CN112562309B - Network taxi appointment scheduling method based on improved Dijkstra algorithm - Google Patents
Network taxi appointment scheduling method based on improved Dijkstra algorithm Download PDFInfo
- Publication number
- CN112562309B CN112562309B CN202110222684.XA CN202110222684A CN112562309B CN 112562309 B CN112562309 B CN 112562309B CN 202110222684 A CN202110222684 A CN 202110222684A CN 112562309 B CN112562309 B CN 112562309B
- Authority
- CN
- China
- Prior art keywords
- driver
- data
- time
- road
- passenger
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/20—Monitoring the location of vehicles belonging to a group, e.g. fleet of vehicles, countable or determined number of vehicles
- G08G1/202—Dispatching vehicles on the basis of a location, e.g. taxi dispatching
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/20—Monitoring the location of vehicles belonging to a group, e.g. fleet of vehicles, countable or determined number of vehicles
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Navigation (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a network taxi appointment scheduling method based on an improved Dijkstra algorithm, which comprises the steps of firstly, acquiring road network data, POI data, taxi journey GPS data and real-time road condition data in a designated area; secondly, selecting a classic Dijkstra algorithm, adjusting the weight on the basis of the Dijkstra algorithm, and comprehensively considering road network distance influence factors and time influence factors based on real-time road conditions to form a new passenger demand and driver matching strategy; and finally, generating random passenger and driver distribution according to the POI data and the comprehensive data of the taxi journey GPS, setting a dispatching range for each randomly generated passenger, and searching for the driver and the route in the dispatching range by improving the Dijkstra algorithm to obtain an optimized route taking the benefits of the passenger, the driver and the network platform into consideration. The invention effectively improves the utilization rate of drivers, the order receiving rate of platforms and reduces the waiting time of passengers.
Description
Technical Field
The invention relates to the technical field of path distance planning, in particular to a network car booking scheduling method based on an improved Dijkstra algorithm.
Background
Based on the rapid development of mobile internet and mobile equipment, the online booking vehicle is popular as a new industry form, provides an interactive platform for passengers and drivers, can acquire requirements and vehicle information in real time, carries out online scheduling, completes matching between the drivers and the passengers, and extracts part of completed orders as platform income. Therefore, whether a network car booking platform can be continuously developed or not is crucial to the balance of benefits among a driver, passengers and the platform. Which in turn depends on the core scheduling algorithm employed by the platform. In the current market and research, a system dispatching monotonicity mode occupies a mainstream position, but in the dispatching mode, the current dispatching algorithm has two modes of considering the road network distance between drivers and passengers and considering the dead time of a driver, and in the actual situation, the driver closest to the passenger may need more time to receive the passenger than other drivers due to complex factors such as real-time road conditions and the like. On the other hand, a scheduling mode integrating interests of two parties or even three parties is not available, and a network appointment platform is used as a third party for providing services, so that it is necessary to comprehensively consider interests of multiple parties. In addition, one of the problems to be considered by the network booking platform is the balance between supply and demand, and the driver naturally obtains low benefit when being in an area with low traffic density for a long time, and meanwhile, the area with high passenger order density also needs more driver services.
Disclosure of Invention
In view of the above, the invention provides a network taxi appointment scheduling method based on an improved Dijkstra algorithm, which selects a classic Dijkstra algorithm, adjusts the weight on the basis of the Dijkstra algorithm, comprehensively considers road network distance influence factors and time influence factors based on real-time road conditions, forms a new passenger demand and driver matching strategy, greatly improves the utilization rate of network taxi appointment drivers, considers the benefits of a network taxi appointment platform and passengers, and has outstanding economic benefits.
In one aspect, the invention provides a network taxi appointment scheduling method based on an improved Dijkstra algorithm, and a packet
The method comprises the following steps:
s1, acquiring road network data, POI data, taxi journey GPS data and real-time road condition data in the designated area;
s2, based on the Dijkstra algorithm, adding road network distance influence factors and time influence factors based on real-time road conditions into the weight to form an improved Dijkstra algorithm;
s3, generating random passenger and driver distribution according to the POI data and the comprehensive data of the taxi journey GPS, setting a dispatching range for each randomly generated passenger, and searching for the driver and the route in the dispatching range by improving the Dijkstra algorithm to obtain the optimized route taking the benefits of the passenger, the driver and the network platform into consideration.
Further, the road network data is downloaded to a road network osm file through an Open Street Map, and the point set and the edge set in the road network data both adopt an mercator projection coordinate system, and the mercator projection coordinate system is converted into an GCj-02 coordinate system, so that a converted road network intersection point set and a converted road network edge set are obtained.
Further, the post-conversion road network cross point set is composed of an Allid column, a lon column and a lat column, wherein the Allid column represents a node ID in the full map, the lon column represents a node longitude, and the lat column represents a node latitude; the converted road network edge set consists of an ID column, a newStart column, a newEnd column, a length column and a freespeed column, wherein the ID column represents an edge ID in the whole map, the newStart column represents a node Allid of a starting point, the newEnd column represents a node Allid of an end point, the length column represents an edge weight, namely the length of a road section, the unit is meter, and the freespeed column represents the highest speed limit, and the unit is meter per second.
Further, the POI data are obtained from the existing APP map software by using a crawler technology, and each POI data comprises longitude, latitude, namePOI name, address, province, city, business area, big _ type, middle _ type and small _ type; the taxi journey GPS data at least comprises a vehicle id, a device number, a direction angle and GPS longitude and latitude; the method comprises the steps of obtaining a gathering set of crowd distribution and people flow in a designated area by integrating POI data and taxi journey GPS data, dividing a map into grids of 1km by 1km, wherein the larger the data volume in the grids is, the larger the people flow and the population density in the area are.
Furthermore, the real-time road condition data is obtained by utilizing a traffic situation interface provided by the existing APP map software API and using a python crawler technology.
Further, the optimized path considering the benefits of the passenger, the driver and the network platform is specifically a path with high success rate of platform order receiving, short waiting time of the passenger and small dead time and distance of the driver.
Further, the step S3 is embodied as:
searching in the dispatching range by taking a randomly generated passenger as an origin, and calculating all drivers in the dispatching rangeDistance of driver with shortest distance in airplane
Calculating the time of the driver with the shortest estimated arrival time under the current road condition
According to the formula (1), calculating the comprehensive weight of each driver, and according to the comprehensive weight of the drivers, selecting the driver with the maximum comprehensive weight from all drivers to receive the order and dispatch:
in the formula (I), the compound is shown in the specification,
the distance of the ith driver from the origin passenger,
the estimated arrival time of the ith driver at the origin passenger under the current road conditions,
is the comprehensive weight of the ith driver, A is the road network distance influence factor, B is the time influence factor, wherein,
the expression of (a) is as follows:
wherein L ismLength of the mth road in the route for the ith driver to pick up the passenger, SmSpeed limit, Rs, for the mth road in the route for the ith driver to pick up and deliver passengersmFor the current congestion situation in the route of the ith driver to pick up the passenger, n represents the road in the route of the driver to pick up the passengerThe total number of stages.
Furthermore, the longitude and latitude of the lower left corner point and the longitude and latitude of the upper right corner point of each grid can obtain real-time road condition data in the area, and the real-time road condition state of the area is determined according to the serious congestion, slow driving and smooth priority of the road condition of each grid, so that a basis is provided for determining the time influence factor.
The invention provides a network taxi appointment scheduling method based on an improved Dijkstra algorithm, which comprises the steps of firstly, acquiring road network data, POI data, taxi journey GPS data and real-time road condition data in a designated area; secondly, selecting a classic Dijkstra algorithm, adjusting the weight on the basis of the Dijkstra algorithm, and comprehensively considering road network distance influence factors and time influence factors based on real-time road conditions to form a new passenger demand and driver matching strategy; and finally, generating random passenger and driver distribution according to the POI data and the comprehensive data of the taxi journey GPS, setting a dispatching range for each randomly generated passenger, and searching for the driver and the route in the dispatching range by improving the Dijkstra algorithm to obtain an optimized route taking the benefits of the passenger, the driver and the network platform into consideration. The invention effectively improves the utilization rate of drivers, the order receiving rate of platforms and reduces the waiting time of passengers.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate an embodiment of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart of a network taxi appointment scheduling method based on an improved Dijkstra algorithm according to an embodiment of the present invention;
FIG. 2 is a global view of a road network in a city;
FIG. 3-1 is a line graph illustrating the effect of late peak real-time traffic on the time when a driver receives passengers;
3-2 is a line graph of the influence of the real-time road conditions of 30 minutes at 20 nights on the time when the driver receives the passengers;
3-3 are line graphs illustrating the influence of 30-minute real-time road conditions at noon 13 on the time when a driver receives passengers;
FIGS. 3-4 are line graphs illustrating the effect of real-time traffic on the idle distance before the driver receives a passenger;
fig. 3-5 are line graphs illustrating the effect of real-time road conditions on the individual power.
Detailed Description
It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict. The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.
To better illustrate the invention, a further explanation is made for the Dijkstra algorithm: the Dijkstra algorithm can calculate the shortest path from one vertex to other vertexes in the graph theory, and solves the shortest path problem in the weighted graph by the idea of the greedy algorithm. The algorithm is expanded layer by layer to the outside by taking the starting point as the center until all nodes are traversed. The Dijkstra algorithm declares an array distance to preserve the shortest distance from the source point to each of the other vertices and a set of vertices for which the shortest path has been found: and T. Initially, the path weight of the source point s is given as 0, that is, distance [ s ] = 0. If there is a directly reachable edge (s, m) for the source point s, distance [ m ] is weight (s, m), and the path length of all other vertices (where s cannot be reached directly) is ∞. Initially, the set T only has a source point s, then selects the minimum value from the distance array, and adds the point to the set T, and at this time, completes a vertex, and then queries whether the newly added vertex can reach other vertices and whether the path length to other points through the vertex is shorter than the path length directly reaching the source point, and if so, replaces the values of the vertices in distance. And then finding the minimum value from the distance, repeating the steps until the set T contains all the vertexes of the graph, and finishing the algorithm.
The traditional Dijkstra algorithm uses the path as weight to perform global breadth-first search, and uses the traditional Dijkstra algorithm to search surrounding drivers by taking passengers as source points and select the driver with the shortest path to dispatch a list.
Fig. 1 is a flowchart of a net taxi scheduling method based on the improved Dijkstra algorithm (dixtre algorithm), according to an embodiment of the present invention. As shown in fig. 1, the scheduling method includes the following steps:
s1, acquiring road network data, POI data, taxi journey GPS data and real-time road condition data in the designated area; specifically, the method comprises the following steps:
the road network data is downloaded to a road network OSM file through an Open Street Map (OSM, Open Map), and a point set and an edge set in the road network data both adopt a mercator projection coordinate system, and the mercator projection coordinate system is converted into an GCj-02 coordinate system (a coordinate system of a geographic information system prepared by the chinese national surveying and mapping bureau) to obtain a converted road network intersection set and a converted road network edge set. The converted road network intersection point set consists of an Allid column, a lon column and a lat column, wherein the Allid column represents a node ID in the whole map, the lon column represents a node longitude, and the lat column represents a node latitude; the converted road network edge set consists of an ID column, a newStart column, a newEnd column, a length column and a freespeed column, wherein the ID column represents an edge ID in the whole map, the newStart column represents a node Allid of a starting point, the newEnd column represents a node Allid of an end point, the length column represents an edge weight, namely the length of a road section, the unit is meter, and the freespeed column represents the highest speed limit, and the unit is meter per second.
POI (Point of Interest) data are obtained from the existing APP map software by using a crawler technology, and each piece of POI data comprises longitude, latitude, namePOI name, address, province, city, business area, big _ type, middle _ type and small _ type; the taxi journey GPS data at least comprises a vehicle id, a device number, a direction angle and GPS longitude and latitude; the method comprises the steps of obtaining a gathering set of crowd distribution and people flow in a designated area by integrating POI data and taxi journey GPS data, dividing a map into grids of 1km by 1km, wherein the larger the data volume in the grids is, the larger the people flow and the people density in the area are, and the opposite is. It should be noted that, the name of namePOI is noun information of POI point, and specific examples are shown in table 1; the big _ type attribute is descriptive words and is a large category, including science and education culture services, scenic spots, company enterprises and the like; the middle _ type attribute will be more specifically a detailed category, for example, science and education culture services including schools, scientific research institutions, science and education culture sites, and the like; the small _ type attribute is the most specific and the most detailed category, such as schools including kindergarten, elementary school, middle school, higher school, etc., see in particular table 2 below:
the real-time road condition data are acquired by using a python crawler technology through a traffic situation interface provided by the existing APP map software API.
It should be noted that the existing APP map software may be a highrise map, a Baidu map, or other APP that can implement a map function.
S2, based on the Dijkstra algorithm, adding road network distance influence factors and time influence factors based on real-time road conditions into the weight to form an improved Dijkstra algorithm;
preferably, the longitude and latitude of the lower left corner point and the longitude and latitude of the upper right corner point of each grid can obtain real-time road condition data in the area, and the real-time road condition state of the area is determined according to the serious congestion, slow driving and smooth priority of the road condition of each grid, so that a basis is provided for determining the time influence factor.
S3, generating random passenger and driver distribution according to the POI data and the comprehensive data of the taxi journey GPS, setting a dispatching range for each randomly generated passenger, and searching for the driver and the route in the dispatching range by improving the Dijkstra algorithm to obtain the optimized route taking the benefits of the passenger, the driver and the network platform into consideration. The steps are embodied as follows:
searching in the dispatching range by using a randomly generated passenger as an origin, and calculating the distance of the driver with the shortest distance from all the drivers in the dispatching range
Calculating the time of the driver with the shortest estimated arrival time under the current road condition
According to the formula (1), calculating the comprehensive weight of each driver, and according to the comprehensive weight of the drivers, selecting the driver with the maximum comprehensive weight from all drivers to receive the order and dispatch:
in the formula (I), the compound is shown in the specification,
the distance of the ith driver from the origin passenger,
the estimated arrival time of the ith driver at the origin passenger under the current road conditions,
is the comprehensive weight of the ith driver, A is the road network distance influence factor, B is the time influence factor, wherein,
the expression of (a) is as follows:
wherein L ismLength of the mth road in the route for the ith driver to pick up the passenger, SmSpeed limit, Rs, for the mth road in the route for the ith driver to pick up and deliver passengersmFor the current congestion situation in the ith driver pick-up passenger path, n represents the total number of road segments in the driver pick-up passenger path.
To better illustrate the technical solution of the present invention, the following is demonstrated by way of example:
fig. 2 is a Map of a city road network provided by an Open Street Map, in which the original data set of the road network includes 16964 nodes (road network intersections) and 37640 links (roads). In a road network osm file downloaded by an Open Street Map, a point set and an edge set both adopt a mercator projection coordinate system, for convenience of experimental analysis, a coordinate system of data in an experiment is unified, the mercator projection coordinate system is converted into a GCJ-02 coordinate, a road network cross point set after conversion is shown in a table 3, an Allid column is a node ID in a full Map, a lon column is a node longitude, and a lat column is a node latitude:
the set of converted road network edges is shown in table 4, where the ID column is an edge ID in the whole map, the newStart column is a node Allid of a starting point, the newEnd column is a node Allid of an end point, the length column is an edge weight, that is, the length of the road segment, is in meters, the freespeed column is the highest speed limit, and the unit is meters per second:
in the example, 66842 strips of POI data are obtained through data cleaning, the number of taxis is about 8000, GPS data recorded every 30 seconds of the urban taxis on a working day is selected for experiment, the data is preprocessed, and repeated and abnormal data are deleted. By integrating POI data and taxi journey GPS data, the crowd distribution and the collective set of the flow of people of the city can be obtained, the map is divided into a road network and divided into grids of 1km by 1km, and the larger the data volume in the grids is, the larger the flow of people and the density of people in the area are, and the contrary is, the larger the flow of people and the density of people in the area are. In the simulation experiment, random passenger and driver distribution is generated according to the comprehensive data of POI and taxi GPS, which accords with the real situation.
In this example, the urban road network is divided into 1km by 1km grids by using road network division, and the total number is 55 rows by 78 columns, so when acquiring road condition data, rectangular area traffic situation interfaces are also adopted for acquisition. And obtaining real-time road condition data in the area by providing the longitude and latitude of the lower left corner point and the longitude and latitude of the upper right corner point of each grid. The original content of the real-time traffic data of each area is shown in the following table 5:
in the example, the selected real-time road condition data are divided into three time periods of 13:30 in the noon, 17:30 in the afternoon and 20:30 in the evening, the original real-time road condition data of the 4290 rectangular areas are acquired in each time period and stored in an Excel table, and the original real-time road condition data are divided into three grades of unblocked (1), slow running (2) and severe congestion (3) according to the analysis of the actual urban data. And determining the real-time road condition state of the area according to the serious congestion, slow running and smooth priority of each rectangular grid road condition, and finally summarizing the road conditions into a real-time road condition table with 55 lines and 78 columns, thereby providing a basis for determining time influence factors.
The invention improves Dijkstra algorithm, adds factors influencing the current road condition on the vehicle speed on the traditional Dijkstra algorithm, compares the traditional Dijkstra algorithm, and compares the success rate of order receiving of an analysis platform, the waiting time of passengers, and the dead time and distance of a driver, thereby proving that the improved Dijkstra algorithm is effective for considering three-party benefits.
The invention selects a road network map as a rectangular area with the east-west span of about 77.8654 kilometers and the north-south span of about 54.5567 kilometers for simulation modeling. In a real network car booking matching scene, the number of network car booking is usually large, and in the modeling process, the calculation amount of a server is increased due to the number of generated network car booking, so that from the viewpoint of saving calculation cost, a scheduling range with the side length of 10 kilometers is set for each randomly generated passenger, and path search is carried out on a scheduled route in the scheduling range.
Randomly generating passenger number x and driver number y, setting dispatching range R, searching in the dispatching range with the passenger as origin, and calculating dispatchingThe distance between the shortest drivers among all drivers in the range
Calculating the time of the driver with the shortest estimated arrival time under the current road condition
According to the formula (1), calculating the comprehensive weight of each driver, and according to the comprehensive weight of the drivers, selecting the driver with the maximum comprehensive weight from all drivers to receive the order and dispatch:
in the formula (I), the compound is shown in the specification,
the distance of the ith driver from the origin passenger,
the estimated arrival time of the ith driver at the origin passenger under the current road conditions,
the total weight of the ith driver is A, the road network distance influence factor and B, the time influence factor.
The driving speed of the driver on the path is influenced by the highest speed limit and the congestion degree in the real-time road condition, the driver can pass through a plurality of roads on the path for picking up and sending the passengers, the congestion condition and the speed limit of each road can be different, so the time for picking up and sending the passengers by the driver is
The expression of (a) is as follows:
wherein L ismLength of the mth road in the route for the ith driver to pick up the passenger, SmSpeed limit, Rs, for the mth road in the route for the ith driver to pick up and deliver passengersmFor the current congestion situation in the ith driver pick-up passenger path, n represents the total number of road segments in the driver pick-up passenger path.
The time for receiving the passenger by the driver is reduced along with the increase of the weight of the considered time through experimental comparison. 3-1-3, the abscissa is the time impact factor and the ordinate is the driver's average pickup time. When the time influence factor of the abscissa is 0, namely the traditional Dijkstra algorithm, the influence of the congestion degree in the late peak on the vehicle speed is not considered, so that the time for receiving the passenger by the driver is more than that of the optimized Dijkstra algorithm. Fig. 3-1 is a line graph illustrating the effect of late peak real-time road conditions on the time when a driver receives passengers. The optimized algorithm, as shown in fig. 3-1, reduces the time for the passenger to wait for the driver by approximately 50 seconds when the time impact factor is 1 in the case of extreme congestion at late peak.
3-2 is a line graph of the influence of the real-time road conditions of 30 minutes at 20 nights on the time when the driver receives the passengers; fig. 3-3 are line graphs illustrating the influence of 30 minutes of real-time road conditions at noon 13 on the time when a driver receives passengers. As shown in fig. 3-2 and 3-3, even in the night of 20 o 'clock and 30 o' clock and 13 o 'clock in the afternoon of 30 o' clock, when the time influence factor is 1, Dijkstra's algorithm considering real-time road conditions can reduce the waiting time of passengers for drivers by more than 30 seconds compared with the conventional Dijkstra's algorithm.
Considering that the weight proportion of the real-time road condition in the improved Dijkstra algorithm is too high, and the weight proportion of the distance considered in the traditional Dijkstra algorithm is too low, in order to reduce the idle distance and time of the driver as much as possible in the process of receiving the passenger by the driver, in the experiment, the influence of the real-time road condition on the idle distance before the driver receives the passenger is compared, and the result is shown in fig. 3-4. In fig. 3-4, the abscissa is a time influence factor, the ordinate is an average driver's route for receiving a single, and the three broken lines in the graph respectively represent influence broken lines of real-time road conditions of 5 pm 30 minutes, 8 pm 30 minutes and 1 pm 30 minutes before the driver receives the passenger's empty distance.
The result shows that the influence of excessive real-time road conditions on the scheduling algorithm is ignored, the idle distance of a driver before the driver receives passengers is increased, and therefore, in order to take the common benefits of drivers and passengers into consideration, the distance influence factor A in the improved Dijkstra algorithm is set to be 0.5, and the time influence factor B is also set to be 0.5. At the moment, the idle time of the driver is greatly reduced, the idle distance is well controlled, and the benefit of the driver is fully guaranteed.
Finally, comparing the success rate of order taking of the observation platform, under the condition that the distance influence factor a is set to 0.5 and the time influence factor B is also set to 0.5, as shown in fig. 3-5, the abscissa of the success rate is the time influence factor, the ordinate is the average success rate of order taking of the driver, and three broken lines in the graph respectively represent the influence broken lines of the real-time road conditions of 5 pm 30 minutes, 8 pm 30 minutes, and 1 pm 30 minutes to the success rate of order taking of the driver, from which it can be known that: the success rate of order taking is higher than that of the traditional Dijkstra algorithm. Therefore, the improved Dijkstra algorithm can give consideration to the benefits of a driver, a passenger and a platform.
According to the invention, a classic Dijkstra algorithm is selected, the weight factor is adjusted on the basis of a road network distance influence factor and a real-time road condition-based time influence factor, and a new passenger demand and driver matching strategy is designed by comprehensively considering passenger waiting time, driver dead time and distance and network car booking platform order receiving rate, so that the utilization rate of the network car booking driver is greatly improved, the benefits of the network car booking platform and the passengers are considered, and the economic benefit is outstanding.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (7)
1. The network taxi appointment scheduling method based on the improved Dijkstra algorithm is characterized by comprising the following steps of:
s1, acquiring road network data, POI data, taxi journey GPS data and real-time road condition data in the designated area;
s2, based on the Dijkstra algorithm, adding road network distance influence factors and time influence factors based on real-time road conditions into the weight to form an improved Dijkstra algorithm;
s3, generating random passenger and driver distribution according to the POI data and the comprehensive data of the taxi journey GPS, setting a dispatching range for each randomly generated passenger, and searching for the driver and the route in the dispatching range by improving the Dijkstra algorithm to obtain an optimized route taking the benefits of the passenger, the driver and the network platform into consideration;
the step S3 is specifically represented as:
searching in the dispatching range by using a randomly generated passenger as an origin, and calculating the distance of the driver with the shortest distance from all the drivers in the dispatching range
Calculating the time of the driver with the shortest estimated arrival time under the current road condition
According to the formula (1), calculating the comprehensive weight of each driver, and according to the comprehensive weight of the drivers, selecting the driver with the maximum comprehensive weight from all drivers to receive the order and dispatch:
in the formula (I), the compound is shown in the specification,
the distance of the ith driver from the origin passenger,
the estimated arrival time of the ith driver at the origin passenger under the current road conditions,
is the comprehensive weight of the ith driver, A is the road network distance influence factor, B is the time influence factor, wherein,
the expression of (a) is as follows:
wherein L ismLength of the mth road in the route for the ith driver to pick up the passenger, SmSpeed limit, Rs, for the mth road in the route for the ith driver to pick up and deliver passengersmFor the current congestion situation in the ith driver pick-up passenger path, n represents the total number of road segments in the driver pick-up passenger path.
2. The improved Dijkstra algorithm-based network appointment scheduling method as claimed in claim 1, wherein the road network data is downloaded into a road network osm file through an Open Street Map, and the point set and the edge set in the road network data are obtained by converting a mercator projection coordinate system into an GCj-02 coordinate system by using a mercator projection coordinate system, so as to obtain a converted road network intersection point set and a converted road network edge set.
3. The improved Dijkstra algorithm-based net appointment scheduling method of claim 2, wherein the set of transformed road network intersection points is composed of an Allid column, a lon column and a lat column, wherein the Allid column represents a node ID in the whole map, the lon column represents a node longitude, and the lat column represents a node latitude; the converted road network edge set consists of an ID column, a newStart column, a newEnd column, a length column and a freespeed column, wherein the ID column represents an edge ID in the whole map, the newStart column represents a node Allid of a starting point, the newEnd column represents a node Allid of an end point, the length column represents an edge weight, namely the length of a road section, the unit is meter, and the freespeed column represents the highest speed limit, and the unit is meter per second.
4. The improved Dijkstra algorithm-based network appointment scheduling method as claimed in claim 1, wherein the POI data is obtained from existing APP map software by using a crawler technology, and each POI data comprises longitude, latitude, namePOI name, address, province, city, business area, big _ type, middle _ type and small _ type; the taxi journey GPS data at least comprises a vehicle id, a device number, a direction angle and GPS longitude and latitude; the method comprises the steps of obtaining a gathering set of crowd distribution and people flow in a designated area by integrating POI data and taxi journey GPS data, dividing a map into grids of 1km by 1km, wherein the larger the data volume in the grids is, the larger the people flow and the population density in the area are.
5. The network appointment scheduling method based on the improved Dijkstra algorithm as claimed in claim 4, wherein the real-time traffic data is obtained by using a python crawler technology through a traffic situation interface provided by existing APP mapping software API.
6. The network appointment scheduling method based on the improved Dijkstra algorithm as claimed in claim 1, wherein the optimized path taking the benefits of the passengers, the drivers and the network appointment platform into consideration is a path with high success rate of platform order taking, short waiting time of the passengers and small dead time and distance of the drivers.
7. The network taxi appointment scheduling method based on the improved Dijkstra algorithm as claimed in claim 5, wherein the real-time road condition data in the area can be obtained by the longitude and latitude of the lower left corner point and the longitude and latitude of the upper right corner point of each grid, the real-time road condition state in the area is determined according to the serious congestion, slow traveling and smooth priority of the road condition of each grid, and a basis is provided for determining the time influence factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110222684.XA CN112562309B (en) | 2021-03-01 | 2021-03-01 | Network taxi appointment scheduling method based on improved Dijkstra algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110222684.XA CN112562309B (en) | 2021-03-01 | 2021-03-01 | Network taxi appointment scheduling method based on improved Dijkstra algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112562309A CN112562309A (en) | 2021-03-26 |
| CN112562309B true CN112562309B (en) | 2021-05-07 |
Family
ID=75034902
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110222684.XA Active CN112562309B (en) | 2021-03-01 | 2021-03-01 | Network taxi appointment scheduling method based on improved Dijkstra algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112562309B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113256015B (en) * | 2021-06-03 | 2021-11-16 | 北京亿心宜行汽车技术开发服务有限公司 | Order prediction-based designated driver scheduling method |
| CN113743767B (en) * | 2021-08-27 | 2022-11-04 | 广东工业大学 | Vehicle dispatching method, system, computer and medium based on time and safety |
| CN114812590A (en) * | 2022-02-26 | 2022-07-29 | 贵州民族大学 | Data-driven taxi route recommendation method |
| CN116108288B (en) * | 2023-02-06 | 2024-09-06 | 苏州大学 | Method and system for searching optimal get-on and get-off points in taxi taking service |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020884A (en) * | 2012-12-06 | 2013-04-03 | 苏州亿倍信息技术有限公司 | Information interaction control method and system |
| CN104639646A (en) * | 2015-02-12 | 2015-05-20 | 北京嘀嘀无限科技发展有限公司 | Method and equipment for processing user request |
| US10637951B2 (en) * | 2016-04-08 | 2020-04-28 | Massachusetts Institute Of Technology | Systems and methods for managing data proxies |
| CN111879329A (en) * | 2020-07-30 | 2020-11-03 | 郑州天迈科技股份有限公司 | Customized public transport passable shortest path calculation method based on A-x algorithm |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853294B (en) * | 2010-05-21 | 2011-12-14 | 中国科学院地理科学与资源研究所 | Multi-mode multi-standard path search method based on genetic algorithm |
| CN102306455B (en) * | 2011-09-07 | 2013-12-11 | 陕西理工学院 | Taxi intelligent service system and service method thereof |
| CN103021167B (en) * | 2012-12-06 | 2014-12-17 | 苏州亿倍信息技术有限公司 | Method and system for taxi information interaction |
| US10861332B2 (en) * | 2016-03-28 | 2020-12-08 | Aisin Aw Co., Ltd. | Communication terminal, server device, route retrieval system, and computer program |
| US20170364968A1 (en) * | 2016-06-16 | 2017-12-21 | Conduent Business Services, Llc | Method and system for cost sharing in a pooled vehicle |
| US10156452B2 (en) * | 2016-11-14 | 2018-12-18 | Conduent Business Service, Llc | Method and system for ridesharing management |
| WO2019018381A1 (en) * | 2017-07-21 | 2019-01-24 | Walmart Apollo, Llc | System and method to enable sharing of delivery resources between non-affiliated entities |
| CN110119955B (en) * | 2018-02-06 | 2022-11-11 | 北京嘀嘀无限科技发展有限公司 | Order rate estimation method and device |
| KR102423172B1 (en) * | 2018-03-20 | 2022-07-22 | 모빌아이 비젼 테크놀로지스 엘티디 | Systems and methods for navigating a vehicle |
| EP3683742A1 (en) * | 2019-01-18 | 2020-07-22 | Naver Corporation | Method for computing at least one itinerary from a departure location to an arrival location |
-
2021
- 2021-03-01 CN CN202110222684.XA patent/CN112562309B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020884A (en) * | 2012-12-06 | 2013-04-03 | 苏州亿倍信息技术有限公司 | Information interaction control method and system |
| CN104639646A (en) * | 2015-02-12 | 2015-05-20 | 北京嘀嘀无限科技发展有限公司 | Method and equipment for processing user request |
| US10637951B2 (en) * | 2016-04-08 | 2020-04-28 | Massachusetts Institute Of Technology | Systems and methods for managing data proxies |
| CN111879329A (en) * | 2020-07-30 | 2020-11-03 | 郑州天迈科技股份有限公司 | Customized public transport passable shortest path calculation method based on A-x algorithm |
Non-Patent Citations (3)
| Title |
|---|
| 基于乘客需求及分布的出租车调度方法技术研究;贾婷;《中国优秀硕士学位论文全文数据库 信息科技辑》;20131225;正文第20-45页 * |
| 基于轨迹数据场的热点区域提取及空间交互分析——以深圳市为例;周博等;《热带地理》;20190131;第39卷(第1期);正文第118页 * |
| 智能巡检系统中导航模块的设计与实现;伍思集;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180315(第03期);正文第15页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112562309A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112562309B (en) | Network taxi appointment scheduling method based on improved Dijkstra algorithm | |
| US10317219B1 (en) | System and method for the calculation and use of travel times in search and other applications | |
| Kong et al. | Mobility dataset generation for vehicular social networks based on floating car data | |
| Alexander et al. | Assessing the impact of real-time ridesharing on urban traffic using mobile phone data | |
| TWI735292B (en) | A method for providing a route for a journey from a source location to a target location | |
| US9677904B2 (en) | Generating travel time data | |
| US7761225B2 (en) | Routing method and system | |
| RU2523192C2 (en) | Methods, devices and map databases for green routing | |
| CN112384758A (en) | Multi-mode method for traffic route selection | |
| CN108985506A (en) | Planning driving path recommended method, prediction technique, acquisition methods and its device | |
| CN111060119B (en) | City slow-travel navigation method and system based on street view visual element scoring | |
| CN112991743B (en) | Real-time traffic risk AI prediction method based on driving path and system thereof | |
| CN110648027B (en) | Self-driving tour digital line interactive generation system and working method thereof | |
| Dezfouli et al. | A novel tour planning model using big data | |
| CN105389751A (en) | Travel service method and device | |
| JP2022526401A (en) | Methods and equipment for searching or comparing sites using routes or route distances between sites and locations within the transportation system. | |
| CN109726868A (en) | Paths planning method, device and storage medium | |
| Tomaras et al. | Crowd-based ecofriendly trip planning | |
| CN103839424B (en) | A kind of taxi empty navigational system and method based on recruiting car expectation value | |
| Bartelme | Geographic information systems | |
| Yang et al. | How the urban microclimate and outdoor thermal comfort can affect intra-city mobility patterns: Evidence from New York City | |
| CN114065055A (en) | Intelligent tour route recommendation method and query system thereof | |
| Phiboonbanakit et al. | How does taxi driver behavior impact their profit? Discerning the real driving from large scale GPS traces | |
| JP4901073B2 (en) | Method for producing simplified digital road network data | |
| CN114791732B (en) | Path planning method, device, equipment and computer readable storage 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |