A global path planning method for unmanned transport vehicles in underground mines

Abstract

The present invention provides a global path planning method for unmanned transport vehicles in underground mines, and the steps are as follows: the high-precision map file of the underground mine is parsed according to the protocol and stored in the database; the high-precision map sections are used as nodes and the connectivity relationship of the sections is used as edges to construct a directed topological graph of the high-precision map; the entry section and the exit section of the turning point of the high-precision map are connected to update the directed topological graph; based on the directed topological graph of the high-precision map, path planning is performed; multiple paths planned are screened to obtain the optimal path. With the help of high-precision maps, this method can realize the global path planning of unmanned transport vehicles considering the turning point in the underground mine environment, effectively solve the problem that ordinary path planning cannot select the optimal path according to factors such as travel time, and also solve the problem that the path planning result is invalid when the starting driving direction and the end point driving direction are inconsistent, which can effectively improve the driving safety and transportation efficiency of unmanned transport vehicles in underground mines.

Description

Technical Field

[0001] The present invention mainly relates to the technical field related to path planning of unmanned vehicles, and specifically to a global path planning method for unmanned transport vehicles in underground mines. Background Art

[0002] In the field of autonomous driving, path planning algorithms have made great progress. The main existing methods include graph-based path planning methods and model-based path planning. The principle of graph-based path planning methods is to find the shortest path from point A to point B on the map, such as the A* algorithm and the Dijkstra algorithm, which are widely used in the field of autonomous driving. Model-based path planning methods use models to predict vehicle behavior and optimize an objective function.

[0003] In the prior art, graph-based path planning methods are already quite mature, but they may encounter difficulties in complex environments because they usually assume that the map is static, while in fact the environment is dynamically changing. In addition, the graph-based path planning algorithm solves the shortest path from point to point. In the field of unmanned driving, the use scenarios of mines are usually not so simple. For example, the driving direction of the unmanned transport vehicle is forward from the starting point to the unloading point. If the unloading point requires the transport vehicle to reverse into the unloading point, then simply performing path planning will not work. It is not just about finding the optimal path, but also considering which path can change the driving direction of the vehicle. At this time, the turning point is involved. Therefore, the solution method that maximizes efficiency is more important. The shortest path may not necessarily maximize efficiency. In addition, when the driving directions of the starting point and the end point are inconsistent, the ordinary graph-based path planning algorithm will fail.

[0004] In the existing technology, model-based path planning methods can handle some dynamic environments, but they require accurate models and a large amount of calculations, have high requirements for equipment, consume a lot of resources, and also have some uncertain factors. They are not applicable in scenarios such as unmanned driving in underground mines, which have strict safety requirements. Summary of the invention

[0005] In order to solve the shortcomings of the current technology, the present invention combines the existing technology and, based on practical applications, provides a global path planning method for unmanned transport vehicles in underground mines. With the help of high-precision maps, the method can realize global path planning of unmanned transport vehicles taking into account turning points in underground mine environments, effectively solving the problem that ordinary path planning cannot select the optimal path according to factors such as travel time, and also solves the problem that the path planning result is invalid when the starting driving direction and the ending point driving direction are inconsistent, which can effectively improve the driving safety and transportation efficiency of unmanned transport vehicles in underground mines.

[0006] The technical solution of the present invention is as follows:

[0007] A global path planning method for an unmanned transport vehicle in an underground mine comprises the following steps:

[0008] S1. Analyze the high-precision map file of the underground mine according to the protocol and store it in the database;

[0009] S2. Construct a directed topological graph of the high-precision map, using the high-precision map sections as nodes and the connectivity relationships of the sections as edges;

[0010] S3, connecting the incoming section and the outgoing section of the U-turn point on the high-precision map, and updating the directed topological map;

[0011] S4. Path planning based on the directed topological graph of the high-precision map;

[0012] S5. Screen multiple paths planned to obtain the optimal path.

[0013] Furthermore, in step S1, according to the high-precision map annotation protocol, each field of the high-precision map is parsed one by one and stored in a database. The database includes a high-precision map point information table, a high-precision map line string information table, a high-precision map section information table, a high-precision map relationship information table, a high-precision map poi point information table, and a high-precision map topology information table.

[0014] Further, step S2 specifically includes:

[0015] First, obtain all the road segment IDs of the high-precision map from the database, and then obtain the four vertices of the road segment according to the road segment ID. The four vertices are the lower left corner, lower right corner, upper left corner, and upper right corner;

[0016] Loop through all road sections. If the upper left corner and upper right corner of section a are the same as the lower left corner and lower right corner of section b, it means that section a points to section b. Then, establish an edge between vertex a and vertex b in the directed topology graph of the high-precision map.

[0017] Furthermore, all edge weights of the directed topology graph of the high-precision map are set to 1.

[0018] The further step S3 is specifically:

[0019] Get the entry and exit section IDs of all U-turn points in the high-precision map, establish a connection relationship between the vertices represented by the entry and exit sections, and update the directed topology graph.

[0020] Further step S4 specifically includes:

[0021] Get the starting and ending sections of the vehicle, determine the vehicle's driving direction on the starting section and the vehicle's driving direction on the ending section, use the starting and ending sections as the starting and ending points of the directed topology graph of the high-precision map, perform path planning based on the A* algorithm, and obtain all available paths for the vehicle from the starting section to the ending section.

[0022] Further, based on the starting coordinate point of the vehicle, determine which road sections the vehicle is in. If there are none, use the road section closest to this point as the starting road section of the vehicle;

[0023] If the difference between the vehicle's starting section direction angle and the vehicle's heading angle at the starting point is less than 90 degrees or greater than 270 degrees, it means that the vehicle is about to travel in the forward direction along the section. Otherwise, it means that the vehicle needs to travel in the reverse direction along the section. The section direction is calculated by finding the point on the center line string of the section closest to the vehicle's position coordinates, and then finding the next point on the center line string to the point, and constructing a vector through these two points. The direction of the vector is the direction of the section.

[0024] The method for obtaining the vehicle's end section is the same as the starting section.

[0025] Further step S5 specifically includes:

[0026] First, the paths are screened based on the driving direction, and then based on the principle of shortest travel time.

[0027] Further path screening based on driving direction includes:

[0028] First, find out how many U-turn points the path passes through. If the starting driving direction is equal to the ending driving direction and the number of U-turn points is an even number, the path is retained; if the starting driving direction is equal to the ending driving direction and the number of U-turn points is an odd number, the path is deleted; if the starting driving direction is not equal to the ending driving direction and the number of U-turn points is an odd number, the path is retained; if the starting driving direction is not equal to the ending driving direction and the number of U-turn points is an even number, the path is deleted.

[0029] Further path screening based on the principle of shortest travel time includes:

[0030] Calculate the forward driving distance and reverse driving distance in the path, and then obtain the vehicle's travel time on the path based on the vehicle's forward driving speed and reverse driving speed. Select the path with the shortest travel time as the optimal path for this path planning.

[0031] Beneficial effects of the present invention:

[0032] The present invention uses high-precision maps for path planning, abstracts the road sections of the high-precision map into nodes of a topological map, abstracts the connectivity of the road sections of the high-precision map into edges of the topological map, and abstracts the high-precision map into a directed topological map. Based on the directed topological map constructed by the high-precision map, the path with the shortest travel time is taken as the optimal path, rather than the shortest path. Based on the directed topological map constructed by the high-precision map, the driving direction of the starting point and the driving direction of the end point are combined to determine the optimal passable path;

[0033] With the help of high-precision maps, this method can realize global path planning of unmanned transport vehicles taking into account turning points in underground mining environments, effectively solving the problem that ordinary path planning cannot select the optimal path based on factors such as travel time, and also solves the problem that the path planning result is invalid when the starting driving direction and the ending point driving direction are inconsistent. It can effectively improve the driving safety and transportation efficiency of unmanned transport vehicles in underground mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the overall process of the present invention. DETAILED DESCRIPTION

[0035] The present invention will be further described with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art may make various changes or modifications to the present invention, and these equivalent forms also fall within the scope limited by the application.

[0036] This embodiment provides a global path planning method for unmanned transport vehicles in underground mining, specifically a global path planning algorithm for unmanned transport vehicles in underground mining that considers turning points. The road sections of the high-precision map are abstracted into nodes of the topological map, and the connectivity attributes of the road sections are used as the edges of the topological map. All road sections of the high-precision map and the connectivity information of the road sections can be abstracted into a directed graph, which can represent the driving trajectory of the unmanned transport vehicle. The A* algorithm can be used to obtain the shortest path from section M to section N. However, the algorithm does not seek the shortest path, but solves the optimal path, that is, the path with the shortest travel time for the unmanned transport vehicle from section M to section N. In addition, the algorithm also considers the driving direction of the unmanned transport vehicle, driving forward from the starting point to the unloading point. If the unloading point requires the transport vehicle to reverse into the vehicle, then simply performing path planning is not feasible. It is not just about finding the optimal path, but also considering which path can change the driving direction of the vehicle. At this time, the turning point is involved. With the help of high-precision maps, this method can realize global path planning of unmanned transport vehicles taking into account turning points in underground mining environments, effectively solving the problem that ordinary path planning cannot select the optimal path based on factors such as travel time, and also solves the problem that the path planning result is invalid when the starting driving direction and the ending point driving direction are inconsistent. It can effectively improve the driving safety and transportation efficiency of unmanned transport vehicles in underground mining.

[0037] In this embodiment, the high-precision map is used as the road data source. First, the high-precision map is made to generate an osm file, and then the high-precision map file is parsed to parse the points, lines, and relationships in the map and store them in a designed database for use by the algorithm. There are many advantages to using high-precision maps as data support for the algorithm. First, the accuracy of high-precision maps can reach the centimeter level, which has higher accuracy for unmanned transport vehicles in underground mines. Since the environment of underground mines is usually narrow, accurate maps can make unmanned transport vehicles safer and more reliable. Second, high-precision maps have rich expressive capabilities and can represent various road conditions, which has higher safety for unmanned transport vehicles in underground mines.

[0038] The path planning method of this embodiment mainly includes five parts: parsing high-precision maps, constructing topology maps, connecting incoming sections and outgoing sections, path planning, and optimal path screening. The specific steps are as follows.

[0039] Step 1: Parse the high-precision map file according to the protocol and store it in the database, specifically:

[0040] According to the HD map annotation protocol, each field of the HD map is parsed one by one and stored in the database. The database contains the HD map point information table, HD map line string information table, HD map road section information table, HD map relationship information table, HD map POI point information table, and HD map topology information table.

[0041] Step 2: Use the HD map sections as nodes and the connectivity of the sections as edges to construct a directed topological graph of the HD map. Specifically:

[0042] The road sections of the HD map are used as the vertices of the topological map, and the connectivity of the HD map sections is used as the edges of the topological map. The specific method is to first obtain the IDs of all road sections of the HD map from the database, and then obtain the four vertices of the road section (lower left corner, lower right corner, upper left corner, and upper right corner) according to the road section ID, and loop through all road sections. If the upper left corner and upper right corner of section a are the same as the lower left corner and lower right corner of section b, it means that section a points to section b, and then establish the edge between vertex a and vertex b in the HD map directed topological map. Since the algorithm does not focus on the shortest path, the weights of all edges in the HD map directed topological map are set to 1.

[0043] Step 3: Connect the incoming section and the outgoing section of the U-turn point on the HD map and update the directed topology map. Specifically:

[0044] In high-precision maps, the incoming and outgoing sections of a U-turn point are often not connected. This results in the inability to plan the directed topology of the high-precision map from the incoming section of the U-turn point to the outgoing section. Therefore, when generating a directed topology, it is necessary to obtain the incoming section ID and the outgoing section ID of all U-turn points in the map, and establish a connection relationship between the vertices representing the incoming section and the outgoing section.

[0045] Step 4: Based on the directed topology of the high-precision map, perform path planning, specifically:

[0046] 1) The starting coordinate point of the vehicle, the heading angle of the vehicle at the starting point, the ending coordinate point of the vehicle, and the driving direction of the vehicle at the ending point are known.

[0047] 2) Determine which road sections the vehicle is on based on the vehicle's starting coordinate point. If there is no road section, find the road section closest to this point as the vehicle's starting section; the same is true for obtaining the ending section.

[0048] 3) Determine the direction of the vehicle on the starting section: If the difference between the vehicle's starting section direction angle and the vehicle's heading angle at the starting point is less than 90 degrees or greater than 270 degrees, it means that the vehicle is about to travel in the forward direction along the section; if the difference between the starting point section direction and the vehicle's heading angle at the starting point is greater than 90 degrees and less than 270 degrees, it means that the vehicle needs to travel in the reverse direction along the section. The method for calculating the section direction is to find the point on the center line string of the section closest to the vehicle's position coordinates, and then find the next point on the center line string to that point. A vector can be constructed through these two points, and the direction of the vector is the direction of the section. The same is true for the terminal section.

[0049] 4) At this time, the starting section of the vehicle, the driving direction of the vehicle on the starting section, the ending section of the vehicle, and the driving direction of the vehicle on the ending section have been determined.

[0050] 5) The starting and ending sections are used as the starting and ending points of the directed topology of the high-precision map, and the path planning is performed based on the A* algorithm. Since the A* algorithm has been proven to be very efficient in both theory and practical applications, it is modified based on this algorithm, from obtaining the shortest path to obtaining all the paths.

[0051] Step 5: Filter multiple paths planned to obtain the optimal path, specifically:

[0052] In step 4, all traversable paths of the vehicle from the starting section to the ending section have been obtained.

[0053] 1) Path screening based on driving direction:

[0054] a. Find the number of U-turn points that the path passes through;

[0055] b. If the starting driving direction is equal to the ending driving direction and the number of U-turn points is an even number, the path is retained; if the starting driving direction is equal to the ending driving direction and the number of U-turn points is an odd number, the path is deleted; if the starting driving direction is not equal to the ending driving direction and the number of U-turn points is an odd number, the path is retained; if the starting driving direction is not equal to the ending driving direction and the number of U-turn points is an even number, the path is deleted.

[0056] 2) Path selection based on the principle of shortest travel time:

[0057] a. Calculate the forward driving distance and reverse driving distance in the path, and then obtain the vehicle's travel time on the path based on the forward driving speed and reverse driving speed of the vehicle;

[0058] b. Select the path with the shortest travel time as the optimal path for this path planning.

[0059] The path planning method provided in this embodiment, with the help of high-precision maps, can realize global path planning of unmanned transport vehicles taking into account turning points in underground mining environments, effectively solve the problem that ordinary path planning cannot select the optimal path according to factors such as travel time, and solves the problem that the path planning result is invalid when the starting driving direction and the ending point driving direction are inconsistent. It can effectively improve the driving safety and transportation efficiency of unmanned transport vehicles in underground mining.

Claims

1. A global path planning method for an unmanned transport vehicle in an underground mine, characterized in that: The steps include: S1. Analyze the high-precision map file of the underground mine according to the protocol and store it in the database; S2. Construct a directed topological graph of the high-precision map, using the high-precision map sections as nodes and the connectivity relationships of the sections as edges; S3, connecting the incoming section and the outgoing section of the U-turn point on the high-precision map, and updating the directed topological map; S4. Path planning based on the directed topological graph of the high-precision map, including: Obtain the starting and ending sections of the vehicle, determine the vehicle's driving direction on the starting section, and the vehicle's driving direction on the ending section, use the starting and ending sections as the starting and ending points of the high-precision map directed topology graph, perform path planning based on the A* algorithm, and obtain all accessible paths from the starting section to the ending section; S5. Screening multiple paths planned by the path planning to obtain the optimal path, specifically including: first, screening the path according to the driving direction, and then screening the path according to the principle of shortest travel time; Path filtering based on driving direction includes: First, find out how many U-turn points the path passes through. If the starting driving direction is equal to the ending driving direction and the number of U-turn points is an even number, the path is retained; if the starting driving direction is equal to the ending driving direction and the number of U-turn points is an odd number, the path is deleted; if the starting driving direction is not equal to the ending driving direction and the number of U-turn points is an odd number, the path is retained; if the starting driving direction is not equal to the ending driving direction and the number of U-turn points is an even number, the path is deleted.

2. The global path planning method for an unmanned transport vehicle for underground mining according to claim 1 is characterized in that: In step S1, according to the high-precision map annotation protocol, each field of the high-precision map is parsed one by one and stored in the database. The database includes a high-precision map point information table, a high-precision map line string information table, a high-precision map section information table, a high-precision map relationship information table, a high-precision map poi point information table, and a high-precision map topology information table.

3. The global path planning method for an unmanned transport vehicle for underground mining according to claim 1 is characterized in that: Step S2 specifically includes: First, obtain all the road segment IDs of the high-precision map from the database, and then obtain the four vertices of the road segment according to the road segment ID. The four vertices are the lower left corner, lower right corner, upper left corner, and upper right corner; Loop through all road sections. If the upper left corner and upper right corner of section a are the same as the lower left corner and lower right corner of section b, it means that section a points to section b. Then, establish an edge between vertex a and vertex b in the directed topology graph of the high-precision map.

4. The global path planning method for an unmanned transport vehicle for underground mining according to claim 3 is characterized in that: The weights of all edges in the directed topology graph of the high-precision map are set to 1.

5. The global path planning method for an unmanned transport vehicle for underground mining according to claim 1 is characterized in that: Step S3 is specifically as follows: Get the entry and exit section IDs of all U-turn points in the high-precision map, establish a connection relationship between the vertices represented by the entry and exit sections, and update the directed topology graph.

6. The global path planning method for an unmanned transport vehicle for underground mining according to claim 1, characterized in that: Based on the starting coordinate point of the vehicle, determine which road sections the vehicle is in. If there are none, use the road section closest to this point as the starting road section of the vehicle. The method for obtaining the ending road section of the vehicle is the same as that of the starting road section. If the difference between the vehicle's starting section direction angle and the vehicle's heading angle at the starting point is less than 90 degrees or greater than 270 degrees, it means that the vehicle is about to travel in the forward direction along the section, otherwise it means that the vehicle needs to travel in the reverse direction along the section.

7. The global path planning method for an unmanned transport vehicle for underground mining according to claim 1 is characterized in that: Path screening based on the principle of shortest travel time specifically includes: Calculate the forward driving distance and reverse driving distance in the path, and then obtain the vehicle's travel time on the path based on the vehicle's forward driving speed and reverse driving speed. Select the path with the shortest travel time as the optimal path for this path planning.

Images