A route planning method for vehicle road testing based on genetic evolution

Abstract

This invention discloses a route planning method for vehicle road testing based on genetic evolution, comprising the following steps: (1) collecting road network information, abstracting one-way roads in the current road network into a directed edge, and two-way roads into two directed edges, and then abstracting the traffic information in the road network into a directed graph G; (2) finding a path p in the directed graph G from a specified starting point to a specified ending point, such that path p effectively covers the edge set of graph G and the path length is minimized; wherein, path p is any one of the edges of a one-way road and any one of the two edges of a two-way road; (3) converting path p into the actual driving path of the navigation vehicle; This invention can automatically calculate according to the road network situation, which means that different cities can share a set of route planning systems, and the algorithm execution does not require manual intervention, thus greatly reducing labor costs; in addition, the evolutionary mechanism inside the genetic algorithm ensures the quality of the solution to a certain extent.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a route planning method for vehicle road testing based on genetic evolution. Background Technology

[0002] The telecommunications industry commonly uses road tests to check the coverage of urban wireless networks. Testers typically operate specialized equipment from inside a car, and as the vehicle cruises through the city, they test the network coverage. To save time and improve efficiency, the vehicle's route needs to be scientifically planned. However, in current road tests, routes are often determined by drivers based on experience, making it difficult to guarantee the scientific accuracy of the route design. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to provide a route planning method for vehicle road testing based on genetic evolution, which can find the route with the shortest total distance by traversing every road, provided that the road network is passable and that any two points in the road network are reachable from each other.

[0004] Technical solution: This invention relates to a route planning method for vehicle road testing based on genetic evolution, comprising the following steps: (1) Collect road network information, abstract one-way roads in the current road network into one directed edge and two-way roads into two directed edges, and then abstract the traffic information in the road network into a directed graph. ; (2) Find the directed graph The path from a specified starting point to a specified ending point. , make the path The edge set of graph G is effectively covered while minimizing the path length; where effective coverage means covering every edge representing a one-way road and any one of the two edges representing each two-way road. (3) Path This is converted into the actual driving route of the navigation vehicle.

[0005] Furthermore, step (2) includes the following steps: (21) Encoding Conversion: Converting path search to... In vector space; (22) Generate a population: pair of edge sets A subset is initially selected; where, The bidirectional road is abstracted into an ordered set of directed edges; (23) Evolutionary population: Edge set The optimal selection is made from a subset of the subset, where... The bidirectional road is abstracted into an ordered set of directed edges.

[0006] Furthermore, step (21) includes the following steps: (211) Given a directed graph The specified starting point is The specified endpoint location is ; in the figure Add a line from the endpoint To the starting point An infinitely long virtual edge is denoted as . ; (212) Suppose that the current road network has Add directions and directed edges to each two-way road to create two-way directed edges, for a total of A set of directed edges, abstracted from bidirectional roads, is given by: Meanwhile, the ordered set of directed edges, abstracted from unidirectional roads, is called a set. Then the diagram The edge set is a set .

[0007] (213) Define the length as A 0-1 vector, where the vector elements take values ​​of This indicates selecting a specific edge, with a value of [value]. If we indicate that a particular edge is not selected, then a 0-1 vector can represent the edge set. Subset selection method; Furthermore, step (22) specifically involves: randomly generating... A length of of Vector, to obtain A population composed of individuals , in, In Each element is randomly assigned a value. or , representing edge set A subset selection scheme; corresponding Let the edge set in be .

[0008] Furthermore, step (23) includes the following steps: (231) Initialization: Initialization The fitness of the global optimal solution is infinite; among which, This indicates the preset time rounds for population evolution; (232) Correct the current solution: For each individual in the population Make corrections to obtain a new one. ; (233) Calculate fitness: for each individual in the population Calculate its fitness; (234) Update the optimal solution: find the individual with the lowest fitness in the population. ;like If the fitness of the solution is less than the current global optimum, then use... Replace the current global optimal solution; otherwise, retain the original optimal solution. (235) Population evolution: the population with higher fitness Each individual is replaced with the current global optimal solution; for each individual in the population, a position is randomly selected for... Conversion, Change to , Change to ; (236) Termination condition determination: Let time Determine the time If the preset time of 2000 is reached, return the path corresponding to the current global optimal solution; otherwise, repeat steps (232)-(235).

[0009] Furthermore, step (232) includes the following steps: (2321) in In the middle, edges are deleted in random order until the result is achieved. The edges lead to a strongly connected subgraph, resulting in the subgraph. ; (2322) From Add edges to the middle in a random order. In, until it makes It is a picture An effective cover of the edge set is obtained, and the edges lead to a strongly connected subgraph, thus yielding the subgraph. ; (2323) will The corresponding edge set The subset selection method is encoded and converted into a 0-1 vector. .

[0010] Furthermore, step (233) includes the following steps: (2331) Update edge sets until it makes The edges are on the graph The effective edge coverage, and the edges lead to strongly connected subgraphs, yielding the subgraph. ; (2332) List the diagram. Each vertex Find the in-degree and out-degree, and calculate the difference between the in-degree and out-degree. ; (2333) Set of points and For an empty set, Repeating points times, join the point set Middle; General Repeating points times, join the point set middle; (2334) Construct a weighted complete bipartite graph ; (2335) Define a set any pair of points in The distance between them is as described above in the original image. The shortest path length in; (2336) Find the graph using the least weight perfect matching method. Matching; (2337) Add the shortest path between matching pairs of points to the graph. Then delete the virtual edges. , to obtain the diagram ; (2338) Find the diagram using the one-stroke drawing method. Zhong Cong arrive The Eulerian path, and the length of this path as vector The degree of adaptability.

[0011] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: Unlike the traditional route planning method that relies on the driver's personal experience, the present invention can automatically calculate according to the road network conditions, which means that different cities can share a set of route planning systems, and the algorithm does not require human intervention, which can greatly reduce labor costs; in addition, the evolutionary mechanism inside the genetic algorithm ensures the quality of the solution to a certain extent. Attached Figure Description

[0012] Figure 1 This is a schematic diagram illustrating the principle of the present invention; Figure 2 This is a flowchart illustrating the evolutionary process involved in this invention. Figure 3 The flowchart for correcting the current solution in this invention Figure 4 This is a flowchart of the fitness calculation method in this invention. Figure 5 This invention provides an example of a road network that includes both two-way and one-way roads. Figure 6 This invention is by Figure 5The directed graph transformed from the road network shown. Figure 7 This invention is by Figure 6 A perfect matching of a bipartite graph obtained from points where "out-degree is not equal to in-degree". Figure 8 For the present invention Figure 7 Add perfect match to Figure 6 The resulting directed graph. Detailed Implementation

[0013] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0014] like Figure 1 As shown, this embodiment of the invention provides a method for road test route planning under real-time traffic conditions, including the following steps: (1) Collect road network information, abstract one-way roads in the current road network into one directed edge and two-way roads into two directed edges, and then abstract the traffic information in the road network into a directed graph. ; (2) Find the directed graph The path from a specified starting point to a specified ending point. , make the path The goal is to achieve an effective coverage of the edge set of graph G while minimizing the path length; where effective coverage means covering every edge representing a one-way road and any one of the two edges representing each two-way road; for example... Figure 2 As shown, it includes the following steps: (21) Encoding Conversion: Converting path search to... In a vector space; including the following steps: (211) Given a directed graph The specified starting point is The specified endpoint location is ; in the figure Add a line from the endpoint To the starting point An infinitely long virtual edge is denoted as . ; (212) Suppose that the current road network has Add directions and directed edges to each two-way road to create two-way directed edges, for a total of A set of directed edges, abstracted from bidirectional roads, is given by: Meanwhile, the ordered set of directed edges, abstracted from unidirectional roads, is called a set. Then the diagram The edge set is a set .

[0015] (213) Define the length as A 0-1 vector, where the vector elements take values ​​of This indicates selecting a specific edge, with a value of [value]. If we indicate that a particular edge is not selected, then a 0-1 vector can represent the edge set. Subset selection method; (22) Generate a population: pair of edge sets A subset is initially selected; where, The bidirectional road is abstracted into an ordered set of directed edges; specifically, it is randomly generated. A length of of Vector, to obtain A population composed of individuals , in, In Each element is randomly assigned a value. or , representing edge set A subset selection scheme, which will corresponding Let the edge set in be .

[0016] (23) Evolutionary population: Edge set The optimal selection is made from a subset of the subset, where... Representing a bidirectional road as an abstraction yields an ordered set of directed edges; this includes the following steps: (231) Initialization: Initialization The fitness of the global optimal solution is infinite; among which, This indicates the preset time rounds for population evolution; (232) Correct the current solution: For each individual in the population Make corrections to obtain a new one. ,Will Vector to graph Effective coverage, such as Figure 3 As shown, it includes the following steps: (2321) in In the middle, edges are deleted in random order until the result is achieved. The edges lead to a strongly connected subgraph, resulting in the subgraph. ; (2322) From Add edges to the middle in a random order. In, until it makes It is a picture An effective cover of the edge set is obtained, and the edges lead to a strongly connected subgraph, thus yielding the subgraph. ; (2323) will The corresponding edge set The subset selection method is encoded and converted into a 0-1 vector. .

[0017] (233) Calculate fitness: for each individual in the population Calculating its fitness is equivalent to calculating... The graph corresponding to the vector The length of the effective coverage path, such as Figure 4 As shown, it includes the following steps: (2331) Update edge sets until it makes The edges are on the graph The effective edge coverage, and the edges lead to strongly connected subgraphs, yielding the subgraph. ; (2332) List the diagram. Each vertex Find the in-degree and out-degree, and calculate the difference between the in-degree and out-degree. ; (2333) Set of points and For an empty set, Repeating points times, join the point set Middle; General Repeating points times, join the point set middle; (2334) Construct a weighted complete bipartite graph ; (2335) Define a set any pair of points in The distance between them is as described above in the original image. The shortest path length in; (2336) Use the least weight perfect matching method (e.g., Edmonds & Johnson 1973) to find the graph Matching; (2337) Add the shortest path between matching pairs of points to the graph. Then delete the virtual edges. , to obtain the diagram ; (2338) Use a one-stroke drawing method (such as Fleury 1883) to find the figure. Zhong Cong arrive The Eulerian path, and the length of this path as vector The degree of adaptability.

[0018] (234) Update the optimal solution: find the individual with the lowest fitness in the population. ;like If the fitness of the solution is less than the current global optimum, then use... Replace the current global optimal solution; otherwise, retain the original optimal solution. (235) Population evolution: the population with higher fitness Each individual is replaced with the current global optimal solution; for each individual in the population, a position is randomly selected for... Conversion, Change to , Change to ; (236) Termination condition determination: Let time Determine the time If the preset time of 2000 is reached, return the path corresponding to the current global optimal solution; otherwise, repeat steps (232)-(235).

[0019] (3) Path This is converted into the actual driving route of the navigation vehicle.

[0020] The actual effects of the present invention are as follows Figures 5-8 ,by Figure 5 Taking a simple road network as an example, it contains seven nodes: s, t, v_1, ..., v_5. The road from v_2 to v_5 is a bidirectional road, and the remaining paths are unidirectional. We need to design the shortest path from s to t to traverse all the roads in the network. First, we abstract bidirectional roads as two directed edges and unidirectional roads as one directed edge. Then, we add a virtual edge from the endpoint t to the initial point s, which gives us the following: Figure 6 The directed graph shown; then, Figure 6 Construct a bipartite graph from nodes with "out-degree greater than in-degree" and "out-degree less than in-degree" (assuming both directed edges between v_2 and v_5 need to be traversed). Use the minimum weight perfect matching method (e.g., Edmonds & Johnson 1973) to find the matching of this bipartite graph, such as... Figure 7 As shown; next, Figure 7 Add the path corresponding to the perfect match shown to Figure 6 Then, by deleting the virtual edge from the endpoint t to the initial point s, we can obtain... Figure 8 Finally, using a one-stroke drawing algorithm (such as Fleury 1883), it is easy to obtain... Figure 8 The Euler path, that is Figure 5 The traversal path is: .

Claims

1. A route planning method for vehicle road testing based on genetic evolution, characterized in that, Includes the following steps: (1) Collect road network information, abstract one-way roads in the current road network into one directed edge and two-way roads into two directed edges, and then abstract the traffic information in the road network into a directed graph. ; (2) Find the directed graph The path from a specified starting point to a specified ending point. , make the path To effectively cover the edge set of graph G while minimizing path length; where effective coverage means covering every edge representing a one-way road and any one of the two edges representing each two-way road; this includes the following steps: (21) Encoding Conversion: Converting path search to... In vector space; (22) Generate a population: pair of edge sets A subset is initially selected; where, The bidirectional road is abstracted into an ordered set of directed edges; (23) Evolutionary population: Edge set The optimal selection is made from a subset of the subset, where... Representing a bidirectional road as an abstraction yields an ordered set of directed edges; this includes the following steps: (231) Initialization: Initialization The fitness of the global optimal solution is infinite; among which, This indicates the preset time rounds for population evolution; (232) Correct the current solution: For each individual in the population Make corrections to obtain a new one. ; (233) Calculate fitness: for each individual in the population Calculate its fitness; (234) Update the optimal solution: find the individual with the lowest fitness in the population. ;like If the fitness of the solution is less than the current global optimum, then use... Replace the current global optimal solution; otherwise, retain the original optimal solution. (235) Population evolution: the population with higher fitness Each individual is replaced with the current global optimal solution; for each individual in the population, a position is randomly selected for... Conversion, Change to ,1 replaced with ; (236) Termination condition determination: Let time Determine the time If the preset time of 2000 is reached, return the path corresponding to the current global optimal solution; otherwise, repeat steps (232)-(235). (3) Path This is converted into the actual driving route of the navigation vehicle.

2. The method for route planning in vehicle road testing based on genetic evolution as described in claim 1, characterized in that, Step (21) includes the following steps: (211) Given a directed graph The specified starting point is The specified endpoint location is ; in the figure Add a line from the endpoint To the starting point An infinitely long virtual edge is denoted as . ; (212) Suppose that the current road network has Add directions and directed edges to each two-way road to create two-way directed edges, for a total of A set of directed edges, abstracted from bidirectional roads, is given by: Meanwhile, the ordered set of directed edges, abstracted from unidirectional roads, is called a set. Then the diagram The edge set is a set ; (213) Define the length as A 0-1 vector, where the vector elements take values ​​of This indicates selecting a specific edge, with a value of [value]. If we indicate that a particular edge is not selected, then a 0-1 vector can represent the edge set. The subset selection method.

3. The method for route planning based on genetic evolution for vehicle road testing according to claim 1, characterized in that, The specific steps (22) are as follows: randomly generate A length of of Vector, to obtain A population composed of individuals ,in, In Each element is randomly assigned a value. or , representing edge set A subset selection scheme; An even positive number represents the population size; corresponding Let the edge set in be .

4. The method for route planning in vehicle road testing based on genetic evolution as described in claim 1, characterized in that, Step (232) includes the following steps: (2321) in In the middle, edges are deleted in random order until the result is achieved. The edges lead to a strongly connected subgraph, resulting in the subgraph. ; (2322) From Add edges to the middle in a random order. In, until it makes It is a picture An effective cover of the edge set is obtained, and the edges lead to a strongly connected subgraph, thus yielding the subgraph. ; (2323) will The corresponding edge set The subset selection method is encoded and converted into a 0-1 vector. .

5. The method for route planning in vehicle road testing based on genetic evolution according to claim 1, characterized in that, The step (233) includes the following steps: (2331) Update edge sets until it makes The edges are on the graph The effective edge coverage, and the edges lead to strongly connected subgraphs, yielding the subgraph. ; (2332) List the diagram. Each vertex Find the in-degree and out-degree, and calculate the difference between the in-degree and out-degree. ; (2333) Set of points and For an empty set, Repeating points times, join the point set Middle; General Repeating points times, join the point set middle; (2334) Construct a weighted complete bipartite graph ; (2335) Define a set any pair of points in The distance between them is as described above in the original image. The shortest path length in; (2336) Find the graph using the least weight perfect matching method. Matching; (2337) Add the shortest path between matching pairs of points to the graph. Then delete the virtual edges. , to obtain the diagram ; (2338) Find the diagram using the one-stroke drawing method. Zhong Cong arrive The Eulerian path, and the length of this path as vector The degree of adaptability.

Images