivery address of the rider to be estimated.
In the prior art, the open source map service framework OSRM may include a graph storage structure constructed according to road network data, and the graph storage structure constructed according to the road network data before replacement may be a point-to-point path constructed according to the length of each road, for example, different paths measured by distance length from a starting point to an end point may be queried in the graph storage structure according to input position coordinates of the starting point and the end point. In the present embodiment, the actual road conditions, for example, the degree of congestion of the road, the traffic speed, the influence of weather, the influence of time zone, and the like are taken into consideration, and even the shortest route is not necessarily the route suitable for the rider. Therefore, in consideration of the fact that the demand of the user who places an order, whether the user is a rider or a merchant, for the time length required for delivery is high, in order to control the time length required for delivery to a great extent within an achievable range, in the embodiment, the graph storage structure in the active map service framework OSRM is a graph storage structure constructed according to the replaced road network data, that is, a point-to-point inter-road path constructed according to the rider passing time of each road is stored in the graph storage structure, and according to the input position coordinates of the delivery start point and the delivery end point, the inter-road path between the delivery start point and the delivery end point is obtained by querying in the graph storage structure, and different paths measured by the length of the rider passing time are adopted between the delivery start point and the delivery end point. The inquired routes between the delivery starting point and the delivery ending point can be multiple, and each inquired route can indicate the rider passing time of the route, so that one route can be selected as the determined route of the rider to be estimated according to the rider passing times indicated by the multiple routes. The route of the rider to be estimated selected in the mode takes the rider passing time, the rider passing speed and the passing distance of each road in the road network into consideration, so that the route of the rider to be estimated is favorable for determining a route suitable for the rider to be estimated.
And step S103, predicting the travel time of the target delivery resources according to the travel path and the prediction model for predicting the travel time.
Specifically, the estimation model for estimating the travel path can be obtained by training according to the collected historical data of the distribution resources in advance. Since the delivery resource in this embodiment is taken as an example of a rider, the estimation model can be obtained by training in advance according to the collected historical data of the rider. A large amount of historical data of riders serves as a training set of the estimation model, each training sample in the training set can comprise relevant data of the riders in the process of completing one delivery, and the relevant data can comprise a path from a delivery starting place to a delivery destination, time spent, weather conditions, a time period of the delivery process, delivery characteristics of the riders and the like. The time period of the distribution process can be a certain time period of a working day or a certain time period of a non-working day. The delivery characteristics of the rider individual may be a waybill capacity value, a time-sliced order pickup rate, a rejection rate, a timeout rate, etc. Taking the unit time as 15 minutes as an example, the time of day can be divided into several time slices, 00: 00 to 00: 15 is the first time slice; 00: 15 to 00: 30 is the second time slice, and so on. That is, at 00: 15 this time, the first time slice is numbered 1, and in the interval 00: at this point 30, a second time slice is obtained, which is numbered 2, and so on. The order receiving rate, the order rejection rate and the timeout rate of a rider in different time slices can be obtained through analysis according to historical distribution data of the rider. Furthermore, XGboost can be adopted to perform off-line training on a training set containing a large number of training samples, so as to train an estimation model for estimating travel time. The XGboost can increase the robustness of a prediction model, and compared with a traditional machine learning algorithm, the XGboost has the advantages of high speed, good effect, capability of processing large-scale data and support of multiple languages and user-defined loss functions.
It is worth mentioning that when the trained estimation model is used for estimating the travel time, the travel time output by the estimation model and the travel time spent in the actual distribution process can be compared to verify the accuracy of the estimated travel time of the estimation model, if the estimated travel time is compared with the actual travel time, the inaccurate times exceed the preset times, the estimation model can be adaptively adjusted, and the preset times can be adjusted by a person skilled in the art according to actual needs. By continuously optimizing the pre-estimated model, the accuracy of the pre-estimated model can continuously approach to the true value.
Of course, in practical application, the estimation model may also be updated at intervals, specifically, a training set used in training the estimation model may be updated in time, and training samples in the training set are historical data of the rider close to the current time. For example, the estimation model is updated once every month, the training set adopted for updating the estimation model at the beginning of 5 months is the rider historical data collected in 4 months, and the training set adopted for updating the estimation model at the beginning of 6 months is the rider historical data collected in 5 months.
Furthermore, after the estimation model is trained, the estimation model can be used for carrying out personalized estimation on the travel time of the rider with different delivery characteristics. Specifically, the determined travel path of the rider to be estimated and the delivery characteristics of the rider to be estimated can be used as input of the estimation model, and the travel time matched with the delivery characteristics of the rider to be estimated is output. In practical application, the collected historical data of the rider can also comprise images of different riders, the images of the riders correspond to the distribution characteristics of the riders, and the distribution characteristics can be as follows: the rider's back order ability value, order taking rate, order rejection rate, timeout rate, the rider's age, work experience, general delivery rate, etc. The image and the travel path of the rider to be estimated can be used as input of a trained estimation model, and therefore the travel time matched with the delivery characteristics of the rider to be estimated is output.
Taking a take-out delivery scene as an example, the target delivery resources are a rider a and a rider B, and it is assumed that during a period from 12 pm to 1 pm, the order taking capability value of the rider a is greater than that of the rider B, the order taking rate of the rider a is 80%, the order rejecting rate is 20%, the timeout rate is 20%, the order taking rate of the rider B is 50%, the order rejecting rate is 50%, and the timeout rate is 40%, that is, the delivery characteristics of the rider a and the rider B are different. Both rider a and rider B have orders to be delivered at 12 am, and the delivery start and target positions are the same. Therefore, in estimating the travel times of both persons, the travel paths of the rider a and the rider B determined from the rider passage times of the respective roads on the 12-point time road network are the same. Inputting the determined rider a travel path and delivery characteristics into the predictive model will output a travel time that matches the delivery characteristics of rider a, say 20 minutes for rider a. Inputting the determined travel path and delivery characteristics of rider B into the predictive model will output a travel time that matches the delivery characteristics of rider B, say 30 minutes for rider B. That is, although the determined travel paths are the same for rider a and rider B, since the rider a has a higher waybill capability value and a higher pick-up rate than rider B, and has a lower reject rate and a lower timeout rate than rider B, the estimated travel time for rider a may be relatively smaller for the same travel path as compared to rider B. In the example, the estimation of personalized travel time can be completed for the rider A and the rider B with different delivery characteristics, and the delivery characteristics of different riders are considered, so that the actual requirements of different riders can be met.
Compared with the prior art, the method and the device have the advantages that the travel route of the target distribution resources is determined through the acquired passing time of each road on the road network, and the travel route suitable for the target distribution resources in each road is determined in consideration of the passing time of each road. According to the determined travel path and the estimation model for estimating the travel time, the travel time of the target delivery resources can be accurately estimated. Moreover, the prediction model is obtained by training according to the collected historical data of the distribution resources in advance, namely the data for training the prediction model is derived from the real historical data of the distribution resources, so that the reference value is high, and the prediction result is more accurate and reliable. Meanwhile, the embodiment of the invention does not depend on crawler data, avoids the problem of low speed of estimated travel time caused by long crawler crawling time, and improves the estimated speed to a certain extent.
The second embodiment of the present invention relates to a method for estimating travel time, and this embodiment provides a specific implementation manner for determining a travel route of a target delivery resource according to the transit time of each road, that is, this embodiment mainly introduces the implementation procedure of step S102 in the first embodiment. As shown in fig. 3, the implementation process of step S102 in this embodiment may include the following steps:
step S301, analyzing the open source map OSM data according to the open source map service framework OSRM, and outputting the road network data.
Step S302, the acquired distribution resource passing time of each road is used for replacing the passing time of each road in the road network data.
Step S301 and step S102 are substantially the same as step S201 and step S202 in the first embodiment, and are not repeated herein to avoid repetition.
Step S303, according to the road network data after replacement, obtaining the shortest path between the starting position and the destination position of the target distribution resource.
Specifically, the shortest route may be a route that requires the shortest transit time to reach the destination location from the start location. In this embodiment, the delivery resources are still taken as an example of a rider, and when the rider prepares to start delivering the order, the position of the rider is taken as the initial position, and the delivery address in the order is taken as the destination position. The map storage structure in the open source map service framework OSRM stores point-to-point routes constructed by using the rider passing time of each road, and the routes from the starting position to the destination position can be inquired in the map storage structure according to the starting position and the destination position. In practical application, a plurality of alternative paths may exist between the initial position and the destination position, and one path with the shortest transit time needs to be selected from the plurality of alternative paths. Since the road network data includes data such as the rider passing time, the rider passing speed, the road distance and the like of each road, in order to select the route with the shortest rider passing time, the rider passing times of the roads included in each optional route can be added to obtain the passing time of the rider from the starting position to the destination position on each route. For example, one alternative route a is composed of a road 1, a road 2 and a road 3, the rider passing time on the road 1, the road 2 and the road 3 in the road network data after the replacement is respectively 10 minutes, 8 minutes and 12 minutes, the rider passing time of the route a is 10+8+12 to 30 minutes, the rider passing time of other alternative routes can be calculated according to the method, and finally the rider passing time of all the alternative routes is compared, and the route with the shortest rider passing time is selected as the route to be estimated of the rider determined according to the road network data after the replacement.
Further, due to the difference in traffic conditions on different roads, in the alternative routes from the start position to the destination position, the required rider passing time may be different even if the distances are the same for different routes, and likewise, the required rider passing time may be the same for different routes at different distances. And if more than one path with the shortest passing time of the rider exists, selecting one path with the shortest passing distance from the paths with the shortest passing time as the finally determined travel path of the rider to be predicted.
Also taking the take-out delivery scenario as an example, the target delivery resource is a rider nail, and the paths that the rider nail can select from the starting position to the destination position and the rider passing time and the passing distance of each path can be shown in the following table 3:
TABLE 3
| Road 1 | Road 2 | Road 3 | Time of each route rider passing | |
| Path A | 10 minutes | 9 minutes | 8 minutes | 27 minutes |
| Path B | 8 minutes | 10 minutes | 10 minutes | 28 minutes |
| Route C | 7 minutes | 10 minutes | 15 minutes | 32 minutes |
| Route D | 12 minutes | 5 minutes | 12 minutes | 29 minutes |
Wherein, the alternative paths include four paths which are respectively a path A, a path B, a path C and a path D, each path needs to pass through three roads, the three roads in each path are not completely the same, the rider passing time of each of the path A, the path B, the path C and the path D can be obtained in the road network data after replacement, the rider passing time of each road is added to obtain the rider passing time of each path, the rider passing time of the path A, the path B, the path C and the path D can be respectively 27 minutes, 28 minutes, 32 minutes and 29 minutes as can be seen from the table 1, the path A with the shortest rider passing time is the path A, therefore, the path A can be determined as the travel path of the rider A, the distribution characteristics of the path A and the rider A are taken as the input of the estimation model, the travel estimation time suitable for the rider A is output, and finishing the personalized estimation of the travel time of the rider armor.
In another example, the alternative paths for the rider nail to reach the destination position from the start position and the rider passing time and distance for each path may be as shown in table 4 below:
TABLE 4
| Road 1 | Road 2 | Road 3 | Time of each route rider passing | |
| Path A (2 kilometers) | 10 minutes | 9 minutes | 10 minutes | 29 minutes |
| Path B (3 kilometers) | 8 minutes | 10 minutes | 12 minutes | 30 minutes |
| Path C (4 kilometers) | 7 minutes | 10 minutes | 15 minutes | 32 minutes |
| Path D (2.5 km) | 12 minutes | 5 minutes | 12 minutes | 29 minutes |
As can be seen from table 2, the rider passing times of the route a and the route D are both the smallest among the finally determined rider passing times of the respective routes, and therefore the passing distances of the route a and the route D can be considered, and the passing distances of the route a and the route D can be obtained by summing the passing distances of the respective roads constituting the route a and the route D. As can be seen from table 2, the passing distances of the path a and the path D are 2 km and 2.5 km, respectively, and therefore the path a can be taken as the finally determined route of the rider's nail to be estimated.
Compared with the prior art, in the embodiment, the transit time in the road network data after replacement is the transit time adapted to the distribution resources, so that the path with the shortest transit time between the starting position and the target position of the target distribution resources can be acquired favorably according to the road network data after replacement. The distribution resource passing time on each road is considered, the most reasonable travel route is determined for the target distribution resources, and therefore the distribution efficiency can be improved.
A third embodiment of the present invention relates to a travel time estimation device, as shown in fig. 4, including:
an obtaining module 401, configured to obtain a passing time of each road on a road network; a determining module 402, configured to determine a route of the target distribution resource according to the passing time of each road; the estimation module 403 estimates the travel time of the target delivery resource according to the travel path and an estimation model for estimating the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
In an example, the obtaining module 401 may be specifically configured to obtain distribution resource trajectory data on each road on the road network; and acquiring the passing time of the distribution resources of each road on the road network according to the acquired distribution resource track data.
In an example, the determining module 402 may be specifically configured to analyze open source map OSM data according to an open source map service framework OSRM, and output road network data; wherein the road network data comprises the passing time of each road; replacing the passing time of each road in the road network data by the acquired distribution resource passing time of each road; and determining the travel path of the target delivery resource according to the replaced road network data.
In an example, the determining module 402 may be specifically configured to obtain a shortest path between a starting position and a destination position of the target distribution resource according to the road network data after replacement; wherein the shortest path is a path with shortest transit time required for reaching the destination position from the starting position; and taking the shortest path as the determined travel path of the target distribution resource.
In one example, the historical data of the dispatched resources may include: delivery characteristics of different delivery resources; the predicting the travel time of the target delivery resource according to the travel path and the prediction model for predicting the travel time specifically comprises: and taking the travel path and the distribution characteristics of the target distribution resources as the input of the pre-estimation model, and outputting the pre-estimated travel time matched with the distribution characteristics of the target distribution resources. Wherein, the distribution characteristics comprise any one or combination of the following: the age, sex, amount of bills of lading, and work experience of the distribution resources.
In one example, the predictive model may be embodied as a model trained offline using XGBoost.
Compared with the prior art, the method and the device have the advantages that the passing time of each road on the road network acquired by the acquisition module is used, the determining module determines the travel route of the target distribution resources, and the passing time of each road is considered, so that the travel route suitable for the target distribution resources can be determined in each road. The estimation module can accurately estimate the travel time of the target delivery resources according to the determined travel path and the estimation model for estimating the travel time. Moreover, the prediction model is obtained by training according to the collected historical data of the distribution resources in advance, namely the data for training the prediction model is derived from the real historical data of the distribution resources, so that the reference value is high, and the prediction result is more accurate and reliable. Meanwhile, the embodiment of the invention does not depend on crawler data, avoids the problem of low speed of estimated travel time caused by long crawler crawling time, and improves the estimated speed to a certain extent.
A fourth embodiment of the present invention relates to an electronic apparatus, as shown in fig. 5, including: at least one processor 501; and a memory 502 communicatively coupled to the at least one processor 501; and a communication component 503 in communicative connection with the scanning device, the communication component 503 receiving and transmitting data under the control of the processor 501; wherein the memory 502 stores instructions executable by the at least one processor 501, the instructions being executable by the at least one processor 501 to implement:
acquiring the passing time of each road on a road network; determining a travel route of target distribution resources according to the passing time of each road; predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
Specifically, the terminal device includes: one or more processors 501 and a memory 502, with one processor 501 being an example in fig. 5. The processor 501 and the memory 502 may be connected by a bus or other means, and fig. 5 illustrates the connection by the bus as an example. Memory 502, which is a non-volatile computer-readable storage medium, may be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The processor 501 executes various functional applications of the device and data processing, i.e., realizes the above-described method of estimating the travel time, by running the nonvolatile software programs, instructions, and modules stored in the memory 502.
The memory 502 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store a list of options, etc. Further, the memory 502 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, memory 502 may optionally include memory 502 located remotely from processor 501, and such remote memory 502 may be connected to an external device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
One or more modules are stored in memory 502 and, when executed by the one or more processors 501, perform the method of travel time estimation in any of the method embodiments described above.
The product can execute the method provided by the embodiment of the application, has corresponding functional modules and beneficial effects of the execution method, and can refer to the method provided by the embodiment of the application without detailed technical details in the embodiment.
A fifth embodiment of the invention relates to a non-volatile storage medium for storing a computer-readable program for causing a computer to perform some or all of the above method embodiments.
That is, as can be understood by those skilled in the art, all or part of the steps in the method for implementing the embodiments described above may be implemented by a program instructing related hardware, where the program is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the method described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It will be understood by those of ordinary skill in the art that the foregoing embodiments are specific examples for carrying out the invention, and that various changes in form and details may be made therein without departing from the spirit and scope of the invention in practice.
The embodiment of the application discloses A1. a method for estimating travel time, which comprises the following steps:
acquiring the passing time of each road on a road network;
determining a travel route of target distribution resources according to the passing time of each road;
predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
A2. According to the method for estimating the travel time described in a1, the obtaining the passing time of each road on the road network specifically includes:
acquiring distribution resource track data on each road on the road network;
and acquiring the passing time of the distribution resources on each road on the road network according to the acquired distribution resource track data.
A3. According to the method for estimating the travel time described in a2, the determining the travel route of the target distribution resource according to the transit time of each road specifically includes:
replacing the passing time of each road in the road network data by using the obtained passing time of each road on the road network of the distribution resources; the road network data is output by analyzing open source map OSM data according to an open source map service framework OSRM;
and determining the travel path of the target distribution resource according to the replaced road network data.
A4. According to the method for estimating the travel time described in a3, the determining the travel path of the target delivery resource according to the road network data after replacement specifically includes:
acquiring the shortest path between the initial position and the target position of the target distribution resource according to the replaced road network data; wherein the shortest path is a path with shortest transit time required for reaching the destination position from the starting position;
and taking the shortest path as a travel path of the target distribution resource.
A5. According to the method for estimating travel time described in a2, the acquiring of the distribution resource trajectory data on each road on the road network specifically includes:
acquiring current environmental characteristics;
and acquiring the distribution resource track data on each road on the road network according to the current environmental characteristics.
A6. According to the method for estimating the travel time of a1, the historical data of the distribution resources includes: delivery characteristics of different delivery resources;
the predicting the travel time of the target delivery resource according to the travel path and the prediction model for predicting the travel time specifically comprises:
and taking the travel path and the distribution characteristics of the target distribution resources as the input of the pre-estimation model, and outputting the pre-estimated travel time matched with the distribution characteristics of the target distribution resources.
A7. The method for estimating the travel time according to a6, wherein the distribution characteristics include any one or a combination of the following:
the back order capability value, the order receiving rate, the order rejecting rate and the timeout rate of the distributed resources.
A8. According to the travel time estimation method described in a1, the estimation model is specifically a model trained offline by using XGBoost.
The embodiment of the present application further discloses b1. a device for estimating travel time, including:
the acquisition module is used for acquiring the passing time of each road on the road network;
the determining module is used for determining a travel route of the target distribution resources according to the passing time of each road;
the estimation module estimates the travel time of the target delivery resources according to the travel path and an estimation model used for estimating the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
The embodiment of the present application further discloses c1. an electronic device, including: a memory storing a computer program and a processor executing the program to perform:
acquiring the passing time of each road on a road network;
determining a travel route of target distribution resources according to the passing time of each road;
predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
C2. According to the electronic device of C1, the acquiring the transit time of each road on the road network specifically includes:
acquiring distribution resource track data on each road on the road network;
and acquiring the passing time of the distribution resources of each road on the road network according to the acquired distribution resource track data.
C3. According to the electronic device of C2, determining the travel route of the target distribution resource according to the transit time of each road specifically includes:
replacing the passing time of each road in the road network data by using the acquired distribution resource passing time of each road; the road network data is output by analyzing open source map OSM data according to an open source map service framework OSRM;
and determining the travel path of the target delivery resource according to the replaced road network data.
C4. According to the electronic device of C3, determining a travel path of the target delivery resource according to the replaced road network data specifically includes:
acquiring the shortest path between the initial position and the target position of the target distribution resource according to the replaced road network data; wherein the shortest path is a path with shortest transit time required for reaching the destination position from the starting position;
and taking the shortest path as the determined travel path of the target distribution resource.
C5. According to the electronic device of C2, the acquiring the distribution resource trajectory data on each road on the road network specifically includes:
acquiring current environmental characteristics;
and acquiring the distribution resource track data on each road on the road network according to the current environmental characteristics.
C6. The electronic device of C1, the historical data of the dispatch resource comprising: delivery characteristics of different delivery resources;
the predicting the travel time of the target delivery resource according to the travel path and the prediction model for predicting the travel time specifically comprises:
and taking the travel path and the distribution characteristics of the target distribution resources as the input of the pre-estimation model, and outputting the pre-estimated travel time matched with the distribution characteristics of the target distribution resources.
C7. The electronic device of C6, wherein the delivery characteristics include any one or a combination of:
the back order capability value, the order receiving rate, the order rejecting rate and the timeout rate of the distributed resources.
C8. According to the electronic device of C1, the pre-estimated model is specifically a model trained offline by using XGBoost.
A non-volatile storage medium storing a computer-readable program for causing a computer to perform the method of estimating a travel time according to any one of a1 to A8 is also disclosed in an embodiment of the present application.
Claims (10)
1. A method for estimating travel time is characterized by comprising the following steps:
acquiring the passing time of each road on a road network;
determining a travel route of target distribution resources according to the passing time of each road;
predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
2. The method for estimating travel time according to claim 1, wherein the obtaining of the travel time of each road in the road network specifically includes:
acquiring distribution resource track data on each road on the road network;
and acquiring the passing time of the distribution resources on each road on the road network according to the acquired distribution resource track data.
3. The method for estimating travel time according to claim 2, wherein the determining a travel route of a target delivery resource according to the transit time of each road specifically comprises:
replacing the passing time of each road in the road network data by using the obtained passing time of each road on the road network of the distribution resources; the road network data is output by analyzing open source map OSM data according to an open source map service framework OSRM;
and determining the travel path of the target distribution resource according to the replaced road network data.
4. The method for estimating travel time according to claim 3, wherein the determining the travel path of the target delivery resource according to the road network data after replacement specifically comprises:
acquiring the shortest path between the initial position and the target position of the target distribution resource according to the replaced road network data; wherein the shortest path is a path with shortest transit time required for reaching the destination position from the starting position;
and taking the shortest path as a travel path of the target distribution resource.
5. The method for estimating travel time according to claim 2, wherein the acquiring of the distribution resource trajectory data on each road on the road network specifically includes:
acquiring current environmental characteristics;
and acquiring the distribution resource track data on each road on the road network according to the current environmental characteristics.
6. The method of claim 1, wherein the historical data of the delivery resources comprises: delivery characteristics of different delivery resources;
the predicting the travel time of the target delivery resource according to the travel path and the prediction model for predicting the travel time specifically comprises:
and taking the travel path and the distribution characteristics of the target distribution resources as the input of the pre-estimation model, and outputting the pre-estimated travel time matched with the distribution characteristics of the target distribution resources.
7. The method of claim 6, wherein the delivery characteristics comprise any one or a combination of the following:
the back order capability value, the order receiving rate, the order rejecting rate and the timeout rate of the distributed resources.
8. An apparatus for estimating travel time, comprising:
the acquisition module is used for acquiring the passing time of each road on the road network;
the determining module is used for determining a travel route of the target distribution resources according to the passing time of each road;
the estimation module estimates the travel time of the target delivery resources according to the travel path and an estimation model used for estimating the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
9. An electronic device, comprising: a memory storing a computer program and a processor executing the program to perform:
acquiring the passing time of each road on a road network;
determining a travel route of target distribution resources according to the passing time of each road;
predicting the travel time of the target delivery resources according to the travel path and a prediction model for predicting the travel time; and the pre-estimation model is obtained by training according to the collected historical data of the distribution resources in advance.
10. A non-volatile storage medium storing a computer-readable program for causing a computer to execute the method of estimating a travel time according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910117958.1A CN111582527A (en) | 2019-02-15 | 2019-02-15 | Travel time estimation method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910117958.1A CN111582527A (en) | 2019-02-15 | 2019-02-15 | Travel time estimation method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111582527A true CN111582527A (en) | 2020-08-25 |
Family
ID=72112580
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910117958.1A Pending CN111582527A (en) | 2019-02-15 | 2019-02-15 | Travel time estimation method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111582527A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112907191A (en) * | 2021-03-23 | 2021-06-04 | 拉扎斯网络科技（上海）有限公司 | Distribution time generation method and device, computer equipment and readable storage medium |
| CN112990813A (en) * | 2021-02-09 | 2021-06-18 | 北京橙心无限科技发展有限公司 | Method and device for determining navigation time |
| CN117635001A (en) * | 2023-11-30 | 2024-03-01 | 阿帕数字科技有限公司 | Distribution method based on time window constraint in supply chain transportation scene |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107945507A (en) * | 2016-10-13 | 2018-04-20 | 腾讯科技（深圳）有限公司 | Travel Time Estimation Method and device |
| US20180350237A1 (en) * | 2016-10-08 | 2018-12-06 | Dalian University Of Technology | Method for estimating distribution of urban road travel time in considering operation state of taxi |
| CN108985506A (en) * | 2018-07-03 | 2018-12-11 | 蔚来汽车有限公司 | Planning driving path recommended method, prediction technique, acquisition methods and its device |
| CN109284946A (en) * | 2017-07-20 | 2019-01-29 | 阿里巴巴集团控股有限公司 | A kind of stroke distances, logistics service journey time determine methods, devices and systems |
-
2019
- 2019-02-15 CN CN201910117958.1A patent/CN111582527A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180350237A1 (en) * | 2016-10-08 | 2018-12-06 | Dalian University Of Technology | Method for estimating distribution of urban road travel time in considering operation state of taxi |
| CN107945507A (en) * | 2016-10-13 | 2018-04-20 | 腾讯科技（深圳）有限公司 | Travel Time Estimation Method and device |
| CN109284946A (en) * | 2017-07-20 | 2019-01-29 | 阿里巴巴集团控股有限公司 | A kind of stroke distances, logistics service journey time determine methods, devices and systems |
| CN108985506A (en) * | 2018-07-03 | 2018-12-11 | 蔚来汽车有限公司 | Planning driving path recommended method, prediction technique, acquisition methods and its device |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990813A (en) * | 2021-02-09 | 2021-06-18 | 北京橙心无限科技发展有限公司 | Method and device for determining navigation time |
| CN112907191A (en) * | 2021-03-23 | 2021-06-04 | 拉扎斯网络科技（上海）有限公司 | Distribution time generation method and device, computer equipment and readable storage medium |
| CN117635001A (en) * | 2023-11-30 | 2024-03-01 | 阿帕数字科技有限公司 | Distribution method based on time window constraint in supply chain transportation scene |
| CN117635001B (en) * | 2023-11-30 | 2024-04-30 | 阿帕数字科技有限公司 | Distribution method based on time window constraint in supply chain transportation scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111966729B (en) | Vehicle track data processing method, device, equipment and storage medium | |
| US10160321B2 (en) | Methods, circuits, devices, systems and associated computer executable code for driver decision support | |
| Zhang et al. | Models, algorithms, and evaluation for autonomous mobility-on-demand systems | |
| CN102147263B (en) | Apparatus and method for planning trips combination | |
| US11386362B1 (en) | System and method for optimizing waste / recycling collection and delivery routes for service vehicles | |
| US20170046653A1 (en) | Planning of transportation requests | |
| CN104121918A (en) | Real-time path planning method and system | |
| CN110118567B (en) | Travel mode recommendation method and device | |
| US9467806B2 (en) | Improvements related to tracking moving objects | |
| CN103177562B (en) | A kind of method and device obtaining information of traffic condition prediction | |
| CN112380448B (en) | Vehicle data processing method and device, computer equipment and storage medium | |
| CN105674994A (en) | Driving route acquisition method and device and navigation equipment | |
| CN111582527A (en) | Travel time estimation method and device, electronic equipment and storage medium | |
| CN110852499A (en) | Method and device for recovering fault vehicle | |
| US20180045527A1 (en) | Systems and Methods for Predicting Vehicle Fuel Consumption | |
| CN111260172B (en) | Information processing method and system and computer equipment | |
| CN109141455A (en) | Navigation path planning method and server | |
| CN114005295A (en) | Method, device, equipment and medium for predicting vehicle energy consumption information | |
| CN103942312A (en) | Public transportation transfer line planning method and device | |
| CN115708343A (en) | Method for collecting data from a set of vehicles | |
| CN109754606B (en) | Method for predicting road congestion condition based on taxi positioning | |
| CN113379159B (en) | Taxi driver passenger searching route recommendation method based on gray model and Markov decision process | |
| WO2020263176A1 (en) | Processing route information | |
| CN113537752A (en) | Traffic transportation big data scheduling method and scheduling platform based on multiple data sources | |
| CN114005294A (en) | Path determination method, device, equipment and medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200825 |