WO2020173044A1 - 无人机路径规划方法、装置、存储介质及计算机设备 - Google Patents

无人机路径规划方法、装置、存储介质及计算机设备 Download PDF

Info

Publication number
WO2020173044A1
WO2020173044A1 PCT/CN2019/098002 CN2019098002W WO2020173044A1 WO 2020173044 A1 WO2020173044 A1 WO 2020173044A1 CN 2019098002 W CN2019098002 W CN 2019098002W WO 2020173044 A1 WO2020173044 A1 WO 2020173044A1
Authority
WO
WIPO (PCT)
Prior art keywords
sampling
sampling point
route
point
target
Prior art date
Application number
PCT/CN2019/098002
Other languages
English (en)
French (fr)
Inventor
周翊民
陈金保
Original Assignee
中国科学院深圳先进技术研究院
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 中国科学院深圳先进技术研究院 filed Critical 中国科学院深圳先进技术研究院
Publication of WO2020173044A1 publication Critical patent/WO2020173044A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/20Instruments for performing navigational calculations

Definitions

  • the present invention relates to the field of drone technology and computer technology, and in particular to a method, device, storage medium and computer equipment for drone path planning.
  • path planning is a key link in the autonomous flight of drones. Its purpose is to plan a meeting in the current environment according to mission requirements.
  • the constrained flight path enables the UAV to avoid collisions during flight and quickly reach the end from the starting point.
  • UAV path planning for three-dimensional complex and changeable environments mostly adopt sampling-based path planning methods, such as Rapidly-exploring Random Tree (RRT) and Probabilistic Roadmap Method. , PRM).
  • RRT Rapidly-exploring Random Tree
  • PRM Probabilistic Roadmap Method.
  • the sampling-based path planning algorithm is to sample in the state space and then perform collision detection. There is no need to preprocess the space model, and the amount of calculation will not change in the high-dimensional environment, which is very suitable for the high-dimensional UAV flight.
  • Unknown complex environment because the map is uniformly sampled, the completion of the route map when there are narrow passages in the map often requires the construction of more sampling points, which leads to an exponential increase in the amount of calculation.
  • a UAV path planning method, device, storage medium and computer equipment are provided, which increase the number of sampling points in the narrow channel with lower complexity, and improve the overall efficiency of route planning .
  • a method for UAV route planning including:
  • the route planning information includes a target planning area, starting point information, and target point information
  • the target planning area includes a feasible area and an obstacle area
  • the optimal route is found in the edge set as the target route.
  • a UAV route planning device which is characterized in that it comprises:
  • a route planning information determining module configured to obtain route planning information, the route planning information includes target planning area, starting point information, and target point information, and the target planning area includes a feasible area and an obstacle area;
  • the vertex set construction module is used to determine a first sampling point in the feasible area by random uniform sampling, determine that the first sampling point does not collide with the obstacle area, and calculate the field strength value corresponding to the first sampling point , Determining that the field strength value is greater than or equal to a preset field strength threshold, and adding the first sampling point to the vertex set;
  • An edge set construction module configured to generate a corresponding edge set according to a preset planner according to the vertex set
  • the route planning module is used to search for the optimal route as the target route in the edge concentration based on the starting point information, the target point information, and the preset route search algorithm.
  • a computer device including a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the aforementioned The steps of the man-machine path planning method.
  • a computer-readable storage medium which stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the aforementioned UAV path planning method.
  • the UAV route planning scheme provided by the embodiment of the present invention is used to increase the average number of sampling points in the narrow channel without increasing the number of sampling points, thereby improving the route planning efficiency Accuracy and overall efficiency.
  • Fig. 1 is a schematic flow chart of a method for planning a UAV path in an embodiment
  • Figure 2 is a schematic diagram of a feasible area in an embodiment
  • FIG. 3 is a schematic flowchart of a vertex set construction process in an embodiment
  • FIG. 4 is a schematic flowchart of a vertex set construction process in an embodiment
  • FIG. 5 is a schematic flowchart of a vertex set construction process in an embodiment
  • FIG. 6 is a schematic flowchart of an edge set construction process in an embodiment
  • FIG. 7 is a schematic diagram of an undirected network diagram structure in an embodiment
  • FIG. 8 is a schematic flowchart of a method for planning a path of a drone in an embodiment
  • Fig. 9 is a schematic diagram of sampling point distribution in an embodiment of the present invention.
  • FIG. 10 is a schematic diagram of sampling point distribution corresponding to the PRM algorithm in the prior art
  • FIG. 11 is a schematic diagram of the distribution of the average number of sampling points in a narrow channel in an embodiment
  • Figure 12 is a schematic structural diagram of a UAV path planning device in an embodiment
  • Fig. 13 is a schematic structural diagram of a computer device that runs the aforementioned UAV path planning method in an embodiment.
  • a UAV path planning method is proposed.
  • the implementation of the method can rely on a computer program, which can run on a computer system based on the von Neumann system, and the computer program can It is an application program for calculating and planning the UAV route.
  • the computer system may be a computer device such as a smart phone, a tablet computer, a personal computer, a server and the like running the above computer program.
  • the computer equipment that runs the above-mentioned UAV route planning method is a computer equipment connected to the UAV equipment, for example, a controller connected to the UAV equipment (the controller may Is a computer device such as a smart phone, a tablet, a personal computer, or a server).
  • the execution of the above-mentioned UAV route planning method may also be based on a UAV device on which is provided A processor, through which the aforementioned UAV route planning method is executed.
  • a UAV path planning method is provided, which specifically includes the following steps S102-S108:
  • Step S102 Obtain route planning information, where the route planning information includes a target planning area, starting point information, and target point information, and the target planning area includes a feasible area and an obstacle area.
  • the target planning area corresponding to the route planning is determined, and the target planning area is the area that needs to be routed.
  • the target planning area may be an image area collected by the drone's camera device and image processing is performed during the flight.
  • the target planning area A is an image area after image recognition/processing.
  • the target planning area includes a feasible area A1 and an obstacle area A2, where the feasible area is an optional area for route planning, and the obstacle area is The area where the obstacle is located is not selectable during the route planning process, and the planned route does not cross or collide with the obstacle area.
  • A1 is a feasible area
  • A2 is an obstacle area.
  • the route planning information that needs to be obtained in advance also includes starting point information and target point information, which are used to determine the starting point and ending point of the route.
  • Step S104 Determine a first sampling point in the feasible area by random uniform sampling, determine that the first sampling point does not collide with the obstacle area, calculate the field strength value corresponding to the first sampling point, and determine the The field strength value is greater than or equal to the preset field strength threshold, and the first sampling point is added to the vertex set.
  • the route planning process of the drone is a process of determining an undirected network diagram corresponding to the target planning area (step S104, step S106), and then determining the optimal route in the undirected network diagram.
  • G(V, E) be the undirected network graph obtained through steps S104 and S106, where V is the vertex set, and E is the edge set corresponding to the vertex set.
  • step S104 is the process of generating the vertex set V.
  • the sampling points in the vertex set are points obtained by sampling in the feasible area, and the obtained first sampling point is confirmed whether it meets the relevant requirements for adding to the vertex set.
  • the acquisition of the first sampling point is a sampling point acquired by random uniform sampling within a feasible area.
  • the first sampling point can be added to the vertex set.
  • the sampling points in order to increase the number of sampling points in the narrow channel, it is necessary to further determine the sampling points.
  • the potential field is introduced, the field strength value U q corresponding to the first sampling point is calculated in the target planning area, and it is determined whether it is greater than the preset field strength threshold U 0 , and only when U q > U 0 Add the first sample point to the vertex set V.
  • step S104 includes the following steps:
  • Step S1041 Randomly and uniformly sample in the feasible area A1, and obtain the sampling point q;
  • Step S1042 Determine whether the sampling point q does not collide with the obstacle area A2;
  • step S1043 calculate the field strength value U q of the sampling point q
  • step S1044 determine whether the field strength value U q of the sampling point q is greater than the field strength threshold U 0 ;
  • step S1045 add the sampling point q to the vertex set V.
  • a threshold n of a point is set as the minimum value of the number of sampling points included in the vertex set.
  • the above method further includes:
  • step S104 is cyclically executed until the number of sampling points included in the vertex set is greater than or equal to the preset point number threshold n.
  • the method further includes: calculating a first distance value between the first sampling point and the sampling points included in the vertex set, and determining the calculation The obtained first distance value is greater than zero.
  • q 1 and q 2 are two points in the feasible area respectively, and the calculation of the distance between the two points can be calculated by the following calculation formula:
  • d(q 1 , q 2 ) is the distance value between points q 1 and q 2 .
  • the first distance value between the sampling point q and the sampling points included in the vertex set V is calculated through the above-mentioned distance value calculation formula, and it is determined that the calculated first distance value is greater than 0, that is to say , Determine that the sampling point q is not an existing point in the vertex set, and avoid repeated sampling point calculation and determination processes.
  • step S1045 it is also necessary to determine again whether the sampling point q does not collide with the obstacle area A2, so as to reconfirm the sampling points added to the vertex set.
  • the method further includes: performing collision detection on the first sampling point, and determining that the first sampling point does not collide with the obstacle area.
  • step S104 a schematic flowchart of the process of generating the vertex set V in step S104 is given.
  • the calculation process of introducing the potential field to calculate the field strength value of the sampling point can be as follows:
  • the preset distance threshold D if d ⁇ D, the calculation formula of the repulsive force p s is:
  • Step S106 Generate a corresponding edge set according to the preset planner according to the vertex set.
  • the corresponding edge set E can be constructed according to the vertex set V, that is, the sampling points q in the vertex set V are connected to each other to generate the edge set E.
  • step S106 includes:
  • Step S1061 Traverse the second sampling points included in the vertex set, and for the traversed second sampling points, determine the neighboring points of the preset point number threshold corresponding to the second sampling point in the vertex set;
  • Step S1062 Connect the second sampling point and neighboring points through a preset planner
  • Step S1063 In the case of a successful connection, add the edge connecting the second sampling point and the neighboring point to the edge set.
  • the neighbors of the preset point threshold (k) are determined by distance calculation (that is, k neighbors corresponding to the sampling point q) are determined, and then the Set a planner (for example, a local planner) to connect the sampling point q and the neighboring point q'. If the connection is successful (that is, the local path without conflict), the edge (q, q') is added to the edge set E. Until all neighbors are connected or discarded.
  • the setting of the point number threshold k is related to the number of sampling points in the narrow channel. In order to increase the number of sampling points in the narrow channel as much as possible, you can set k The value is set as large as possible, but it will also increase the amount of calculation in the construction of the edge set E.
  • Figure 7 shows an example of the construction of an undirected network graph.
  • the process of determining the k neighboring points corresponding to the sampling point q is determined by calculating the distance between the sampling point and other sampling points in the vertex set. That is to say, the step of determining the neighboring points of the preset point number threshold corresponding to the second sampling point in the vertex set further includes: calculating the difference between the second sampling point and the second sampling point included in the vertex set Sort the second distance value from small to large, and obtain the second sampling point corresponding to the first preset point number threshold second distance value as the neighboring point.
  • Step S108 According to the starting point information and the target point information, according to the preset path search algorithm, the optimal route is found in the edge set as the target route.
  • the UAV route can be found in the undirected network diagram, and different path search algorithms can be selected as needed to find the optimal route as the target route of the UAV route planning.
  • the route with the shortest distance is found in the edge set E, and the route with the shortest distance found is taken as the target route.
  • the number of sampling points in the narrow channel is significantly increased in the vertex set structure in Figure 9.
  • the average number of sampling points in the narrow channel under the same value of n and the selection of different k values is shown in Figure 11 in the narrow channel under the PRM algorithm without introducing the potential field.
  • the average number of sampling points and the average number of sampling points in the narrow channel under the UAV route planning method provided by the present invention, the number of sampling points in the narrow channel of the present invention is significantly increased.
  • the embodiment of the present invention also provides a UAV route planning device.
  • the UAV route planning device includes:
  • the route planning information determining module 102 is configured to obtain route planning information, where the route planning information includes a target planning area, starting point information, and target point information, and the target planning area includes a feasible area and an obstacle area;
  • the vertex set construction module 104 is configured to determine a first sampling point in the feasible area by random uniform sampling, determine that the first sampling point does not collide with the obstacle area, and calculate the field strength corresponding to the first sampling point Value, determining that the field strength value is greater than or equal to a preset field strength threshold, and adding the first sampling point to the vertex set;
  • the edge set construction module 106 is configured to generate a corresponding edge set according to the vertex set according to a preset planner
  • the route planning module 108 is configured to find the optimal route in the edge cluster as the target route according to the starting point information, the target point information, and the preset route search algorithm.
  • the vertex set construction module 104 is further configured to cyclically execute the determination in the feasible region by random uniform sampling when the number of sampling points contained in the vertex set is less than a preset point threshold. The first sampling point, determining that the first sampling point does not collide with the obstacle area, calculating the field strength value corresponding to the first sampling point, determining that the field strength value is greater than or equal to a preset field strength threshold, and The first sampling point is added to the vertex set.
  • the vertex set construction module 104 is also used to calculate the repulsive force of the first sampling point relative to the obstacle area; calculate the attractive force of the first sampling point relative to the obstacle area according to the target point information ; Based on the repulsive force and the attractive force, the field strength value corresponding to the first sampling point is calculated.
  • the first distance value between the first sampling point and the sampling points included in the vertex set is calculated, and it is determined that the calculated first distance value is greater than zero.
  • the edge set construction module 106 is further configured to traverse the second sampling points included in the vertex set, and for the second sampling points traversed, determine in the vertex set that they correspond to the second sampling points The second sampling point and the neighboring point are connected through the preset planner; if the connection is successful, the edge connecting the second sampling point and the neighboring point is added to the edge set.
  • the edge set construction module 106 is also used to calculate a second distance value between the second sampling point and the second sampling point included in the vertex set; and the second distance value is reduced to For large sorting, the second sampling point corresponding to the second distance value of the previous preset point number threshold is obtained as the neighboring point.
  • the route planning module 108 is further configured to search for the route with the shortest distance in the edge cluster according to a preset quick search algorithm, and use the route with the shortest distance as the target route.
  • Fig. 13 shows an internal structure diagram of a computer device in an embodiment.
  • the computer device may specifically be a server.
  • the computer device includes a processor, a memory, and a network interface connected through a system bus.
  • the memory includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium of the computer device stores an operating system and may also store a computer program.
  • the processor can enable the processor to implement the UAV route planning method.
  • a computer program may also be stored in the internal memory, and when the computer program is executed by the processor, the processor can execute the UAV route planning method.
  • the network interface is used to communicate with the outside.
  • 13 is only a block diagram of part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied.
  • the specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.
  • the UAV route planning method provided in this application can be implemented in the form of a computer program, and the computer program can be run on a computer device as shown in FIG. 13.
  • the memory of the computer device can store various program templates that make up the UAV route planning device.
  • a computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the following steps:
  • the foregoing computer equipment in one of the embodiments, when the foregoing computer program is executed by the processor, it is further configured to perform the following steps:
  • the route planning information includes a target planning area, starting point information, and target point information
  • the target planning area includes a feasible area and an obstacle area
  • the optimal route is found in the edge set as the target route.
  • a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the processor executes the following steps:
  • the route planning information includes a target planning area, starting point information, and target point information
  • the target planning area includes a feasible area and an obstacle area
  • the optimal route is found in the edge set as the target route.
  • the UAV route planning scheme provided by the embodiment of the present invention is used to increase the average number of sampling points in the narrow channel without increasing the number of sampling points, thereby improving the route planning efficiency Accuracy and overall efficiency.
  • UAV route planning method UAV route planning device, computer equipment and computer readable storage medium belong to the same inventive concept.
  • UAV route planning method, UAV route planning device, and computer equipment And the content involved in the computer-readable storage medium is mutually applicable.
  • Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Channel (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
  • SRAM static RAM
  • DRAM dynamic RAM
  • SDRAM synchronous DRAM
  • DDRSDRAM double data rate SDRAM
  • ESDRAM enhanced SDRAM
  • SLDRAM synchronous chain Channel
  • memory bus Radbus direct RAM
  • RDRAM direct memory bus dynamic RAM
  • RDRAM memory bus dynamic RAM

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

提供了一种无人机路线规划方法、装置、存储介质及计算机设备,其中方法包括:获取路线规划信息,路线规划信息包括目标规划区域、起始点信息、目标点信息,目标规划区域包括可行区域和障碍区域(S102);通过随机均匀采样方式在可行区域确定第一采样点,确定第一采样点与障碍区域无碰撞,计算第一采样点对应的场强值,确定场强值大于或等于预设的场强阈值,将第一采样点添加到顶点集(S104);根据顶点集按照预设的规划器生成对应的边缘集(S106);根据起始点信息、目标点信息,按照预设的路径搜索算法,在边缘集中查找最优路线作为目标路线(S108)。该无人机路线规划方法及装置提高了路线规划的准确性和规划效率。

Description

无人机路径规划方法、装置、存储介质及计算机设备 技术领域
本发明涉及无人机技术领域和计算机技术领域,尤其涉及一种无人机路径规划方法、装置、存储介质及计算机设备。
背景技术
随着计算机技术的发展,带动新一轮的人工智能革命,无人机智能化成为趋势,而路径规划是无人机自主飞行的关键环节,其目的是根据任务需求在当前环境中规划一条满足约束条件的飞行路径,使无人机在飞行过程中避免碰撞并且快速从起点到达终点。
与2D路径规划不同的是,3D环境中的无人机路径规划的困难随着动态约束而呈指数增长,并且运动约束变得更加复杂。在相关技术方案中,针对三维复杂多变环境的无人机路径规划多采用基于采样的路径规划方法,例如快速随机搜索树法(Rapidly-exploring Random Tree,RRT)和概率地图算法(Probabilistic Roadmap Method,PRM)。基于采样的路径规划算法是在状态空间中进行采样然后进行碰撞检测,不需要对空间模型预处理,而且在高维环境中计算量也不会改变,很适合无人机飞行所处的高维未知复杂环境。但是,由于是在地图均匀采样,当地图中存在狭窄通道时完成线路图往往需要构造较多采样点,这就导致了计算量随之指数级增长。
也就是说,现有的路线规划方法中对于狭窄通道中的采样点确定及路线规划,为了提高狭窄通道路线规划的准确性和有效性,需要提高采样点数量,这就导致了计算量过大、采样点利用率过低。
发明内容
基于此,在本发明中,提供了一种无人机路径规划方法、装置、存储介质及计算机设备,在较低复杂度下提高了狭窄通道中采样点的数量,提高了路线规划的整体效率。
在本发明的第一方面,提供了一种无人机路线规划方法,包括:
获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
根据所述顶点集按照预设的规划器生成对应的边缘集;
根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
在本发明的第二方面,提供了一种无人机路线规划装置,其特征在于,包括:
路线规划信息确定模块,用于获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
顶点集构造模块,用于通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
边缘集构造模块,用于根据所述顶点集按照预设的规划器生成对应的边缘集;
路线规划模块,用于根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
在本发明的第三方面,还提出了一种计算机设备,包括存储器和处理器,所述存储器存储有计算机程序,所述计算机程序被所述处理器执行时,使得所述处理器执行前述无人机路径规划方法的步骤。
在本发明的第四方面,还提出了一种计算机可读存储介质,存储有计算机程序,所述计算机程序被处理器执行时,使得所述处理器执行前述无人机路径规划方法的步骤。
实施本发明实施例,将具有如下有益效果:
采用了上述无人机路径规划方法、装置、存储介质及计算机设备之后,在对无人机路线进行规划的过程中,在可行区域内进行随机均匀采样,对采样点不仅需要进行与障碍区域的无碰撞检测,还需要引入势场对该采样点的场强值 进行计算,只有在计算得到的场强值大于一定值的情况下,才将该采样点添加到顶点集中;然后根据该顶点集通过规划器生成对应的边缘集,然后根据顶点集、边缘集组成的无向网络图进行路径搜索,以获取无人机规划的目标路线。也就是说,通过引入势场,尽可能多的保留了可行区域中狭窄通道内的采样点,增加了狭窄通道内采样点的平均数量。
相较于RRT算法、或PRM算法,采用本发明实施例提供的无人机路线规划方案,在没有增加采样点数量的情况下,提高狭窄通道内采样点的平均数量,从而提高了路线规划的准确性和整体效率。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
其中:
图1为一个实施例中一种无人机路径规划方法的流程示意图;
图2为一个实施例中可行区域示意图;
图3为一个实施例中顶点集构造过程的流程示意图;
图4为一个实施例中顶点集构造过程的流程示意图;
图5为一个实施例中顶点集构造过程的流程示意图;
图6为一个实施例中边缘集构造过程的流程示意图;
图7为一个实施例中无向网络图构造示意图;
图8为一个实施例中无人机路径规划方法的流程示意图;
图9为一个本发明的一个实施例中采样点分布示意图;
图10为现有技术的PRM算法对应的采样点分布示意图;
图11为一个实施例中狭窄通道内平均采样点数量分布示意图;
图12为一个实施例中一种无人机路径规划装置的结构示意图;
图13为一个实施例中运行上述无人机路径规划方法的计算机设备的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
在本实施例中,特提出了一种无人机路径规划方法,该方法的实现可依赖于计算机程序,该计算机程序可运行于基于冯诺依曼体系的计算机系统之上,该计算机程序可以是对无人机路线进行计算和规划的应用程序。该计算机系统可以是运行上述计算机程序的例如智能手机、平板电脑、个人电脑、服务器等计算机设备。
需要说明的是,在本实施例中,运行上述无人机路线规划方法的计算机设备为一与无人机设备连接的计算机设备,例如,与无人机设备连接的控制器(该控制器可以是智能手机、平板电脑、个人电脑或服务器等计算机设备),在另一个实施例中,上述无人机路线规划方法的执行还可以是基于一无人机设备,该无人机设备上设置有处理器,通过该处理器来执行上述无人机路线规划方法。
如图1所示,在一个实施例中,提供了一种无人机路径规划方法,具体包括如下步骤S102-S108:
步骤S102:获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域。
在需要对无人机的路线进行规划的情况下,确定路线规划对应的目标规划区域,该目标规划区域即为需要进行线路规划的区域。在一个具体的实施例中,目标规划区域可以是在飞行过程中通过无人机的摄像设备采集并进行图像处理之后的图像区域。
在本实施例中,目标规划区域A为经过图像识别/处理之后的图像区域,在目标规划区域中包含了可行区域A1和障碍区域A2,其中可行区域为路线规划的可选区域,障碍区域为障碍物所在的区域,在路线规划的过程中为不可选区域,且规划的路线与障碍区域也不存在交叉或碰撞。例如,在如图2所示的场景,A1为可行区域,A2为障碍区域。
进一步的,对无人机的路线进行规划,需要提前获取的路线规划信息还包括起始点信息、目标点信息,用于确定路线的起始点和终点。
步骤S104:通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集。
在本实施例中,对无人机的路线规划过程为确定与目标规划区域对应的无向网络图(步骤S104、步骤S106)、然后在该无向网络图中确定最优路线的过程。
记G(V,E)为通过步骤S104、S106获取的无向网络图,其中V为顶点集,E为与顶点集对应的边缘集。
具体的,上述步骤S104即为生成顶点集V的过程。
在本实施例中,顶点集中的采样点,是通过在可行区域内采样获取的点,并且对获取到的第一采样点进行确认,是否符合添加到顶点集的相关要求。且,在一个具体的实施例中,第一采样点的获取是在可行区域内进行随机均匀采样获取的采样点。
在获取到第一采样点q之后,还需要对该第一采样点q进行碰撞检测,也就是说,需要确定第一采样点q与障碍区域A2是无碰撞的,与障碍区域无碰撞的采样点才能作为顶点集V中的顶点。
在一般情况下,在确定第一采样点与障碍区域无碰撞之后即可将该第一采样点添加到顶点集。在本实施例中,为了提高狭窄通道内的采样点数量,还需要对采样点进行进一步的判断。
具体的,引入势场,在目标规划区域内计算第一采样点对应的场强值U q,并且判断是否大于预设的场强阈值U 0,只有在U q>U 0的情况下,才将第一采样点添加到顶点集V。
进一步的,在一个具体的实施例中,如图3所示,上述步骤S104包括如下步骤:
步骤S1041:在可行区域A1中随机均匀采样,获取采样点q;
步骤S1042:判断采样点q是否与障碍区域A2无碰撞;
若是,执行步骤S1043:计算采样点q的场强值U q;步骤S1044:判断采 样点q的场强值U q是否大于场强阈值U 0
若是,执行步骤S1045:将采样点q添加到顶点集V。
在本实施例中,为了保证路线规划的合理性和准确性,还需要保证顶点集中的采样点数量超过一定值,例如,设定一点数阈值n作为顶点集中包含的采样点数量的最低值。
具体的,在一个实施例中,如图4所示,上述方法还包括:
在所述顶点集包含的采样点数量小于预设的点数阈值n的情况下,循环执行步骤S104,直至顶点集中包含的采样点数量大于或等于预设的点数阈值n。
上述通过随机均匀采样方式在所述可行区域确定第一采样点的步骤之后还包括:计算所述第一采样点与所述顶点集包含的采样点之间的第一距离值,确定所述计算得到的第一距离值大于0。
q 1、q 2分别为可行区域内的两个点,二者之间的距离值的计算可以通过如下计算公式进行计算:
Figure PCTCN2019098002-appb-000001
其中,d(q 1,q 2)为点q 1、q 2之间的距离值。
在本实施例中,通过上述距离值计算公式,计算采样点q与顶点集V中包含的采样点之间的第一距离值,并且确定该计算得到的第一距离值大于0,也就是说,确定采样点q不是顶点集中已有的点,避免重复的采样点计算和确定过程。
需要说明的是,在本实施例中,在步骤S1045之前,还需要再次判断采样点q是否与障碍区域A2无碰撞,从而对添加到顶点集的采样点进行再次确认。
具体的,上述将所述第一采样点添加到顶点集的步骤之前,还包括:对第一采样点进行碰撞检测,确定所述第一采样点与所述障碍区域无碰撞。
如图5所示,在一个具体的实施例中,给出了步骤S104中生成顶点集V的过程的流程示意图。
另外,在本实施例中,引入势场计算采样点的场强值的计算过程可以如下:
在无向网络图中引入势场,计算每一个采样点点在网络图中的来自障碍物的排斥力p s和吸引力p g,然后通过排斥力和吸引力来计算该采样点的场强值:
U q=p s+p g
具体的,采样点来自障碍物的排斥力p s的计算过程如下:
按照距离值计算公式计算当前采样点q(坐标(x,y))与障碍物中心(坐标(x 0,y 0))之间的距离d:
预设距离阈值D,若d≦D,排斥力p s的计算公式为:
Figure PCTCN2019098002-appb-000002
其中,d 2=d/100+1,d 0=2,C=800,且其中d 0和C可以根据需要设置成其他值。
若d>D,p s=0。
需要说明的是,在本实施例中,D=100,或者可以根据实际的需要,设置其他的距离阈值。
采样点对应的吸引力为:
p g=α*((x-g 1) 2+(y-g 1) 2)
其中,α为常数,且在一个实施例中,α=1/700,(g 1,g 2)为目标点坐标。
需要说明的是,在本实施例中,还可以根据引入势场函数设置其他的场强值计算公式,在本实施例中不做限定。
步骤S106:根据所述顶点集按照预设的规划器生成对应的边缘集。
在无向网络图的顶点集的构造完成之后,即可根据顶点集V构造相应的边缘集E,即将顶点集V中的采样点q相互之间进行连接,从而生成边缘集E。
具体的,如图6所示,步骤S106包括:
步骤S1061:遍历所述顶点集中包含的第二采样点,对于遍历到的第二采样点,在所述顶点集中确定与所述第二采样点对应的预设点数阈值的邻点;
步骤S1062:通过预设的规划器连接第二采样点和邻点;
步骤S1063:在连接成功的情况下,将连接所述第二采样点与邻点的边添 加到边缘集。
也就是说,对于顶点集中包含的每一个采样点q,通过距离值计算来确定其预设点数阈值(k)的邻点(即确定与采样点q对应的k个邻点),然后通过预设的规划器(例如,本地规划器)来连接采样点q与邻点q’,如果连接成功(即无冲突本地路径),则将边(q,q')添加到边缘集E中。直至所有的邻点均被连接或丢弃。
需要说明的是,在本实施例中,点数阈值k的设定关系到了狭窄通道中采样点的数量,为了尽可能的提高狭窄通道内采样点的数量,可在<n的情况下,将k值设置得尽可能的大,但是同时也会增加边缘集E构造过程的计算量。
在对每一个采样点均执行上述邻点确定、规划器连接、添加边缘集/丢弃的过程之后,即完成了边缘集E的构建。
例如,如图7所示,图7给出了无向网络图的构建例子。
需要说明的是,在本实施例中,在确定与采样点q对应的k个邻点的过程,是通过计算采样点与顶点集中的其他采样点之间的距离值来确定的。也就是说,上述在所述顶点集中确定与所述第二采样点对应的预设点数阈值的邻点的步骤,还包括:计算所述第二采样点与顶点集包含的第二采样点之间的第二距离值;对所述第二距离值进行自小到大的排序,获取与前预设点数阈值个第二距离值对应的第二采样点作为邻点。
步骤S108:根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
在无向网络图构建完成以后,即可在该无向网络图中查找无人机路线,并且可以根据需要选择不同的路径搜索算法来查找最优路线作为无人机路线规划的目标路线。
例如,按照预设的快速搜索算法,在边缘集E中查找距离最短的路线,将查找到的距离最短的路线作为目标路线。
需要说明的是,在本实施例中,不对路线搜索的具体算法进行限定,可以是任一的在无向网络图中查找相应路线的搜索算法均属于本发明实施例的保护范围。
如图8所示,展示了上述无人机路线规划方法的流程示意图。
采用上述无人机路线规划方法,可在不增加采样次数和采样点数量的情况下尽可能多的保留障碍物边缘的采样点,增加狭窄通道内的采样点数量,更加有效的完成路线图的构建。
如图9所示,图9给出了在n=50、n=100、n=200的情况下,目标规划区域内的顶点集的相应示意图,相对于图10给出在未引入势场的PRM算法构建的顶点集的相关示意图来讲,图9中的顶点集构造上,狭窄通道中的采样点数量明显增加。
进一步的,如图11所示,在相同的n值下,对于不同的k值的选择下的狭窄通道内平均采样点数,图11中给出了未引入势场的PRM算法下的狭窄通道内的平均采样点数以及采用本发明给出的无人机路线规划方法下的狭窄通道内的平均采样点数,本发明的狭窄通道中的采样点数量明显增加。
如图12所示,本发明实施例还提供一种无人机路线规划装置。具体的,如图12所示,所述无人机路线规划装置包括:
路线规划信息确定模块102,用于获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
顶点集构造模块104,用于通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
边缘集构造模块106,用于根据所述顶点集按照预设的规划器生成对应的边缘集;
路线规划模块108,用于根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
在其中一个实施例中,顶点集构造模块104还用于在所述顶点集包含的采样点数量小于预设的点数阈值的情况下,循环执行所述通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集的。
在其中一个实施例中,顶点集构造模块104还用于计算第一采样点相对于所述障碍区域的排斥力;根据所述目标点信息计算第一采样点相对于所述障碍区域的吸引力;基于所述排斥力和所述吸引力,计算所述第一采样点对应的场强值。
在其中一个实施例中,计算所述第一采样点与所述顶点集包含的采样点之间的第一距离值,确定所述计算得到的第一距离值大于0。
在其中一个实施例中,边缘集构造模块106还用于遍历所述顶点集中包含的第二采样点,对于遍历到的第二采样点,在所述顶点集中确定与所述第二采样点对应的预设点数阈值的邻点;通过预设的规划器连接第二采样点和邻点;在连接成功的情况下,将连接所述第二采样点与邻点的边添加到边缘集。
在其中一个实施例中,边缘集构造模块106还用于计算所述第二采样点与顶点集包含的第二采样点之间的第二距离值;对所述第二距离值进行自小到大的排序,获取与前预设点数阈值个第二距离值对应的第二采样点作为邻点。
在其中一个实施例中,路线规划模块108还用于按照预设的快速搜索算法,在所述边缘集中查找距离最短的路线,将查找到的距离最短的路线作为目标路线。
图13示出了一个实施例中计算机设备的内部结构图。该计算机设备具体可以是服务器。如图13所示,该计算机设备包括通过系统总线连接的处理器、存储器和网络接口。其中,存储器包括非易失性存储介质和内存储器。该计算机设备的非易失性存储介质存储有操作系统,还可存储有计算机程序,该计算机程序被处理器执行时,可使得处理器实现无人机路线规划方法。该内存储器中也可储存有计算机程序,该计算机程序被处理器执行时,可使得处理器执行无人机路线规划方法。网络接口用于与外部进行通信。本领域技术人员可以理解,图13中示出的结构,仅仅是与本申请方案相关的部分结构的框图,并不构成对本申请方案所应用于其上的计算机设备的限定,具体的计算机设备可以包括比图中所示更多或更少的部件,或者组合某些部件,或者具有不同的部件布置。
在一个实施例中,本申请提供的无人机路线规划方法可以实现为一种计算机程序的形式,计算机程序可在如图13所示的计算机设备上运行。计算机设 备的存储器中可存储组成无人机路线规划装置的各个程序模板。
一种计算机设备,包括存储器和处理器,所述存储器存储有计算机程序,所述计算机程序被所述处理器执行时,使得所述处理器执行如下步骤:
上述计算机设备,在其中一个实施例中,上述计算机程序被所述处理器执行时,还用于执行以下步骤:
获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
根据所述顶点集按照预设的规划器生成对应的边缘集;
根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
一种计算机可读存储介质,存储有计算机程序,所述计算机程序被处理器执行时,使得所述处理器执行如下步骤:
获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
根据所述顶点集按照预设的规划器生成对应的边缘集;
根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
上述无人机路径规划方法、装置、计算机设备及计算机可读存储介质中,在对无人机路线进行规划的过程中,在可行区域内进行随机均匀采样,对采样点不仅需要进行与障碍区域的无碰撞检测,还需要引入势场对该采样点的场强值进行计算,只有在计算得到的场强值大于一定值的情况下,才将该采样点添加到顶点集中;然后根据该顶点集通过规划器生成对应的边缘集,然后根据顶点集、边缘集组成的无向网络图进行路径搜索,以获取无人机规划的目标路线。 也就是说,通过引入势场,尽可能多的保留了可行区域中狭窄通道内的采样点,增加了狭窄通道内采样点的平均数量。
相较于RRT算法、或PRM算法,采用本发明实施例提供的无人机路线规划方案,在没有增加采样点数量的情况下,提高狭窄通道内采样点的平均数量,从而提高了路线规划的准确性和整体效率。
需要说明的是,上述无人机路线规划方法、无人机路线规划装置、计算机设备和计算机可读存储介质属于同一个发明构思,无人机路线规划方法、无人机路线规划装置、计算机设备和计算机可读存储介质中涉及的内容可相互适用。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成,所述的程序可存储于一非易失性计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用,均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
以上实施例的各技术特征可以进行任意的组合,为使描述简洁,未对上述实施例中的各个技术特征所有可能的组合都进行描述,然而,只要这些技术特征的组合不存在矛盾,都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式,其描述较为具体和详细,但并不能因此而理解为对本申请专利范围的限制。应当指出的是,对于本领域的普通技术人员来说,在不脱离本申请构思的前提下,还可以做出若干变形和改进,这些都属于本申请的保护范围。因此,本申请专利的保护范围应以所附权利要求为准。

Claims (10)

  1. 一种无人机路线规划方法,其特征在于,包括:
    获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
    通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
    根据所述顶点集按照预设的规划器生成对应的边缘集;
    根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
  2. 根据权利要求1所述的无人机路线规划方法,其特征在于,所述方法还包括:
    在所述顶点集包含的采样点数量小于预设的点数阈值的情况下,循环执行所述通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集的步骤。
  3. 根据权利要求1所述的无人机路线规划方法,其特征在于,所述计算所述第一采样点对应的场强值的步骤,包括:
    计算第一采样点相对于所述障碍区域的排斥力;
    根据所述目标点信息计算第一采样点相对于所述障碍区域的吸引力;
    基于所述排斥力和所述吸引力,计算所述第一采样点对应的场强值。
  4. 根据权利要求1所述的无人机路线规划方法,其特征在于,所述通过随机均匀采样方式在所述可行区域确定第一采样点的步骤之后还包括:
    计算所述第一采样点与所述顶点集包含的采样点之间的第一距离值,确定所述计算得到的第一距离值大于0。
  5. 根据权利要求1所述的无人机路线规划方法,其特征在于,所述根据所述顶点集按照预设的规划器生成对应的边缘集的步骤,包括:
    遍历所述顶点集中包含的第二采样点,对于遍历到的第二采样点,在所述顶点集中确定与所述第二采样点对应的预设点数阈值的邻点;
    通过预设的规划器连接第二采样点和邻点;
    在连接成功的情况下,将连接所述第二采样点与邻点的边添加到边缘集。
  6. 根据权利要求5所述的无人机路线规划方法,其特征在于,所述在所述顶点集中确定与所述第二采样点对应的预设点数阈值的邻点的步骤,还包括:
    计算所述第二采样点与顶点集包含的第二采样点之间的第二距离值;
    对所述第二距离值进行自小到大的排序,获取与前预设点数阈值个第二距离值对应的第二采样点作为邻点。
  7. 根据权利要求1所述的无人机路线规划方法,其特征在于,所述根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线的步骤,还包括:
    按照预设的快速搜索算法,在所述边缘集中查找距离最短的路线,将查找到的距离最短的路线作为目标路线。
  8. 一种无人机路线规划装置,其特征在于,包括:
    路线规划信息确定模块,用于获取路线规划信息,所述路线规划信息包括目标规划区域、起始点信息、目标点信息,所述目标规划区域包括可行区域和障碍区域;
    顶点集构造模块,用于通过随机均匀采样方式在所述可行区域确定第一采样点,确定所述第一采样点与所述障碍区域无碰撞,计算所述第一采样点对应的场强值,确定所述场强值大于或等于预设的场强阈值,将所述第一采样点添加到顶点集;
    边缘集构造模块,用于根据所述顶点集按照预设的规划器生成对应的边缘集;
    路线规划模块,用于根据起始点信息、目标点信息,按照预设的路径搜索算法,在所述边缘集中查找最优路线作为目标路线。
  9. 一种计算机设备,包括存储器和处理器,所述存储器存储有计算机程序,所述计算机程序被所述处理器执行时,使得所述处理器执行如权利要求1至7中任一项所述方法的步骤。
  10. 一种计算机可读存储介质,存储有计算机程序,所述计算机程序被处理器执行时,使得所述处理器执行如权利要求1至7中任一项所述方法的步骤。
PCT/CN2019/098002 2019-02-28 2019-07-26 无人机路径规划方法、装置、存储介质及计算机设备 WO2020173044A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910151761.X 2019-02-28
CN201910151761.XA CN109737967A (zh) 2019-02-28 2019-02-28 无人机路径规划方法、装置、存储介质及计算机设备

Publications (1)

Publication Number Publication Date
WO2020173044A1 true WO2020173044A1 (zh) 2020-09-03

Family

ID=66368690

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/098002 WO2020173044A1 (zh) 2019-02-28 2019-07-26 无人机路径规划方法、装置、存储介质及计算机设备

Country Status (2)

Country Link
CN (1) CN109737967A (zh)
WO (1) WO2020173044A1 (zh)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109737967A (zh) * 2019-02-28 2019-05-10 中国科学院深圳先进技术研究院 无人机路径规划方法、装置、存储介质及计算机设备
CN110262567B (zh) * 2019-06-27 2022-04-15 深圳市道通智能航空技术股份有限公司 一种路径中继点空间生成方法、装置和无人机
CN110414588A (zh) * 2019-07-23 2019-11-05 广东小天才科技有限公司 图片标注方法、装置、计算机设备和存储介质
CN111060103B (zh) * 2019-12-11 2021-12-10 安徽工程大学 一种局部动态寻优避障的路径规划方法
CN112162566B (zh) * 2020-09-04 2024-01-16 深圳市创客火科技有限公司 路线规划方法、电子设备及计算机可读存储介质
CN112669445B (zh) * 2020-12-23 2022-09-20 中国人民武装警察部队警官学院 应用于格斗对抗模拟数据的校验方法、系统及存储介质
CN112829797B (zh) * 2021-01-05 2023-01-06 北京全路通信信号研究设计院集团有限公司 一种线路点的参数获取方法、装置、设备及存储介质
CN113148507B (zh) * 2021-03-11 2022-08-26 上海电气慧程智能系统有限公司 一种生物样本存取路径的规划方法、装置及电子设备
CN113778090A (zh) * 2021-09-13 2021-12-10 上海电机学院 基于蚁群优化和prm算法的移动机器人路径规划方法
CN114010310B (zh) * 2021-09-15 2023-10-13 苏州中科华影健康科技有限公司 一种路径规划方法、装置、电子设备及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104155998A (zh) * 2014-08-27 2014-11-19 电子科技大学 一种基于势场法的航迹规划方法
CN104516356A (zh) * 2015-01-08 2015-04-15 西北工业大学 基于rrt的动态障碍规避算法
US20170241790A1 (en) * 2016-02-24 2017-08-24 Honda Motor Co., Ltd. Path plan generating apparatus for mobile body
CN108415461A (zh) * 2018-05-28 2018-08-17 济南大学 一种无人飞行器的航迹规划
CN109343528A (zh) * 2018-10-30 2019-02-15 杭州电子科技大学 一种节能的无人机路径规划避障方法
CN109737967A (zh) * 2019-02-28 2019-05-10 中国科学院深圳先进技术研究院 无人机路径规划方法、装置、存储介质及计算机设备

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101667029B1 (ko) * 2009-08-10 2016-10-17 삼성전자 주식회사 로봇의 경로 계획방법 및 장치
CN103760904B (zh) * 2014-02-13 2016-10-19 北京工业大学 一种语音播报式智能车辆路径规划装置与实施方法
CN108171796A (zh) * 2017-12-25 2018-06-15 燕山大学 一种基于三维点云的巡检机器人视觉系统及控制方法
CN108469822B (zh) * 2018-04-04 2020-12-15 天津理工大学 一种室内导盲机器人在动态环境下的路径规划方法
CN108582073B (zh) * 2018-05-02 2020-09-15 北京邮电大学 一种基于改进的随机路标地图法的机械臂快速避障方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104155998A (zh) * 2014-08-27 2014-11-19 电子科技大学 一种基于势场法的航迹规划方法
CN104516356A (zh) * 2015-01-08 2015-04-15 西北工业大学 基于rrt的动态障碍规避算法
US20170241790A1 (en) * 2016-02-24 2017-08-24 Honda Motor Co., Ltd. Path plan generating apparatus for mobile body
CN108415461A (zh) * 2018-05-28 2018-08-17 济南大学 一种无人飞行器的航迹规划
CN109343528A (zh) * 2018-10-30 2019-02-15 杭州电子科技大学 一种节能的无人机路径规划避障方法
CN109737967A (zh) * 2019-02-28 2019-05-10 中国科学院深圳先进技术研究院 无人机路径规划方法、装置、存储介质及计算机设备

Also Published As

Publication number Publication date
CN109737967A (zh) 2019-05-10

Similar Documents

Publication Publication Date Title
WO2020173044A1 (zh) 无人机路径规划方法、装置、存储介质及计算机设备
WO2020186678A1 (zh) 无人机三维地图构建方法、装置、计算机设备及存储介质
US20190332944A1 (en) Training Method, Apparatus, and Chip for Neural Network Model
US20230089380A1 (en) Neural network construction method and apparatus
CN107506786B (zh) 一种基于深度学习的属性分类识别方法
CN110221600B (zh) 路径规划方法、装置、计算机设备和存储介质
TW201832134A (zh) 人臉辨識訓練方法、裝置、電子設備、電腦可讀儲存介質及電腦程式產品
CN108229347B (zh) 用于人识别的拟吉布斯结构采样的深层置换的方法和装置
CN109871821B (zh) 自适应网络的行人重识别方法、装置、设备及存储介质
Handa et al. Scalable active matching
CN109543634B (zh) 定位过程中的数据处理方法、装置、电子设备和存储介质
CN114648640B (zh) 一种目标物单体化方法、装置、设备及存储介质
Montaut et al. Collision detection accelerated: An optimization perspective
CN115018039A (zh) 一种神经网络蒸馏方法、目标检测方法以及装置
CN110472588B (zh) 锚点框确定方法、装置、计算机设备和存储介质
US20220222852A1 (en) Methods and systems for generating end-to-end model to estimate 3-dimensional(3-d) pose of object
Sugahara et al. Exact learning augmented naive Bayes classifier
CN111177190B (zh) 数据处理方法、装置、电子设备及可读存储介质
WO2022017129A1 (zh) 目标对象检测方法、装置、电子设备及存储介质
WO2022141494A1 (zh) 三维点云聚类方法、装置、计算机设备和存储介质
CN114913330A (zh) 点云部件分割方法、装置、电子设备与存储介质
CN115586767A (zh) 一种多机器人路径规划方法和装置
KR102097505B1 (ko) 무인기의 이동 경로 생성 방법 및 장치, 컴퓨터 판독 가능한 기록 매체 및 컴퓨터 프로그램
JP2023548201A (ja) タスク学習システムおよび方法、ならびに関連デバイス
CN113723540A (zh) 一种基于多视图的无人驾驶场景聚类方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19917347

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19917347

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19917347

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 02/02/2022)