CN108009666A - The preferential optimal path computation method of level based on dynamic road network - Google Patents
The preferential optimal path computation method of level based on dynamic road network Download PDFInfo
- Publication number
- CN108009666A CN108009666A CN201610966010.XA CN201610966010A CN108009666A CN 108009666 A CN108009666 A CN 108009666A CN 201610966010 A CN201610966010 A CN 201610966010A CN 108009666 A CN108009666 A CN 108009666A
- Authority
- CN
- China
- Prior art keywords
- road network
- level
- areas
- section
- optimal path
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Navigation (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of preferential optimal path computation method of level based on dynamic road network, including:Step 1, road network is divided by level according to category of roads, generates the corresponding Voronoi diagram of each level road network respectively;Step 2, route searching is carried out using the search strategy of stratification, searches for the main part and component of optimal path respectively based on Shortest Path Searching method;Step 3, optimal path is generated according to the main part of optimal path and component.The present invention can greatly reduce running time of the in-trips vehicles on road, and can respond the change of traffic flow modes in road network, and then provide reliable real time solution for the Path selection of vehicle driving, so that trip scheme has more adaptability and reliability.
Description
Technical field
The invention belongs to technical field of intelligent traffic, more particularly to a kind of preferential optimal path of level based on dynamic road network
Computational methods.
Background technology
In recent years, how an optimal path is found between the beginning and end of trip to grind as the hot topic of intelligent transportation
Study carefully topic.What it not only directly affects people goes out line efficiency, also relates to urban development and environmental pollution etc. and asks
Topic.It for example, vehicle travels on the poor road of a road conditions, can not only increase Trip Costs, and be further exacerbated by road friendship
The situation of pass blocking plug, and thing followed tail gas pollution problem.Many professional GIS softwares (such as ArcGIS)
Both provide and solved in road network by the function of the shortest path of source location to target location, but conventional method is all base
In static traffic stream information, not with reference to traffic flow on city road network at any time space-variant the characteristics of have to in-trips vehicles
Effect induction.
With the progress of Information and Communication Technology and the extensive use of GPS sensor, there is scholar to start with mobile object
Track dynamically monitors it to analyze its behavior.Wherein, the GPS track data of taxi because it includes
Instant road speed can accurately reflect road state and enjoy focus of attention.For example, there are some researchers to pass through meter
The current frequency of each section of road is calculated to extract the experience of driver and then provide effective navigation Service for other people.Also scholar
The traffic property of Over-size transport vehicle is studied from GPS track data.There is scholar to take up recently actual in understanding vehicle
Running route and GPS analog lines between relation.Nevertheless, how to be extracted more from substantial amounts of GPS track data
Useful information, is always an outstanding question in route planning industry.
In addition, during vehicle driving route is planned, the research for the cognitive behavior that can combine the mankind, which is not much, to be seen,
Although there is research to think that people are more likely to from turning on less route during trip by and being not limited solely to thing
The distance managed in meaning is most short, because this will greatly reduce the burden of people's cognition, but not further to the spy of these roads
Point carries out detailed analysis.In fact, the grade of road network is a base attribute of road, it is function according to road, position
Put and drive a vehicle capacity and the road network in whole city has been divided into different subsets.For the longer traffic path of distance, utilize
Road network grade guides traveler the burden that can substantially reduce its cognition.The reason for another is important is due to high-grade road
Net has often corresponded to preferable driving road-condition, and traveler passes through from high class road network can significantly save the time of trip.
Following bibliography involved in text:
[1]Chen,B.Y.,et al.,2014.Map-matching algorithm for large-scale low-
frequency floating car data.International Journal of Geographical Information
Science,28(1),22–38.doi:10.1080/13658816.2013.816427
[2]Daltona A.M.,Jones A.P.,Panter J.,and Ogilvie D.(2015),Are GIS-
modeled routes a useful proxy for the actual routes followed by commuters,
Journal of Transport &Health,2,219-229
[3]Geisberger R,Sanders P,Schultes D,and Vetter C 2012Exact routing
in large road networks using contraction hierarchies.Transportation Science
46:388–404
[4]Hess S.,Quddus M.,Rieser-Schussler N.,Daly A.(2015),Developing
advanced route choice models for heavy goods vehicles using GPS data,
Transportation Research Part E,77,29-44
[5]Hu J.H.,Huang Z.,Deng J.(2013),A hierarchical path planning method
using the experience of taxi drivers,Procedia-Social and Behavioral Sciences,
96,1898-1909
[6]Jia T.and Jiang B.(2012),Exploring human activity patterns using
taxicab static points,ISPRS International Journal of Geo-Information,1(1),89-
107.
[7]Jiang B.and Liu X.(2011),Computing the fewest-turn map directions
based on the connectivity of natural roads,International Journal of
Geographical InformationScience,25(7),1069-1082
[8]Jung S.and Pramanik S.(2002),An efficient path computation
modelfor hierarchically structured topographical road maps,IEEE
Transactionson Knowledge and Data Engineering,14(5):1029-1046.
[9]Meng L.K.,Hu Z.H.,Huang C.Q.,Zhang W.,and Jia T.(2015),Optimized
Route Selection Method based on the Turns of Road Intersections:A Case Study
onOversized Cargo Transportation,ISPRS Int.J.Geo-Inf.4(4),2428-2445
[10]Nejad M.M.,Lena Mashayekhy L.,RatnaBabuChinnamR.B.,and Phillips
A.(2016),Hierarchicaltime-dependent shortestpathalgorithms for vehicle
routingunder ITS,IIE Transactions,48(2),158-169
The content of the invention
The object of the present invention is to provide a kind of preferential optimal path computation method of level based on dynamic road network.
The present invention imitates the thought process that the mankind design trip route, i.e. traveler selects advanced road first, and first
Geographically whether it is connected without considering road.Then not connected high-grade road is progressively connected using the road network of inferior grade
Road, untill all roads are connected completely.Specifically, the present invention constructs a road network changed over time and therewith
Corresponding Voronoi diagram, Voronoi diagram is not only remarkably improved the connectivity of road, and then improves the reliability of optimal path.
Secondly, the invention also provides a kind of hierarchical path searching method based on global and local, i.e., from high-level road network recursively
Go out an end to end path to low-level road network search.
To reach above-mentioned purpose, the present invention adopts the following technical scheme that:
The preferential optimal path computation method of level based on dynamic road network, including:
Step 1, road network is divided into the road network of K grade according to category of roads, the road network of K grade is layered,
The road network of K level is obtained, generates the corresponding Voronoi diagram of each level road network respectively;Wherein, the road network of kth level is by kth etc.
Level, the road network of the 1st grade of the grade ... of kth -1 are formed, and k=1,2 ... K, K values take the category of roads quantity of road network;
Step 2, route searching is carried out based on Voronoi diagram, this step includes the main part of search optimal path and divides
Branch part;
The main part of the search optimal path further comprises:
The 2.1 initiating searches level using linear search algorithm by source point O and target point D matching road networks, by initiating searches
For the corresponding Voronoi diagram of level road network as current Voronoi diagram, O and the V areas where D are denoted as starting point V areas and terminal V respectively
Area, V areas, that is, Voronoi area;
2.2 carry out Shortest Path Searching based on current Voronoi diagram;
2.3 sequentially check the connectedness that shortest path includes section, to disconnected front and rear two section, with front and rear two-way
The minimum enclosed rectangle of the union in V areas is search range where section, performs sub-step 2.4;If being connected comprising all sections,
Terminate, perform sub-step 2.5;
2.4 judge whether current search level road network is lowest hierarchical level road network, if so, based on all layers in search range
Level road net data carries out Shortest Path Searching, and the path searched is saved in path candidate data set;Otherwise, by current search
Voronoi diagram corresponding to next level road network of level road network is as current Voronoi diagram, and traversal is current in search range
All V areas of Voronoi diagram, determine in current Voronoi diagram according to the spatial relationship between road section in current Voronoi diagram
Starting point V areas and terminal V areas, perform sub-step 2.2;
The component of the search optimal path further comprises:
2.5 current search level road networks be initialized as sub-step 2.1 matched initiating searches level road network next layer
Level road network;
2.6 are based on the corresponding Voronoi diagram of current search level road network, using the V areas where O or D as search range, are working as
The V areas intersected with the search range are found in preceding search level road network;
2.7 judge whether the intersecting V areas are V areas where main part, if so, by under current search level road network
One level road network performs sub-step 2.6 as current search level road network;Otherwise, it is corresponding based on current search level road network
Voronoi diagram, using the V areas where O as starting point V areas or using the V areas where D as terminal V areas, using V areas where main part as eventually
Point V areas or starting point V areas;
2.8 carry out Shortest Path Searching based on the corresponding Voronoi diagram of current search level road network;
2.9 using next level road network of current search level road network as current search level road network, based on current search
The corresponding Voronoi diagram of level road network, using the V areas where O as starting point V areas or using the V areas where D as terminal V areas, to be searched for
Shortest path where V areas be terminal V areas or starting point V areas, perform sub-step 2.8;
2.10 repetition sub-steps 2.8~2.9 are until it is lowermost layer that O or D, which drops down onto shortest path or current search level road network,
Level road network;
Step 3, the main part of optimal path is obtained according to step 2 and component generates optimal path.
The road network to K grade described in step 1 is layered, and is specially:
(1) section is divided and is merged using following rule:
The division principle is:
When a plurality of section intersects at an intersection point, and the intersection point is different for the endpoint in a plurality of section when, to a plurality of road
Section follows when division that " inferior grade road network is interrupted by high class road network, and high class road network is not beaten by inferior grade road network into line splitting
It is disconnected ";
The combination principle is:
To only having the two of an intersection point sections, if the intersection point is at the same time the endpoint in two sections, to two sections into
Row merges;
(2) road network after being divided and merged based on section is layered, wherein, the road network of kth level is by kth grade, the
The road network of the 1st grade of k-1 grades ... is formed.
The starting level that search road network is determined using linear search method described in sub-step 2.1, is specially:
The line of source point O and target point D are extended to the 10% of line total length, that is, obtain searching operators, obtains and searches
The level in all sections that rope operator intersects, highest level originate level.
In sub-step 2.1, if the V areas where O and D are identical, section in the identical V areas is saved in path candidate data set
Afterwards, terminate, then perform sub-step 2.5.
In sub-step 2.2, using weight of the vehicle by history average time in each V areas as each V areas, based on current
Voronoi diagram carries out Shortest Path Searching and obtains shortest path.
The vehicle is obtained with the following method by the history average time in each V areas:
The sample track point in V areas is obtained, the history average speeds of vehicle in V areas, V are calculated using sample track point
The physical length divided by history average speeds in section, i.e., the history average time that vehicle passes through V areas are included in area.
In sub-step 2.3, preceding a road section in two sections before and after connection is saved in path candidate data set;If connection
Two sections in latter bar section be the last item section, then latter bar section is also saved in path candidate data set.
Step 3 is specially:
Main part and component institute are proceeded as follows comprising section:
(1) overlapping section is removed;
(2) intersection point in section before and after being calculated using the spatial relationship between section;
(3) two sections of head and the tail are trimmed, removes unreasonable part on head and the tail section;
(4) former and later two intersection points are sequentially connected, up to optimal path.
Step 3 further includes:
The nearest neighbor point of source point and/or target point on two sections of head and the tail is found, as the true source point of optimal path
And/or target point.
Compared to the prior art, the invention has the advantages that and beneficial effect:
The present invention not only can reflect the mode of thinking of human cognitive well, and also organically combine classics
Dijkstra's algorithm and the searching algorithm based on level road network.It is demonstrated experimentally that the present invention can greatly reduce in-trips vehicles in road
On running time, and the present invention can respond the changes of traffic flow modes in road network, and then be the Path selection of vehicle driving
Reliable real time solution is provided, so that trip scheme has more adaptability and reliability.
Brief description of the drawings
Fig. 1 is the schematic diagram of road network and its Voronoi diagram, wherein, figure (a) is road network and its Voronoi diagram, and figure (b) is
Topological relation in each level road network between each section, figure (c) are the topological relation of each section Voronoi diagram in each level road network;
Fig. 2 is the specific product process schematic diagram of Voronoi diagram;
Fig. 3 is the path connected degree distribution schematic diagram of road network;
Fig. 4 is the path connected degree distribution schematic diagram of Voronoi diagram;
Fig. 5 is the history average speeds distribution map of city road network in different periods;
Fig. 6 is urban road network's road speed spatial and temporal distributions schematic diagram of b periods, c periods, d periods and e periods in Fig. 5,
Wherein, urban road network's road speed spatial and temporal distributions schematic diagram that (a) is the b periods is schemed, figure (b) is period c relative to period b city
The change schematic diagram of city's road network uplink vehicle speed, figure (c) show for period d relative to the change of period c city road network road speed
It is intended to, figure (d) is change schematic diagrams of the period e relative to period d city road network road speed;
Fig. 7 is flow diagram of the present invention;
The optimal path schematic diagram that Fig. 8 is obtained by embodiment;
Fig. 9 is obtained the service condition of roads at different levels in optimal path result by embodiment, wherein, figure (a) is embodiment
The middle service condition that roads at different levels in optimal path result are obtained using method A;(b) is schemed to use method B institutes in embodiment
Obtain the service condition of roads at different levels in optimal path result;(c) is schemed by obtaining optimal path knot using method C in embodiment
The service condition of roads at different levels in fruit;
Figure 10 obtains the statistical chart of road section length and section quantity accounting in optimal path result by embodiment, wherein,
Figure (a) is obtained the statistical chart of road section length accounting in optimal path result by embodiment;Figure (b) is obtained optimal by embodiment
The statistical chart of route result Road segment number accounting;
Figure 11 is the computational efficiency comparison diagram of three kinds of methods in embodiment.
Embodiment
Road network is actually a data acquisition system with hierarchical structure associated with road network grade.Mathematically, put down
Face road network is expressed as G (V, E), and V represents the set of intersection, and E represents the set in section.For example, there is three grades
Road network can be expressed as HG (V', E'), V' is by V1、V2And V3The set of composition, E' are by E1、E2And E3The set of composition,
V1、V2、V3The set of the intersection endpoint of different levels, E are represented respectively1、E2、E3The section of different levels is represented respectively
Set.Wherein:
E1=e | e ∈ E and e.class=1 };
E2=e | e ∈ E and e.class=1or 2 };
E3=e | e ∈ E and e.class=1or 2or 3 };
E.class represents the grade of section e, E1Represent the set in the section of the first level, the set is by the first estate
Section is formed;E2Represent the set in the section of the second level, which is made of the section of the first estate and the second grade;E3Table
Show the set in the section of third layer level, which is made of the section of the first estate, the second grade and the tertiary gradient.
V1=v | v ∈ V and e (v) .class=1 },
V2=v | v ∈ V and e (v) .class=1or 2 };
V3=v | v ∈ V and e (v) .class=1or 2or 3 };
The grade of road, V where e (v) .class represent intersection endpoint v1Represent the intersection of the first level
The set of mouth endpoint, the set are made of the intersection endpoint of the first estate road;V2Represent the intersection of the second level
The set of endpoint, the set are made of the intersection endpoint of the first estate road and the second grade road;V3Represent third layer level
Intersection endpoint set, the set by the first estate road, the second grade road and tertiary gradient road intersection
Mouth endpoint is formed.
HG once (V', E') construction complete, its corresponding Voronoi diagram can also be obtained accordingly.Fig. 1 is a tertiary road
The schematic diagram of net and its Voronoi diagram.Wherein, Fig. 1 (a) show the first level, the second level and third layer level road network and
Its Voronoi diagram, it should be noted that each road corresponds to a Voronoi area, and inferior grade road network is by high class road network
Interrupted, but high class road network will not be interrupted by inferior grade road network.Fig. 1 (b) is shown in each level road network between each section
Topological relation.Fig. 1 (c) shows the topological relation of the corresponding Voronoi diagram in each section in each level road network.Can from figure
Go out, the data structure based on Voronoi diagram substantially improves the connectedness between road network.
Therefore, the generating process of Voronoi diagram can be described as following three steps, with reference to figure 2:
First, planar road net data are filtered according to category of roads, road network is divided into three sub-networks, is remembered respectively
For the first estate road network, the second grade road network, tertiary gradient road network.
Secondly, road network is layered according to hierarchical rule, obtains the road network of three levels, be denoted as the first level road respectively
Net, the second level road network, third layer level road network.
Used hierarchical rule is as follows:
(1) when a plurality of section intersects at an intersection point, and the intersection point is different for the endpoint in a plurality of section when, it is a plurality of to this
Section follows when division that " inferior grade road network is interrupted by high class road network, and high class road network is not beaten by inferior grade road network into line splitting
It is disconnected ".
When section is intersected, inferior grade road network is interrupted by high class road network, and high class road network is not beaten by inferior grade road network
Disconnected, this is primarily to avoided producing failure at inferior grade road network search path, and then improve the reliability of route searching.Beat
The operation of open circuit net is shown in Fig. 1 (a) primarily to ensure that inferior grade road network can effectively be connected high class road network.
(2) when two sections have and an only intersection point, and the intersection point is the endpoint in this two sections at the same time, merges this at this time
Two sections, so can effectively reduce section quantity.
(3) road network after being divided and merged based on section is layered, wherein, third layer level road network is by the first estate road
Net, the second grade road network and tertiary gradient road network are formed, and the second level road network is by the first estate road network and the second grade road network structure
Into the first level road network, that is, the first estate road network.
Finally, corresponding Voronoi diagram is generated based on each level road network respectively.It is corresponding by being generated to road network
Voronoi diagram, the connectedness of road can greatly enhance, and see Fig. 3~4.
In the present invention, by generating corresponding Voronoi area to each section so that the mileage chart layer data conversion of wire
For planar data so that the connectedness between original section is changed.It is original not connect but adjacent road in spatial dimension
Section but becomes adjacent key element in planar figure layer, and the road originally connected still maintains connective in planar figure layer,
This just considerably increases the possibility of Path selection.
When Voronoi area connects, and when actual section does not connect, using the Voronoi area where respective stretch as
Search range looks for the road net data of next level, and optimum route search process is carried out in the road net data found.With this
Analogize, until the section in Voronoi area connects or obtain two-by-two bottom level road net data.Finally, on each level road
The section fragment assembly found in net, that is, obtain final optimal path.
After the hierarchical road network construction complete based on Voronoi diagram, by distributing the row under different periods for each section
Vehicle velocity value, you can the road speed road net model changed over time.In present embodiment, it was divided into work by one week
Day and weekend, were divided into 12 periods, the road speed distributed is all sample tracks in corresponding Voronoi area by one day
The average value of history road speed of the point under day part.Fig. 5 is shown in different periods, and the history of city road network is averagely driven a vehicle
Velocity contour, therefrom it is not difficult to find that different grades of road network shows similar velocity pattern, i.e., possess on daytime compared with
Slow road speed and possess faster road speed at night.Secondly, average driving of the high class road network than inferior grade road network
Speed is much faster, this just illustrates to guide vehicle to travel to be beneficial to improve it on high class road network and go out line efficiency.By cutting
The VELOCITY DISTRIBUTION of four periods upper road network is taken to find that on period b, most of highest ranking road networks possess the maximum of road speed
Value, and the road speed of the road close to city center is relatively slow, it appears road is very crowded, sees Fig. 6 (a).Fig. 6 (b) is shown
Changing values of the period c relative to period b city road network uplink vehicle speed, must apparently, and road speed averagely have dropped
0.89km/h, and the road speed on most level-one road network decreases drastically.Fig. 6 (c) shows that period d is opposite
In the changing value of period c city road network uplink vehicle speed, it can be found that the road speed average in road network have dropped about 0.66
Kilometer/hour, most change is still occurred on turnpike road.Fig. 6 (d) shows period e relative to period d, road network
Average speeds about have dropped 0.7km/h, and changed section is concentrated mainly on the road network of the tertiary gradient.
The present invention carries out route searching, including global search and local search two parts using the search strategy of stratification.
Global search is to build the road network skeleton formed by high-level road network, the road that global search is calculated using Voronoi diagram
Footpath is not necessarily truly connected.Local search then solves the problems, such as that high-level road network is disjunct using low-level road network.In fact,
This is a recursive procedure that gap between high-level road network is constantly sutured using low-level road network, ultimately form one it is optimal
Trip route.It can be seen from the above that searching method of the present invention is very suitable for handling the computational problem of optimal path in extensive road network,
It has filled up traditional dijkstra's algorithm and in ArcGIS based on the gap between level road network search algorithm.
Searching method of the present invention mainly includes following three big steps:
First, the main part of optimal path is searched for
This step is to find the main part of optimal path.Given source point O and target point D, utilizes linear search method
Determine the initial ranging level of search road network, the line of source point O and target point D are extended the 10% of its total length, that is, searched
Rope operator.The level in all sections intersected with searching operators is obtained, highest level originates level;Obtain initiating searches level
Road network and corresponding Voronoi diagram, using the corresponding Voronoi diagram of initiating searches level road network as current Voronoi diagram, really
Determine the Voronoi area where source point O and target point D.
If source point O is identical with the Voronoi area where target point D, section in the identical Voronoi area is saved in
Path candidate data set.If the Voronoi area where source point O and target point D differs, by the Voronoi areas where O and D
Domain is averaged respectively as starting point Voronoi area and terminal Voronoi area with vehicle by the history of each Voronoi area
Time is the weight of each Voronoi area, Shortest Path Searching is carried out based on current Voronoi diagram, to obtain shortest path.
Sequentially check and obtain the connectedness that shortest path includes section, if front and rear two sections connect, by previous bar
Section is saved in path candidate data set.If latter bar section is the last item section in two sections of connection, will be latter
Bar section is also saved in path candidate data set.If front and rear two sections do not connect, Voronoi where front and rear two sections is sought
The union in region, search range is used as using the minimum enclosed rectangle of the union.
Traveled through in search range all in Voronoi diagram corresponding to next level road network of current search level road network
Voronoi area, at this time, using next level road network as current search level road network, the corresponding Voronoi of next level road network
Figure is current Voronoi diagram.Determined according to the spatial relationship between road section in current Voronoi diagram in current Voronoi diagram
Starting point Voronoi area and terminal Voronoi area, and based on current Voronoi diagram carry out Shortest Path Searching.Judgement is searched
Whether the true path in the shortest path that rope obtains in two adjacent Voronoi areas intersects, if intersecting, this is adjacent
Section is saved in path candidate data set in Voronoi area., will be previous in the adjacent Voronoi area if non-intersect
The minimum enclosed rectangle of the union of Voronoi area and latter Voronoi area is searched as search range, the recurrence shortest path
Rope process.
According to depth-first search strategy, this step search process would be repeated for until real section is connected or
Search the road network of lowest hierarchical level.Each Voronoi area includes a real section, but Voronoi area is adjacent does not represent
Real section also connects, and there is the path connected only between source point and target point, then shows that real roads have been connected.
Vehicle is achieved in that by the history average time of Voronoi area:
The sample track point in Voronoi area is obtained, calculating vehicle in Voronoi area using sample track point goes through
History average speeds, the interior physical length divided by history average speeds for including section of Voronoi area, i.e. vehicle lead to
Spend the history average time of Voronoi area.
2nd, the component of searching route
This step is found from main part to source point O or the optimal path of target point D, it is substantially by main part
Respectively to the search procedure of source point O or target point D Step wise approximations.Specifically, first, with where source point O or target D points
Voronoi area is search range, is found in the Voronoi diagram corresponding to next level road network in initiating searches level road network
The Voronoi area intersected with the search range.Where judging whether the intersecting Voronoi area is main part
Voronoi area, if so, continuing to find in the Voronoi diagram corresponding to next level road network in current search level road network
The Voronoi area intersected with search range, where the intersecting Voronoi area searched out is not main part
Voronoi area or current search level road network have been lowest hierarchical level road network.At this time, it is corresponding in current search level road network
Shortest Path Searching is carried out in Voronoi diagram.
The above process is circulated, until the Voronoi area where source point O or target point D is corresponding for lowest hierarchical level road network
Voronoi area, or source point O or target point D have fallen on shortest paths.The section that each recursive calculation obtains is stored in time
Routing footpath data set.
3rd, the generation of optimal path
The orderly section collection based on road network level, i.e. path candidate data set can be obtained by step 2 and three.It is based on
Path candidate data set, first, removes the overlapping section in path candidate data set.Then, using the spatial relationship between section,
The intersection point in section before and after calculating.Then, two sections of head and the tail are trimmed, and removes unreasonable part on head and the tail section.Most
Afterwards, former and later two intersection points are sequentially connected, that is, have obtained real optimal path.
If in component obtained by step 2, source point O and/or target point D are not fallen within any section of component, also
It need to be operated:By neighbor analysis, find point immediate from source point and/or target point in head and the tail section and be used as optimal path
True source point and/or target point.
Fig. 7 is the rough schematic of searching route of the present invention, uses the lines of different thicknesses to represent different levels in figure
Road net data, two solid black points represent source point O and target point D respectively.First, searched for from high to low most according to road network level
The main part of shortest path, including the first estate section of heavy line, the of the second grade section of middle heavy line and fine line
Three grade sections.Then, the component of optimal path is searched for, for O points, when searching lowest hierarchical level road network, O points place
Voronoi area one is directly subordinate to the Voronoi area of main part, using subpoint of the O points on main part head sections as
One endpoint of optimal path.For D points, it can be seen that shortest path gradually passes through the second grade section from nearest main part
Approached with tertiary gradient section to D points, be eventually found D points another end of subpoint as optimal path on shortest paths
Point, has thus obtained final optimal path.
For the superiority of the verification present invention, using Wuhan City's road network as experimental data, choose four typical period of time and make
For research object, and compared with conventional method (level road network search algorithm is based in classical dijkstra's algorithm and ArcGIS)
Compared with.It should be noted that the present invention uses time of vehicle operation as weight, and road physical length is then utilized in conventional method
As weight.Therefore, journey time (T) and haul distance (D) are two parameters as benchmark.Meanwhile to wherein 12 groups with
Machine selection O/D points and its between optimal path done detailed comparison and analysis.
Shown in table 1 is the path analysis results contrast of three kinds of methods under four typical period of time, in terms of haul distance
From the point of view of, the method for the present invention is longer than the path that traditional dijkstra's algorithm is calculated by about 23%, but than in ArcGIS
Want short by 30% in the path being calculated based on level road network search algorithm.From the point of view of journey time, the method for the present invention is in four allusion quotations
The path difference that the path being calculated under the type period is obtained than other two methods it is fast by 41% and 21%, 34% and 10%,
39% and 19%, 39% and 19%.In addition it is also found that the hourage ratio method B spent by method C is less, because high
The road of grade is better utilized.These are the result is that encouraging, because it indicate that the present invention is exactly in method B
Between C, but also the section of traffic congestion is efficiently avoid, significantly reduce the travel time.It has been found that for
For a pair of O/D, the time consumption change gone on a journey under different periods is smaller, therefore selects period b as concrete analysis
Period, as shown in Figure 8, it is the optimal path of 12 groups of experiments in table 1, wherein light gray, Dark grey, black path represent respectively
The optimal path that method A, B, C are obtained.
The contrast of haul distance and journey time in 1 three kinds of experimental results of table
In table 2, D represents haul distance, and T represents journey time, and A represents the method for the present invention, and B represents traditional Dijkstra
Algorithm, C represent the searching algorithm based on level road network in ArcGIS, and B (+) represents increment of the method for the present invention relative to method B
Percentage, B (-) represent decrement percentage of the method for the present invention relative to method B, and C (-) represents the method for the present invention relative to method
The decrement percentage of C.
In the optimal path result that analysis is tried to achieve by the method for the present invention the service condition of roads at different levels (different stage road
Length accounting and section quantity accounting), as shown in Figure 9.It can be seen from the figure that two in the optimal path that the method for the present invention obtains
Level road network occupies 72%, this is almost twice of tradition Dijkstra methods, is the search based on level road network in ArcGIS
Four times of algorithm.And three-level road network occupies about 66% ratio in the path being calculated by traditional Dijkstra methods
Weight.Level-one road occupies about 78% proportion in searching algorithm based on level road network in ArcGIS.For section quantity
For, also obtain similar result.These experimental datas further demonstrate the method for the present invention between conventional method and
Between searching algorithm based on level road network in ArcGIS.Therefore, the method for the present invention can effectively prevent traditional algorithm deviation with from
On rudimentary road network by the drawbacks of, traffic congestion often occurs for the road network of these inferior grades.Meanwhile further through visiting by phased manner
Ask rudimentary road network, significantly reduce the length of distance to improve the reliability of the searching algorithm based on level road network.
In order to strengthen demonstration, have selected 1000 groups of O/D points on period b has carried out contrast experiment to the present invention again, for road
The length of line, the method for the present invention ratio method B will grow 28%, but ratio method C wants short by 11%.This is protected substantially with previous experiments result
Hold consistent.Meanwhile compared with method C, the method for the present invention is more sensitive to path length.For journey time, present invention side
The path that method obtains its journey time ratio method B and method C has lacked 26% and 9%, this also matches with experimental result above,
The stability of the method for the present invention result is also demonstrated at the same time.In addition as shown in Figure 10, the utilization power of road is contrasted substantially with before
The experimental result in face is consistent, and wherein secondary road occupies main proportion, and path total length reaches 63%, section total quantity
Reach 61%.And mainly occupy proportion in method B is three-level road network, mainly occupy proportion in method C is level-one road network.More
What is interesting is, it has been found that method C has the quality of data of road network certain requirement, result in and generates in this experiment
7.3% analysis failure.
The computational efficiency of three of the above method is compared and analyzed, below by gradual increased road network scale,
The variation tendency of measuring and calculation efficiency.
See Figure 11, (such as containing the road net data in 500 sections), the calculating of three kinds of methods when road network scale is smaller
Efficiency does not have too big difference.However, as network size increases, the computational efficiency of method B shows exponential growth, this is remote
Beyond other two methods and it cannot meet the needs of real-time route analysis.And the method for the present invention and method C road network scale compared with
When big, real-time navigation Service can be provided, embodied obvious advantage.Analysis result shows that the method for the present invention is than tradition side
Method is more efficient, and the method for the present invention and method C are maintained at similar in one in level, this is mainly due to the method for the present invention and adopts
The computational methods searched for the thoughtcast of spatial level and corresponding global and local, therefore by this method application
In on a large-scale road network carry out optimal path analysis be it is feasible be also it is effective.
In short, it is longer than the path that dijkstra's algorithm obtains using the path length that the method for the present invention is calculated, but
It is basic shorter than the path length by being obtained based on level road network search algorithm.But than both from journey time
It is short;Therefore, the method for the present invention is better than other two methods.
Claims (9)
1. the preferential optimal path computation method of level based on dynamic road network, it is characterized in that, including:
Step 1, road network is divided into the road network of K grade according to category of roads, the road network of K grade is layered, obtains K
The road network of a level, generates the corresponding Voronoi diagram of each level road network respectively;Wherein, the road network of kth level is by kth grade,
The road network of the 1st grade of k-1 grades ... is formed, and k=1,2 ... K, K values take the category of roads quantity of road network;
Step 2, route searching is carried out based on Voronoi diagram, this step includes the main part and branch of search optimal path
Point;
The main part of the search optimal path further comprises:
The 2.1 initiating searches level using linear search algorithm by source point O and target point D matching road networks, by initiating searches level
For the corresponding Voronoi diagram of road network as current Voronoi diagram, O and the V areas where D are denoted as starting point V areas and terminal V areas, V respectively
Area, that is, Voronoi area;
2.2 carry out Shortest Path Searching based on current Voronoi diagram;
2.3 sequentially check the connectedness that shortest path includes section, to disconnected front and rear two section, with front and rear two-way section institute
It is search range in the minimum enclosed rectangle of the union in V areas, performs sub-step 2.4;If being connected comprising all sections, knot
Beam, performs sub-step 2.5;
2.4 judge whether current search level road network is lowest hierarchical level road network, if so, based on all level roads in search range
Network data carries out Shortest Path Searching, and the path searched is saved in path candidate data set;Otherwise, by current search level
Voronoi diagram corresponding to next level road network of road network is as current Voronoi diagram, and traversal is current in search range
All V areas of Voronoi diagram, determine in current Voronoi diagram according to the spatial relationship between road section in current Voronoi diagram
Starting point V areas and terminal V areas, perform sub-step 2.2;
The component of the search optimal path further comprises:
2.5 current search level road networks be initialized as sub-step 2.1 matched initiating searches level road network next level road
Net;
2.6 are based on the corresponding Voronoi diagram of current search level road network, using the V areas where O or D as search range, are searched currently
The V areas intersected with the search range are found in rope level road network;
2.7 judge whether the intersecting V areas are V areas where main part, if so, next layer by current search level road network
Level road network performs sub-step 2.6 as current search level road network;Otherwise, it is corresponding based on current search level road network
Voronoi diagram, using the V areas where O as starting point V areas or using the V areas where D as terminal V areas, using V areas where main part as eventually
Point V areas or starting point V areas;
2.8 carry out Shortest Path Searching based on the corresponding Voronoi diagram of current search level road network;
2.9 using next level road network of current search level road network as current search level road network, based on current search level
The corresponding Voronoi diagram of road network, using the V areas where O as starting point V areas or using the V areas where D as terminal V areas, to be searched for most
V areas where short path are terminal V areas or starting point V areas, perform sub-step 2.8;
2.10 repetition sub-steps 2.8 ~ 2.9 are until it is lowest hierarchical level road that O or D, which drops down onto shortest path or current search level road network,
Net;
Step 3, the main part of optimal path is obtained according to step 2 and component generates optimal path.
2. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
The road network to K grade described in step 1 is layered, and is specially:
(1)Section is divided and merged using following rule:
The division principle is:
When a plurality of section intersects at an intersection point, and the intersection point is different for the endpoint in a plurality of section when, to a plurality of section into
Line splitting, when division, follow " inferior grade road network is interrupted by high class road network, and high class road network is not interrupted by inferior grade road network ";
The combination principle is:
To only having the two of an intersection point sections, if the intersection point is at the same time the endpoint in two sections, which is closed
And;
(2)Road network after being divided and merged based on section is layered, wherein, the road network of kth level is by kth grade, kth -1 etc.
The road network of the 1st grade of level ... is formed.
3. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
The starting level that search road network is determined using linear search method described in sub-step 2.1, is specially:
The line of source point O and target point D are extended to the 10% of line total length, that is, obtain searching operators, acquisition and searching operators
The level in intersecting all sections, highest level originate level.
4. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
In sub-step 2.1, if the V areas where O and D are identical, after section is saved in path candidate data set in the identical V areas,
Terminate, then perform sub-step 2.5.
5. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
In sub-step 2.2, using weight of the vehicle by history average time in each V areas as each V areas, based on current Voronoi diagram
Carry out Shortest Path Searching and obtain shortest path.
6. the preferential optimal path computation method of level as claimed in claim 5 based on dynamic road network, it is characterized in that:
The vehicle is obtained with the following method by the history average time in each V areas:
The sample track point in V areas is obtained, calculates the history average speeds of vehicle in V areas using sample track point, in V areas
The physical length divided by history average speeds in included section, i.e., the history average time that vehicle passes through V areas.
7. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
In sub-step 2.3, preceding a road section in two sections before and after connection is saved in path candidate data set;If the two of connection
Latter bar section is the last item section in bar section, then latter bar section is also saved in path candidate data set.
8. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
Step 3 is specially:
Main part and component institute are proceeded as follows comprising section:
(1)Remove overlapping section;
(2)The intersection point in section before and after being calculated using the spatial relationship between section;
(3)Two sections of head and the tail are trimmed, remove unreasonable part on head and the tail section;
(4)Former and later two intersection points are sequentially connected, up to optimal path.
9. the preferential optimal path computation method of level as claimed in claim 1 based on dynamic road network, it is characterized in that:
Step 3 further includes:
Find the nearest neighbor point of source point and/or target point on two sections of head and the tail, true source point as optimal path and/or
Target point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610966010.XA CN108009666B (en) | 2016-10-28 | 2016-10-28 | Hierarchical priority optimal path calculation method based on dynamic road network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610966010.XA CN108009666B (en) | 2016-10-28 | 2016-10-28 | Hierarchical priority optimal path calculation method based on dynamic road network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108009666A true CN108009666A (en) | 2018-05-08 |
CN108009666B CN108009666B (en) | 2020-04-10 |
Family
ID=62048375
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610966010.XA Active CN108009666B (en) | 2016-10-28 | 2016-10-28 | Hierarchical priority optimal path calculation method based on dynamic road network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108009666B (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109101743A (en) * | 2018-08-28 | 2018-12-28 | 武汉市众向科技有限公司 | A kind of construction method of high-precision road net model |
CN109166311A (en) * | 2018-08-15 | 2019-01-08 | 广州市康软信息科技有限公司 | A kind of region first aid monitoring method, system and device based on vehicle scheduling |
CN109506669A (en) * | 2018-12-28 | 2019-03-22 | 斑马网络技术有限公司 | Dynamic path planning method, device, system and storage medium |
CN110032381A (en) * | 2019-02-27 | 2019-07-19 | 丰图科技(深圳)有限公司 | Road condition update method, system and the storage medium of electronic map |
CN110567459A (en) * | 2018-06-05 | 2019-12-13 | 北京京东尚科信息技术有限公司 | Path planning method and device |
CN111027743A (en) * | 2019-11-06 | 2020-04-17 | 深圳市城市交通规划设计研究中心有限公司 | OD optimal path searching method and device based on hierarchical road network |
CN111047019A (en) * | 2018-10-12 | 2020-04-21 | 兰州交通大学 | Construction algorithm of network weighted Voronoi graph |
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 |
CN112882466A (en) * | 2021-01-12 | 2021-06-01 | 上海电力大学 | Fusion hierarchical planning and A*Shared electric vehicle path planning method of algorithm |
CN113204348A (en) * | 2021-04-30 | 2021-08-03 | 北京连山科技股份有限公司 | Domestic road network data compiling method based on linkage hierarchy |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1988362A1 (en) * | 2007-05-04 | 2008-11-05 | Harman Becker Automotive Systems GmbH | Route determination method and device |
CN102081658A (en) * | 2011-01-13 | 2011-06-01 | 北京超图软件股份有限公司 | Hierarchical road network-based path search method and device |
CN103047990A (en) * | 2012-12-24 | 2013-04-17 | 北京交通发展研究中心 | Multi-path selection method based on hierarchical backbone network |
-
2016
- 2016-10-28 CN CN201610966010.XA patent/CN108009666B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1988362A1 (en) * | 2007-05-04 | 2008-11-05 | Harman Becker Automotive Systems GmbH | Route determination method and device |
CN102081658A (en) * | 2011-01-13 | 2011-06-01 | 北京超图软件股份有限公司 | Hierarchical road network-based path search method and device |
CN103047990A (en) * | 2012-12-24 | 2013-04-17 | 北京交通发展研究中心 | Multi-path selection method based on hierarchical backbone network |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110567459B (en) * | 2018-06-05 | 2021-09-07 | 北京京东乾石科技有限公司 | Path planning method and device |
CN110567459A (en) * | 2018-06-05 | 2019-12-13 | 北京京东尚科信息技术有限公司 | Path planning method and device |
CN109166311A (en) * | 2018-08-15 | 2019-01-08 | 广州市康软信息科技有限公司 | A kind of region first aid monitoring method, system and device based on vehicle scheduling |
CN109101743B (en) * | 2018-08-28 | 2023-01-17 | 武汉市众向科技有限公司 | Method for constructing high-precision road network model |
CN109101743A (en) * | 2018-08-28 | 2018-12-28 | 武汉市众向科技有限公司 | A kind of construction method of high-precision road net model |
CN111047019B (en) * | 2018-10-12 | 2022-09-23 | 兰州交通大学 | Network weighting Voronoi diagram construction method based on improved pulse coupling neural network |
CN111047019A (en) * | 2018-10-12 | 2020-04-21 | 兰州交通大学 | Construction algorithm of network weighted Voronoi graph |
CN109506669A (en) * | 2018-12-28 | 2019-03-22 | 斑马网络技术有限公司 | Dynamic path planning method, device, system and storage medium |
CN110032381A (en) * | 2019-02-27 | 2019-07-19 | 丰图科技(深圳)有限公司 | Road condition update method, system and the storage medium of electronic map |
CN110032381B (en) * | 2019-02-27 | 2022-10-04 | 丰图科技(深圳)有限公司 | Method, system and storage medium for updating road state of electronic map |
CN111027743A (en) * | 2019-11-06 | 2020-04-17 | 深圳市城市交通规划设计研究中心有限公司 | OD optimal path searching method and device based on hierarchical road network |
CN111027743B (en) * | 2019-11-06 | 2022-11-25 | 深圳市城市交通规划设计研究中心有限公司 | OD optimal path searching method and device based on hierarchical road network |
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 |
CN111879329B (en) * | 2020-07-30 | 2022-02-22 | 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 |
CN112882466A (en) * | 2021-01-12 | 2021-06-01 | 上海电力大学 | Fusion hierarchical planning and A*Shared electric vehicle path planning method of algorithm |
CN113204348A (en) * | 2021-04-30 | 2021-08-03 | 北京连山科技股份有限公司 | Domestic road network data compiling method based on linkage hierarchy |
Also Published As
Publication number | Publication date |
---|---|
CN108009666B (en) | 2020-04-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108009666A (en) | The preferential optimal path computation method of level based on dynamic road network | |
Lang et al. | The new metropolis: Rethinking megalopolis | |
CN102538806B (en) | A kind of paths planning method and relevant device | |
CN105825672B (en) | A kind of city guide method for extracting region based on floating car data | |
CN109959388B (en) | Intelligent traffic refined path planning method based on grid expansion model | |
Yang et al. | Multi-objective highway alignment optimization incorporating preference information | |
CN102081658B (en) | Hierarchical road network-based path search method and device | |
CN102788584B (en) | The energy consumption prediction unit of road grade data generating device and generation method, vehicle console device and vehicle | |
CN108177648A (en) | A kind of energy management method of the plug-in hybrid vehicle based on intelligent predicting | |
CN106225797A (en) | A kind of paths planning method | |
CN103221785B (en) | For trying to achieve the method for residue flying power | |
CN113155145B (en) | Lane-level path planning method for automatic driving lane-level navigation | |
CN101614551A (en) | Sneakernet method for building up and device, method for searching path and device | |
CN112734176B (en) | Charging station building method and device, terminal equipment and readable storage medium | |
CN102435200A (en) | Rapid path planning method | |
CN104731963A (en) | Grid path recommending method and system based on internet of vehicle | |
CN102607582A (en) | Method for planning guidance route by means of using city access points as guideposts and system | |
CN105841709A (en) | Method for planning car driving path | |
CN106373384B (en) | Outlying district regular bus circuit Real-time Generation | |
Song et al. | Exploring pareto routes in multi-criteria urban bicycle routing | |
CN111879329A (en) | Customized public transport passable shortest path calculation method based on A-x algorithm | |
Ji-hua et al. | A hierarchical path planning method using the experience of taxi drivers | |
CN109800903A (en) | A kind of profit route planning method based on taxi track data | |
CN109166307B (en) | Visual metaphor expression method and system for time-space accessibility of traffic network | |
KR101270375B1 (en) | A method of analyzing optimal paths across a continuous landscape by considering an existing road network |
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 |