WO2022198993A1 - 一种机械臂运动规划方法、装置、可读存储介质及机械臂 - Google Patents

一种机械臂运动规划方法、装置、可读存储介质及机械臂 Download PDF

Info

Publication number
WO2022198993A1
WO2022198993A1 PCT/CN2021/124619 CN2021124619W WO2022198993A1 WO 2022198993 A1 WO2022198993 A1 WO 2022198993A1 CN 2021124619 W CN2021124619 W CN 2021124619W WO 2022198993 A1 WO2022198993 A1 WO 2022198993A1
Authority
WO
WIPO (PCT)
Prior art keywords
point
planning
path
tree
pose
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.)
Ceased
Application number
PCT/CN2021/124619
Other languages
English (en)
French (fr)
Inventor
陈金亮
刘益彰
熊友军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ubtech Robotics Corp
Original Assignee
Ubtech Robotics Corp
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 Ubtech Robotics Corp filed Critical Ubtech Robotics Corp
Publication of WO2022198993A1 publication Critical patent/WO2022198993A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Program-controlled manipulators
    • B25J9/16—Program controls
    • B25J9/1656—Program controls characterised by programming, planning systems for manipulators
    • B25J9/1664—Program controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
    • B25J9/1666—Avoiding collision or forbidden zones
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J18/00—Arms
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Program-controlled manipulators
    • B25J9/16—Program controls
    • B25J9/1679—Program controls characterised by the tasks executed

Definitions

  • the present application belongs to the field of robotics technology, and in particular relates to a method, device, computer-readable storage medium and a robotic arm for motion planning of a robotic arm.
  • a tree is generally constructed with the initial pose as the root node, and the search is randomly extended in the entire free space until the target pose is reached.
  • the embodiments of the present application provide a robotic arm motion planning method, device, computer-readable storage medium, and robotic arm, so as to solve the problem that the existing robotic arm motion planning method is relatively inefficient, takes a long time, and The obtained motion planning path often has a large gap with the optimal path.
  • a first aspect of the embodiments of the present application provides a method for planning a motion of a robotic arm, which may include:
  • a first expansion point corresponding to a first planning tree is determined in the free space according to the pose sampling points; the first planning tree is a plan extending from a preset initial pose to a preset target pose Tree;
  • a second expansion point corresponding to the second planning tree is determined in the free space according to the first expansion point;
  • the second planning tree is a planning tree expanded from the target pose to the initial pose;
  • a motion planning path of the robotic arm is determined according to the first planning tree and the second planning tree.
  • the adding the first extension point into the first planning tree according to a preset path cost function may include:
  • a node whose distance from the first expansion point in the first planning tree is less than a preset distance threshold is used as the adjacent point of the first expansion point;
  • the first path cost is the path cost from the initial pose to the adjacent point and the path cost from the adjacent point to the first extension point the sum of path costs;
  • the first extension point is added to the first planning tree by taking the adjacent point with the smallest first path cost as the parent node of the first extension point.
  • the performing optimization processing on the first planning tree may include:
  • the second path cost is from The path cost from the initial pose to the adjacent point
  • the third path cost is the path cost from the initial pose to the first extension point and the path from the adjacent point to the first extension point the sum of the costs
  • the first extension point is used as the parent node of the adjacent point.
  • the path cost function is:
  • path is the target path
  • qn is the nth node on the target path
  • 1 ⁇ n ⁇ N N is the number of nodes on the target path
  • similarity(qn,qn+1) is qn and qn+ Cosine similarity between 1
  • cost(path) is the path cost of the target path.
  • the path cost function is:
  • dist(qn, qn+1) is the Euclidean distance between qn and qn+1.
  • the path cost function is:
  • Weight1 and Weight2 are both preset weight coefficients.
  • the determining the first expansion point corresponding to the first planning tree in the free space according to the pose sampling point may include:
  • Described extension line is the connection line between described pose sampling point and described to-be-expanded node
  • the candidate point is determined as the first extension point corresponding to the first planning tree.
  • a second aspect of the embodiments of the present application provides a robotic arm motion planning device, which may include:
  • the random sampling module is used for random sampling in the free space of the manipulator to obtain the pose sampling points of the manipulator;
  • a first expansion point determination module configured to determine a first expansion point corresponding to the first planning tree in the free space according to the pose sampling point; the first planning tree is a direction from a preset initial pose to The extended planning tree of the preset target pose;
  • a first planning tree optimization module configured to add the first extension point into the first planning tree according to a preset path cost function, and perform optimization processing on the first planning tree
  • a second expansion point determining module configured to determine a second expansion point corresponding to the second planning tree in the free space according to the first expansion point; the second planning tree is a direction from the target pose to the Describe the planning tree of the initial pose expansion;
  • a second planning tree optimization module configured to add the second expansion point into the second planning tree according to the path cost function, and perform optimization processing on the second planning tree;
  • a motion planning path determination module configured to determine the motion of the robotic arm according to the first planning tree and the second planning tree when the first planning tree and the second planning tree satisfy a preset coincidence condition Motion planning path.
  • the first planning tree optimization module may include:
  • an adjacent point determination unit configured to use a node whose distance from the first expansion point in the first planning tree is less than a preset distance threshold as the adjacent point of the first expansion point;
  • the first calculation unit is used to calculate the first path cost corresponding to each adjacent point according to the path cost function; the first path cost is the path cost from the initial pose to the adjacent point and the path cost from the adjacent point to the adjacent point. the sum of the path costs of the first extension point;
  • An extension point adding unit configured to add the first extension point to the first planning tree by taking the adjacent point with the smallest first path cost as the parent node of the first extension point.
  • first planning tree optimization module may also include:
  • the second calculation unit is configured to, for each adjacent point except the parent node of the first extension point, calculate the second path cost and the third path cost corresponding to the adjacent point according to the path cost function;
  • the second path cost is the path cost from the initial pose to the adjacent point;
  • the third path cost is the path cost from the initial pose to the first extension point and the path cost from the adjacent point to the adjacent point. The sum of the path costs of the first extension point;
  • a parent node changing unit configured to use the first extension point as the parent node of the adjacent point if the third path cost is less than the second path cost.
  • the first extension point determination module may include:
  • a node to be expanded determination unit configured to use the node with the shortest distance from the pose sampling point in the first planning tree as the node to be expanded;
  • a distance calculation unit configured to calculate the distance between the pose sampling point and the node to be expanded
  • a first determining unit configured to determine the pose sampling point as the first planning tree if the distance between the pose sampling point and the node to be expanded is less than or equal to a preset step size the corresponding first extension point;
  • a candidate point determination unit configured to set the distance between the extended connection line and the node to be extended as the step if the distance between the pose sampling point and the node to be extended is greater than the step size
  • the long point is determined as a candidate point
  • the extended connection is the connection between the pose sampling point and the node to be extended
  • a second determining unit configured to determine the candidate point as a first extension point corresponding to the first planning tree if the candidate point is in the free space.
  • a third aspect of the embodiments of the present application provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, any one of the above-mentioned robotic arm motion planning methods is implemented. step.
  • a fourth aspect of the embodiments of the present application provides a robotic arm, including a memory, a processor, and a computer program stored in the memory and executable on the processor, when the processor executes the computer program The steps of implementing any one of the above-mentioned robotic arm motion planning methods.
  • a fifth aspect of the embodiments of the present application provides a computer program product, which, when the computer program product runs on a robotic arm, causes the robotic arm to execute the steps of any of the above-mentioned methods for motion planning of a robotic arm.
  • the beneficial effects of the embodiments of the present application are: the first planning tree extended from the initial pose to the target pose and the second planning tree extended from the target pose to the initial pose are used in the embodiments of the present application.
  • the bidirectional parallel expansion of the motion planning path is realized, and the path cost function is introduced in the expansion process to optimize the planning tree, so as to improve the efficiency and reduce the time-consuming, and make the obtained motion planning path closer to the optimal path. optimal path.
  • FIG. 1 is a flowchart of an embodiment of a method for planning a motion of a robotic arm in an embodiment of the present application
  • Fig. 2 is the schematic flow chart of adding the first extension point into the first planning tree
  • FIG. 3 is a structural diagram of an embodiment of a robotic arm motion planning device in an embodiment of the present application.
  • FIG. 4 is a schematic block diagram of a robotic arm in an embodiment of the present application.
  • the term “if” may be contextually interpreted as “when” or “once” or “in response to determining” or “in response to detecting” .
  • the phrases “if it is determined” or “if the [described condition or event] is detected” may be interpreted, depending on the context, to mean “once it is determined” or “in response to the determination” or “once the [described condition or event] is detected. ]” or “in response to detection of the [described condition or event]”.
  • the set composed of the poses of all joints of the manipulator can be described by the configuration space, that is, the C space, and when the manipulator and obstacles are considered, the C space will be divided into two spaces , namely the obstacle space C Obs and the free space C free .
  • C Obs can be described by the following formula:
  • C Obs is a set of poses that meet the following conditions: the pose (ie q) belongs to the C space and the robot arm (ie Robot(q)) in this pose and the obstacle (ie Obs) intersect not empty, the two collide.
  • C free can be described by the following formula:
  • C free is the complement of C Obs .
  • C free is an optional area in which the robotic arm can move and find the optimal trajectory and the shortest time.
  • the basic idea of the manipulator motion planning algorithm is to establish a path diagram for solving the manipulator motion planning problem based on C free .
  • the specific process can include:
  • the collision-free path ⁇ of the manipulator is obtained by sampling according to the motion planning algorithm of the manipulator in the free space C free , and the path must satisfy ⁇ [0,1] ⁇ C free , where 0 represents the starting time point of the motion planning , 1 represents the termination time point of the motion planning, the pose of the robot arm at any planning time point between the start time point and the termination time point of the path is within the range of the free space C free , and the path
  • an embodiment of a robotic arm motion planning method in the embodiment of the present application may include:
  • step S101 random sampling is performed in the free space of the manipulator to obtain the pose sampling points of the manipulator.
  • Step S102 Determine a first expansion point corresponding to the first planning tree in the free space according to the pose sampling points.
  • the first planning tree is a planning tree extended from a preset initial pose to a preset target pose, that is, a forward-expanding planning tree. In the initial state, the first planning tree has only one node, that is, the initial pose.
  • the node with the shortest distance from the pose sampling point in the first planning tree is used as the node to be expanded, and the distance between the pose sampling point and the node to be expanded is calculated.
  • the pose sampling point is determined as the first expansion point corresponding to the first planning tree.
  • the specific value of the step size may be set according to the actual situation, which is not specifically limited in this embodiment of the present application. If the distance between the pose sampling point and the node to be extended is greater than the step size, the point on the extension line whose distance from the node to be extended is the step size is determined as a candidate point, and the extension line here is The connection between the pose sampling point and the node to be expanded.
  • step S101 determines whether the candidate point is in the obstacle space. If the candidate point is in the obstacle space, return to step S101 to perform random sampling again; if the candidate point is in the free space, determine the candidate point as the first expansion point corresponding to the first planning tree.
  • Step S103 Add the first expansion point into the first planning tree according to the preset path cost function, and perform optimization processing on the first planning tree.
  • different path cost functions may be used to evaluate the pros and cons of paths according to actual conditions.
  • the path cost can be calculated based on cosine similarity in the path cost function, which is as follows:
  • path is the target path
  • qn is the nth node on the target path
  • 1 ⁇ n ⁇ N N is the number of nodes on the target path
  • similarity(qn,qn+1) is the difference between qn and qn+1 Cosine similarity
  • cost(path) is the path cost of the target path.
  • the path cost can be calculated based on the Euclidean distance in the path cost function, which is as follows:
  • dist(qn, qn+1) is the Euclidean distance between qn and qn+1.
  • the path cost can be calculated by combining cosine similarity and Euclidean distance in the path cost function, and the path cost function is as follows:
  • Weight1 and Weight2 are both preset weight coefficients, and their specific values may be set according to actual conditions, which are not specifically limited in this embodiment of the present application.
  • the process of adding the first extension point to the first planning tree may specifically include the following steps:
  • Step S1031 a node whose distance from the first expansion point in the first planning tree is less than a preset distance threshold is used as an adjacent point of the first expansion point.
  • the specific value of the distance threshold may be set according to the actual situation, which is not specifically limited in this embodiment of the present application. Generally, the distance threshold should be larger than the step size of each expansion.
  • Step S1032 Calculate the first path cost corresponding to each adjacent point according to the path cost function.
  • the first path cost is the sum of the path cost from the initial pose to the adjacent point and the path cost from the adjacent point to the first extension point.
  • Step S1033 taking the adjacent point with the smallest first path cost as the parent node of the first extension point, and adding the first extension point into the first planning tree.
  • the first extension point becomes a new node of the first planning tree, and the connection between the first extension point and its parent node becomes a new edge of the first planning tree.
  • the path cost of each newly generated node can be reduced as much as possible.
  • the first planning tree may be further optimized. Specifically, for each adjacent point except the parent node of the first extension point, the second path cost and the third path cost corresponding to the adjacent point are calculated according to the path cost function respectively.
  • the second path cost is the path cost from the initial pose to the adjacent point
  • the third path cost is the sum of the path cost from the initial pose to the first extension point and the path cost from the adjacent point to the first extension point and. If the third path cost is greater than or equal to the second path cost, the adjacent point does not need to be optimized; if the third path cost is less than the second path cost, the first extension point is used as the parent node of the adjacent point, that is, the adjacent point is deleted. The edge between the adjacent point and its original parent node, and the connection between the first extension point and the adjacent point becomes a new edge of the first planning tree.
  • path optimization can be performed to reduce the path cost of some nodes. From an overall perspective, not every node optimized for the path will appear in the final generated path, but in the process of generating the random tree, each path optimization creates an opportunity to reduce the final path cost as much as possible .
  • Step S104 Determine a second extension point corresponding to the second planning tree in the free space according to the first extension point.
  • the second planning tree is a planning tree extended from the target pose to the initial pose, that is, a planning tree extended in reverse. In the initial state, the second planning tree has only one node, that is, the target pose.
  • the node with the shortest distance from the first expansion point in the second planning tree is used as the node to be expanded, and the distance between the first expansion point and the node to be expanded is calculated.
  • the first extension point is determined as the second extension point corresponding to the second planning tree. If the distance between the first extension point and the node to be extended is greater than the step size, the point on the extension line whose distance from the node to be extended is the step size is determined as a candidate point, and the extension line here is The connection between the first expansion point and the node to be expanded.
  • step S101 If the candidate point is in the obstacle space, return to step S101 to perform random sampling again; if the candidate point is in the free space, the candidate point is determined as the second expansion point corresponding to the second planning tree.
  • Step S105 Add the second expansion point into the second planning tree according to the path cost function, and perform optimization processing on the second planning tree.
  • Step S105 is similar to step S103, and the specific process can refer to the detailed description in step S103, which will not be repeated here.
  • Step S106 when the first planning tree and the second planning tree satisfy the preset coincidence condition, determine the motion planning path of the robotic arm according to the first planning tree and the second planning tree.
  • the coincidence condition is that the first planning tree and the second planning tree are connected, that is, the two have the same node.
  • the coincidence condition is not met, return to step S101 to re-sample randomly; when this condition is met, the first planning tree and the second planning tree are combined into a whole, because the first planning tree is from the initial pose to the target pose
  • the second planning tree is extended from the target pose to the initial pose. After the two are combined during the expansion, a complete path from the initial pose to the target pose will be formed, which is the motion planning path of the robotic arm. .
  • the embodiment of the present application realizes the bidirectional parallel expansion of the motion planning path through the first planning tree extending from the initial pose to the target pose and the second planning tree extending from the target pose to the initial pose, and In the expansion process, the path cost function is introduced to optimize the planning tree, so that the obtained motion planning path is closer to the optimal path while improving the efficiency and reducing the time-consuming.
  • FIG. 3 shows a structural diagram of an embodiment of a robotic arm motion planning apparatus provided in an embodiment of the present application.
  • a robotic arm motion planning device may include:
  • the random sampling module 301 is used to perform random sampling in the free space of the manipulator to obtain the pose sampling points of the manipulator;
  • a first expansion point determination module 302 configured to determine a first expansion point corresponding to a first planning tree in the free space according to the pose sampling points; the first planning tree is a preset initial pose The planning tree extended to the preset target pose;
  • a first planning tree optimization module 303 configured to add the first extension point into the first planning tree according to a preset path cost function, and perform optimization processing on the first planning tree;
  • the second expansion point determining module 304 is configured to determine a second expansion point corresponding to the second planning tree in the free space according to the first expansion point; the second planning tree is a direction from the target pose to the the expanded planning tree of the initial pose;
  • a second planning tree optimization module 305 configured to add the second expansion point into the second planning tree according to the path cost function, and perform optimization processing on the second planning tree;
  • a motion planning path determination module 306 configured to determine the robotic arm according to the first planning tree and the second planning tree when the first planning tree and the second planning tree satisfy a preset coincidence condition motion planning path.
  • the first planning tree optimization module may include:
  • an adjacent point determination unit configured to use a node whose distance from the first expansion point in the first planning tree is less than a preset distance threshold as the adjacent point of the first expansion point;
  • the first calculation unit is used to calculate the first path cost corresponding to each adjacent point according to the path cost function; the first path cost is the path cost from the initial pose to the adjacent point and the path cost from the adjacent point to the adjacent point. the sum of the path costs of the first extension point;
  • An extension point adding unit configured to add the first extension point to the first planning tree by taking the adjacent point with the smallest first path cost as the parent node of the first extension point.
  • first planning tree optimization module may also include:
  • the second calculation unit is configured to, for each adjacent point except the parent node of the first extension point, calculate the second path cost and the third path cost corresponding to the adjacent point according to the path cost function;
  • the second path cost is the path cost from the initial pose to the adjacent point;
  • the third path cost is the path cost from the initial pose to the first extension point and the path cost from the adjacent point to the adjacent point. The sum of the path costs of the first extension point;
  • a parent node changing unit configured to use the first extension point as the parent node of the adjacent point if the third path cost is less than the second path cost.
  • the first extension point determination module may include:
  • a node to be expanded determination unit configured to use the node with the shortest distance from the pose sampling point in the first planning tree as the node to be expanded;
  • a distance calculation unit configured to calculate the distance between the pose sampling point and the node to be expanded
  • a first determining unit configured to determine the pose sampling point as the first planning tree if the distance between the pose sampling point and the node to be expanded is less than or equal to a preset step size the corresponding first extension point;
  • a candidate point determination unit configured to set the distance between the extended connection line and the node to be extended as the step if the distance between the pose sampling point and the node to be extended is greater than the step size
  • the long point is determined as a candidate point
  • the extended connection is the connection between the pose sampling point and the node to be extended
  • a second determining unit configured to determine the candidate point as a first extension point corresponding to the first planning tree if the candidate point is in the free space.
  • FIG. 4 shows a schematic block diagram of a robotic arm provided by an embodiment of the present application. For convenience of description, only parts related to the embodiment of the present application are shown.
  • the robotic arm 4 of this embodiment includes: a processor 40 , a memory 41 , and a computer program 42 stored in the memory 41 and executable on the processor 40 .
  • the processor 40 executes the computer program 42
  • the steps in each of the above embodiments of the robotic arm motion planning method are implemented, for example, steps S101 to S106 shown in FIG. 1 .
  • the processor 40 executes the computer program 42
  • the functions of the modules/units in the above device embodiments for example, the functions of the modules 301 to 306 shown in FIG. 3 are implemented.
  • the computer program 42 may be divided into one or more modules/units, and the one or more modules/units are stored in the memory 41 and executed by the processor 40 to complete the this application.
  • the one or more modules/units may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program 42 in the robotic arm 4 .
  • FIG. 4 is only an example of the robot arm 4, and does not constitute a limitation to the robot arm 4. It may include more or less components than the one shown, or combine some components, or different components
  • the robotic arm 4 may further include an input and output device, a network access device, a bus, and the like.
  • the processor 40 may be a central processing unit (Central Processing Unit, CPU), or other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
  • the storage 41 may be an internal storage unit of the robotic arm 4 , such as a hard disk or a memory of the robotic arm 4 .
  • the memory 41 can also be an external storage device of the robotic arm 4, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) equipped on the robotic arm 4 Card, Flash Card, etc. Further, the memory 41 may also include both an internal storage unit of the robotic arm 4 and an external storage device.
  • the memory 41 is used to store the computer program and other programs and data required by the robotic arm 4 .
  • the memory 41 can also be used to temporarily store data that has been output or will be output.
  • the disclosed apparatus/robot and method may be implemented in other ways.
  • the device/manipulator embodiments described above are only illustrative.
  • the division of modules or units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units. Or components may be combined or may be integrated into another system, or some features may be omitted, or not implemented.
  • the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.
  • the integrated modules/units if implemented in the form of software functional units and sold or used as independent products, may be stored in a computer-readable storage medium.
  • the present application can implement all or part of the processes in the methods of the above embodiments, and can also be completed by instructing the relevant hardware through a computer program.
  • the computer program can be stored in a computer-readable storage medium, and the computer When the program is executed by the processor, the steps of the foregoing method embodiments can be implemented.
  • the computer program includes computer program code, and the computer program code may be in the form of source code, object code, executable file or some intermediate form, and the like.
  • the computer-readable storage medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory) ), random access memory (RAM, Random Access Memory), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable storage medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, computer-readable Storage media exclude electrical carrier signals and telecommunications signals.

Landscapes

  • Engineering & Computer Science (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Manipulator (AREA)

Abstract

本申请属于机器人技术领域，尤其涉及一种机械臂运动规划方法、装置、计算机可读存储介质及机械臂。所述方法通过从初始位姿向目标位姿扩展的第一规划树和从目标位姿向初始位姿扩展的第二规划树实现了运动规划路径的双向并进扩展，且在扩展过程中引入了路径代价函数来对规划树进行优化，从而在提升效率，减少耗时的同时，使得所得到的运动规划路径更加趋近于最优路径。

Description

一种机械臂运动规划方法、装置、可读存储介质及机械臂
本申请要求于2021年03月22日在中国专利局提交的、申请号为202110301129.6的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请属于机器人技术领域，尤其涉及一种机械臂运动规划方法、装置、计算机可读存储介质及机械臂。
背景技术
现有技术在对机械臂进行运动规划时，一般会构建一棵将初始位姿作为根节点的树，在整个自由空间中随机扩展搜索，直至到达目标位姿为止。这种方式虽然可以得到机械臂的运动规划路径，但是效率较为低下，耗时极长，且所得到的运动规划路径往往与最优路径存在较大的差距。
技术问题
有鉴于此，本申请实施例提供了一种机械臂运动规划方法、装置、计算机可读存储介质及机械臂，以解决现有的机械臂运动规划方法效率较为低下，耗时极长，且所得到的运动规划路径往往与最优路径存在较大的差距的问题。
技术解决方案
本申请实施例的第一方面提供了一种机械臂运动规划方法，可以包括：
在机械臂的自由空间中进行随机采样，得到所述机械臂的位姿采样点；
根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点；所述第一规划树为从预设的初始位姿向预设的目标位姿扩展的规划树；
根据预设的路径代价函数将所述第一扩展点添加入所述第一规划树中，并对所述第一规划树进行优化处理；
根据所述第一扩展点在所述自由空间中确定与第二规划树对应的第二扩展点；所述第二规划树为从所述目标位姿向所述初始位姿扩展的规划树；
根据所述路径代价函数将所述第二扩展点添加入所述第二规划树中，并对所述第二规划树进行优化处理；
当所述第一规划树与所述第二规划树满足预设的重合条件时，根据所述第一规划树与所述第二规划树确定所述机械臂的运动规划路径。
进一步地，所述根据预设的路径代价函数将所述第一扩展点添加入所述第一规划树中，可以包括：
将所述第一规划树中与所述第一扩展点的距离小于预设的距离阈值的节点作为所述第 一扩展点的临近点；
根据所述路径代价函数分别计算与各个临近点对应的第一路径代价；所述第一路径代价为从所述初始位姿至临近点的路径代价与从临近点至所述第一扩展点的路径代价之和；
以第一路径代价最小的临近点作为所述第一扩展点的父节点，将所述第一扩展点添加入所述第一规划树中。
进一步地，所述对所述第一规划树进行优化处理，可以包括：
对于除所述第一扩展点的父节点之外的各个临近点，分别根据所述路径代价函数计算与该临近点对应的第二路径代价和第三路径代价；所述第二路径代价为从所述初始位姿至该临近点的路径代价；所述第三路径代价为从所述初始位姿至所述第一扩展点的路径代价与从该临近点至所述第一扩展点的路径代价之和；
若所述第三路径代价小于所述第二路径代价，则将所述第一扩展点作为该临近点的父节点。
可选地，所述路径代价函数为：
Figure PCTCN2021124619-appb-000001
其中，path为目标路径，qn为所述目标路径上的第n个节点，1≤n≤N，N为所述目标路径上的节点数目，similarity(qn,qn+1)为qn与qn+1之间的余弦相似度，cost(path)为所述目标路径的路径代价。
可选地，所述路径代价函数为：
Figure PCTCN2021124619-appb-000002
其中，dist(qn,qn+1)为qn与qn+1之间的欧氏距离。
优选地，所述路径代价函数为：
Figure PCTCN2021124619-appb-000003
其中，Weight1和Weight2均为预设的权重系数。
进一步地，所述根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点，可以包括：
将所述第一规划树中与所述位姿采样点的距离最短的节点作为待扩展节点；
计算所述位姿采样点与所述待扩展节点之间的距离；
若所述位姿采样点与所述待扩展节点之间的距离小于或等于预设的步长，则将所述位姿采样点确定为与所述第一规划树对应的第一扩展点；
若所述位姿采样点与所述待扩展节点之间的距离大于所述步长，则将扩展连线上与所 述待扩展节点之间的距离为所述步长的点确定为候选点；所述扩展连线为所述位姿采样点与所述待扩展节点之间的连线；
若所述候选点在所述自由空间内，则将所述候选点确定为与所述第一规划树对应的第一扩展点。
本申请实施例的第二方面提供了一种机械臂运动规划装置，可以包括：
随机采样模块，用于在机械臂的自由空间中进行随机采样，得到所述机械臂的位姿采样点；
第一扩展点确定模块，用于根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点；所述第一规划树为从预设的初始位姿向预设的目标位姿扩展的规划树；
第一规划树优化模块，用于根据预设的路径代价函数将所述第一扩展点添加入所述第一规划树中，并对所述第一规划树进行优化处理；
第二扩展点确定模块，用于根据所述第一扩展点在所述自由空间中确定与第二规划树对应的第二扩展点；所述第二规划树为从所述目标位姿向所述初始位姿扩展的规划树；
第二规划树优化模块，用于根据所述路径代价函数将所述第二扩展点添加入所述第二规划树中，并对所述第二规划树进行优化处理；
运动规划路径确定模块，用于当所述第一规划树与所述第二规划树满足预设的重合条件时，根据所述第一规划树与所述第二规划树确定所述机械臂的运动规划路径。
进一步地，所述第一规划树优化模块可以包括：
临近点确定单元，用于将所述第一规划树中与所述第一扩展点的距离小于预设的距离阈值的节点作为所述第一扩展点的临近点；
第一计算单元，用于根据所述路径代价函数分别计算与各个临近点对应的第一路径代价；所述第一路径代价为从所述初始位姿至临近点的路径代价与从临近点至所述第一扩展点的路径代价之和；
扩展点添加单元，用于以第一路径代价最小的临近点作为所述第一扩展点的父节点，将所述第一扩展点添加入所述第一规划树中。
进一步地，所述第一规划树优化模块还可以包括：
第二计算单元，用于对于除所述第一扩展点的父节点之外的各个临近点，分别根据所述路径代价函数计算与该临近点对应的第二路径代价和第三路径代价；所述第二路径代价为从所述初始位姿至该临近点的路径代价；所述第三路径代价为从所述初始位姿至所述第一扩展点的路径代价与从该临近点至所述第一扩展点的路径代价之和；
父节点变更单元，用于若所述第三路径代价小于所述第二路径代价，则将所述第一扩展点作为该临近点的父节点。
进一步地，所述第一扩展点确定模块可以包括：
待扩展节点确定单元，用于将所述第一规划树中与所述位姿采样点的距离最短的节点作为待扩展节点；
距离计算单元，用于计算所述位姿采样点与所述待扩展节点之间的距离；
第一确定单元，用于若所述位姿采样点与所述待扩展节点之间的距离小于或等于预设的步长，则将所述位姿采样点确定为与所述第一规划树对应的第一扩展点；
候选点确定单元，用于若所述位姿采样点与所述待扩展节点之间的距离大于所述步长，则将扩展连线上与所述待扩展节点之间的距离为所述步长的点确定为候选点；所述扩展连线为所述位姿采样点与所述待扩展节点之间的连线；
第二确定单元，用于若所述候选点在所述自由空间内，则将所述候选点确定为与所述第一规划树对应的第一扩展点。
本申请实施例的第三方面提供了一种计算机可读存储介质，所述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实现上述任一种机械臂运动规划方法的步骤。
本申请实施例的第四方面提供了一种机械臂，包括存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现上述任一种机械臂运动规划方法的步骤。
本申请实施例的第五方面提供了一种计算机程序产品，当计算机程序产品在机械臂上运行时，使得机械臂执行上述任一种机械臂运动规划方法的步骤。
有益效果
本申请实施例与现有技术相比存在的有益效果是：本申请实施例通过从初始位姿向目标位姿扩展的第一规划树和从目标位姿向初始位姿扩展的第二规划树实现了运动规划路径的双向并进扩展，且在扩展过程中引入了路径代价函数来对规划树进行优化，从而在提升效率，减少耗时的同时，使得所得到的运动规划路径更加趋近于最优路径。
附图说明
为了更清楚地说明本申请实施例中的技术方案，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申请的一些实施例，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它的附图。
图1为本申请实施例中一种机械臂运动规划方法的一个实施例流程图；
图2为将第一扩展点添加入第一规划树中的示意流程图；
图3为本申请实施例中一种机械臂运动规划装置的一个实施例结构图；
图4为本申请实施例中一种机械臂的示意框图。
本发明的实施方式
为使得本申请的发明目的、特征、优点能够更加的明显和易懂，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，下面所描述的实施例仅仅是本申请一部分实施例，而非全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其它实施例，都属于本申请保护的范围。
应当理解，当在本说明书和所附权利要求书中使用时，术语“包括”指示所描述特征、整体、步骤、操作、元素和/或组件的存在，但并不排除一个或多个其它特征、整体、步骤、操作、元素、组件和/或其集合的存在或添加。
还应当理解，在此本申请说明书中所使用的术语仅仅是出于描述特定实施例的目的而并不意在限制本申请。如在本申请说明书和所附权利要求书中所使用的那样，除非上下文清楚地指明其它情况，否则单数形式的“一”、“一个”及“该”意在包括复数形式。
还应当进一步理解，在本申请说明书和所附权利要求书中使用的术语“和/或”是指相关联列出的项中的一个或多个的任何组合以及所有可能组合，并且包括这些组合。
如在本说明书和所附权利要求书中所使用的那样，术语“如果”可以依据上下文被解释为“当...时”或“一旦”或“响应于确定”或“响应于检测到”。类似地，短语“如果确定”或“如果检测到[所描述条件或事件]”可以依据上下文被解释为意指“一旦确定”或“响应于确定”或“一旦检测到[所描述条件或事件]”或“响应于检测到[所描述条件或事件]”。
另外，在本申请的描述中，术语“第一”、“第二”、“第三”等仅用于区分描述，而不能理解为指示或暗示相对重要性。
在本申请实施例中，可以通过位姿空间(configuration space)描述机械臂所有关节的位姿构成的集合，即C空间，而当考虑机械臂和障碍物时，C空间会分为两个空间，即障碍物空间C Obs和自由空间C free。
其中，C Obs可以用下式进行描述：
Figure PCTCN2021124619-appb-000004
即C Obs为满足以下条件的位姿构成的集合：该位姿(即q)属于C空间且在该位姿下的机械臂(即Robot(q))与障碍物(即Obs)相交不为空，两者发生碰撞。
相应地，C free可以用下式进行描述：
C free＝C-C Obs
即C free为C Obs的补集。
C free是机械臂能够在其中运动并发现轨迹最优、时间最短的可选区域。机械臂运动规划算法的基本思想是基于C free建立用于解决机械臂运动规划问题的路径图，其具体流程可以包括：
(1)确定机械臂的工作空间W；
(2)确定工作空间W中的障碍物Obs以及机械臂Robot；
(3)确定机械臂对应的C空间，以及障碍物空间C Obs和自由空间C free；
(4)确定机械臂的初始位姿q init和目标位姿q en d；
(5)在自由空间C free中根据机械臂运动规划算法采样获得机械臂的无碰撞路径τ，该路径需满足τ[0,1]→C free，其中，0表示运动规划的起始时间点，1表示运动规划的终止时间点，该路径在从起始时间点至终止时间点之间的任意一个规划时间点时的机械臂位姿均处在自由空间C free的范围内，且该路径在起始时间点时的机械臂位姿为初始位姿，即τ[0]＝q init，该路径在终止时间点时的机械臂位姿为目标位姿，即τ[1]＝q end。
请参阅图1，本申请实施例中一种机械臂运动规划方法的一个实施例可以包括：
步骤S101、在机械臂的自由空间中进行随机采样，得到机械臂的位姿采样点。
步骤S102、根据位姿采样点在自由空间中确定与第一规划树对应的第一扩展点。
其中，第一规划树为从预设的初始位姿向预设的目标位姿扩展的规划树，也即正向扩展的规划树。在起始状态下，第一规划树仅有一个节点，即初始位姿。
首先，将第一规划树中与位姿采样点的距离最短的节点作为待扩展节点，并计算位姿采样点与待扩展节点之间的距离。
若位姿采样点与待扩展节点之间的距离小于或等于预设的步长，则将位姿采样点确定为与第一规划树对应的第一扩展点。该步长的具体取值可以根据实际情况进行设置，本申请实施例对其不作具体限定。若位姿采样点与待扩展节点之间的距离大于该步长，则将扩展连线上与待扩展节点之间的距离为该步长的点确定为候选点，此处的扩展连线为位姿采样点与待扩展节点之间的连线。
若候选点在障碍物空间内，则返回步骤S101重新进行随机采样；若候选点在自由空间内，则将候选点确定为与第一规划树对应的第一扩展点。
步骤S103、根据预设的路径代价函数将第一扩展点添加入第一规划树中，并对第一规划树进行优化处理。
在本申请实施例中，可以根据实际情况使用不同的路径代价函数来对路径的优劣进行评价。
可选地，在路径代价函数中可以基于余弦相似度来计算路径代价，该路径代价函数如 下所示：
Figure PCTCN2021124619-appb-000005
其中，path为目标路径，qn为目标路径上的第n个节点，1≤n≤N，N为目标路径上的节点数目，similarity(qn,qn+1)为qn与qn+1之间的余弦相似度，cost(path)为目标路径的路径代价。
可选地，在路径代价函数中可以基于欧式距离来计算路径代价，该路径代价函数如下所示：
Figure PCTCN2021124619-appb-000006
其中，dist(qn,qn+1)为qn与qn+1之间的欧氏距离。
优选地，在路径代价函数中可以结合余弦相似度和欧式距离来计算路径代价，该路径代价函数如下所示：
Figure PCTCN2021124619-appb-000007
其中，Weight1和Weight2均为预设的权重系数，其具体取值可以根据实际情况进行设置，本申请实施例对其不作具体限定。
如图2所示，将第一扩展点添加入第一规划树中的过程具体可以包括如下步骤：
步骤S1031、将第一规划树中与第一扩展点的距离小于预设的距离阈值的节点作为第一扩展点的临近点。
距离阈值的具体取值可以根据实际情况进行设置，本申请实施例对其不作具体限定。一般地，该距离阈值应大于每次扩展的步长。
步骤S1032、根据路径代价函数分别计算与各个临近点对应的第一路径代价。
其中，第一路径代价为从初始位姿至临近点的路径代价与从临近点至第一扩展点的路径代价之和。
步骤S1033、以第一路径代价最小的临近点作为第一扩展点的父节点，将第一扩展点添加入第一规划树中。
此时，第一扩展点成为第一规划树的一个新节点，而第一扩展点与其父节点的连线成为第一规划树的一条新边。通过这样的节点添加过程，可以尽可能地减小每次新生成的节点的路径代价。
优选地，在将第一扩展点添加入第一规划树中之后，还可以进一步对第一规划树进行优化处理。具体地，对于除第一扩展点的父节点之外的各个临近点，分别根据路径代价函数计算与该临近点对应的第二路径代价和第三路径代价。其中，第二路径代价为从初始位 姿至该临近点的路径代价，第三路径代价为从初始位姿至第一扩展点的路径代价与从该临近点至第一扩展点的路径代价之和。若第三路径代价大于或等于第二路径代价，则无需对该临近点进行优化；若第三路径代价小于第二路径代价，则将第一扩展点作为该临近点的父节点，即删除该临近点与其原来的父节点之间的边，并将第一扩展点与该临近点的连线成为第一规划树的一条新边。
通过这样的优化处理，每当生成了新的节点后，可以通过路径优化，使得某些节点的路径代价减少。如果以整体的眼光看，并不是每一个路径优化的节点都会出现在最终生成的路径中，但在生成随机树的过程中，每一次的路径优化都尽可能的为最终路径代价减小创造机会。
步骤S104、根据第一扩展点在自由空间中确定与第二规划树对应的第二扩展点。
其中，第二规划树为从目标位姿向初始位姿扩展的规划树，也即反向扩展的规划树。在起始状态下，第二规划树仅有一个节点，即目标位姿。
首先，将第二规划树中与第一扩展点的距离最短的节点作为待扩展节点，并计算第一扩展点与待扩展节点之间的距离。
若第一扩展点与待扩展节点之间的距离小于或等于预设的步长，则将第一扩展点确定为与第二规划树对应的第二扩展点。若第一扩展点与待扩展节点之间的距离大于该步长，则将扩展连线上与待扩展节点之间的距离为该步长的点确定为候选点，此处的扩展连线为第一扩展点与待扩展节点之间的连线。
若候选点在障碍物空间内，则返回步骤S101重新进行随机采样；若候选点在自由空间内，则将候选点确定为与第二规划树对应的第二扩展点。
步骤S105、根据路径代价函数将第二扩展点添加入第二规划树中，并对第二规划树进行优化处理。
步骤S105与步骤S103类似，具体的过程可以参见步骤S103中的详细叙述，此处不再赘述。
步骤S106、当第一规划树与第二规划树满足预设的重合条件时，根据第一规划树与第二规划树确定机械臂的运动规划路径。
该重合条件为第一规划树与第二规划树产生连接，即两者存在相同的节点。当不满足该重合条件时，返回步骤S101重新进行随机采样；当满足该条件时，第一规划树与第二规划树结合成一个整体，由于第一规划树是从初始位姿向目标位姿扩展，而第二规划树是从目标位姿向初始位姿扩展，两者在扩展途中结合后将形成一条从初始位姿到目标位姿的完整路径，该路径即为机械臂的运动规划路径。
综上所述，本申请实施例通过从初始位姿向目标位姿扩展的第一规划树和从目标位姿 向初始位姿扩展的第二规划树实现了运动规划路径的双向并进扩展，且在扩展过程中引入了路径代价函数来对规划树进行优化，从而在提升效率，减少耗时的同时，使得所得到的运动规划路径更加趋近于最优路径。
应理解，上述实施例中各步骤的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本申请实施例的实施过程构成任何限定。
对应于上文实施例所述的一种机械臂运动规划方法，图3示出了本申请实施例提供的一种机械臂运动规划装置的一个实施例结构图。
本实施例中，一种机械臂运动规划装置可以包括：
随机采样模块301，用于在机械臂的自由空间中进行随机采样，得到所述机械臂的位姿采样点；
第一扩展点确定模块302，用于根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点；所述第一规划树为从预设的初始位姿向预设的目标位姿扩展的规划树；
第一规划树优化模块303，用于根据预设的路径代价函数将所述第一扩展点添加入所述第一规划树中，并对所述第一规划树进行优化处理；
第二扩展点确定模块304，用于根据所述第一扩展点在所述自由空间中确定与第二规划树对应的第二扩展点；所述第二规划树为从所述目标位姿向所述初始位姿扩展的规划树；
第二规划树优化模块305，用于根据所述路径代价函数将所述第二扩展点添加入所述第二规划树中，并对所述第二规划树进行优化处理；
运动规划路径确定模块306，用于当所述第一规划树与所述第二规划树满足预设的重合条件时，根据所述第一规划树与所述第二规划树确定所述机械臂的运动规划路径。
进一步地，所述第一规划树优化模块可以包括：
临近点确定单元，用于将所述第一规划树中与所述第一扩展点的距离小于预设的距离阈值的节点作为所述第一扩展点的临近点；
第一计算单元，用于根据所述路径代价函数分别计算与各个临近点对应的第一路径代价；所述第一路径代价为从所述初始位姿至临近点的路径代价与从临近点至所述第一扩展点的路径代价之和；
扩展点添加单元，用于以第一路径代价最小的临近点作为所述第一扩展点的父节点，将所述第一扩展点添加入所述第一规划树中。
进一步地，所述第一规划树优化模块还可以包括：
第二计算单元，用于对于除所述第一扩展点的父节点之外的各个临近点，分别根据所述路径代价函数计算与该临近点对应的第二路径代价和第三路径代价；所述第二路径代价 为从所述初始位姿至该临近点的路径代价；所述第三路径代价为从所述初始位姿至所述第一扩展点的路径代价与从该临近点至所述第一扩展点的路径代价之和；
父节点变更单元，用于若所述第三路径代价小于所述第二路径代价，则将所述第一扩展点作为该临近点的父节点。
进一步地，所述第一扩展点确定模块可以包括：
待扩展节点确定单元，用于将所述第一规划树中与所述位姿采样点的距离最短的节点作为待扩展节点；
距离计算单元，用于计算所述位姿采样点与所述待扩展节点之间的距离；
第一确定单元，用于若所述位姿采样点与所述待扩展节点之间的距离小于或等于预设的步长，则将所述位姿采样点确定为与所述第一规划树对应的第一扩展点；
候选点确定单元，用于若所述位姿采样点与所述待扩展节点之间的距离大于所述步长，则将扩展连线上与所述待扩展节点之间的距离为所述步长的点确定为候选点；所述扩展连线为所述位姿采样点与所述待扩展节点之间的连线；
第二确定单元，用于若所述候选点在所述自由空间内，则将所述候选点确定为与所述第一规划树对应的第一扩展点。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的装置，模块和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述或记载的部分，可以参见其它实施例的相关描述。
图4示出了本申请实施例提供的一种机械臂的示意框图，为了便于说明，仅示出了与本申请实施例相关的部分。
如图4所示，该实施例的机械臂4包括：处理器40、存储器41以及存储在所述存储器41中并可在所述处理器40上运行的计算机程序42。所述处理器40执行所述计算机程序42时实现上述各个机械臂运动规划方法实施例中的步骤，例如图1所示的步骤S101至步骤S106。或者，所述处理器40执行所述计算机程序42时实现上述各装置实施例中各模块/单元的功能，例如图3所示模块301至模块306的功能。
示例性的，所述计算机程序42可以被分割成一个或多个模块/单元，所述一个或者多个模块/单元被存储在所述存储器41中，并由所述处理器40执行，以完成本申请。所述一个或多个模块/单元可以是能够完成特定功能的一系列计算机程序指令段，该指令段用于描述所述计算机程序42在所述机械臂4中的执行过程。
本领域技术人员可以理解，图4仅仅是机械臂4的示例，并不构成对机械臂4的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件，例如所述机械 臂4还可以包括输入输出设备、网络接入设备、总线等。
所述处理器40可以是中央处理单元(Central Processing Unit，CPU)，还可以是其它通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)或者其它可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
所述存储器41可以是所述机械臂4的内部存储单元，例如机械臂4的硬盘或内存。所述存储器41也可以是所述机械臂4的外部存储设备，例如所述机械臂4上配备的插接式硬盘，智能存储卡(Smart Media Card，SMC)，安全数字(Secure Digital，SD)卡，闪存卡(Flash Card)等。进一步地，所述存储器41还可以既包括所述机械臂4的内部存储单元也包括外部存储设备。所述存储器41用于存储所述计算机程序以及所述机械臂4所需的其它程序和数据。所述存储器41还可以用于暂时地存储已经输出或者将要输出的数据。
所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，仅以上述各功能单元、模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能单元、模块完成，即将所述装置的内部结构划分成不同的功能单元或模块，以完成以上描述的全部或者部分功能。实施例中的各功能单元、模块可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中，上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。另外，各功能单元、模块的具体名称也只是为了便于相互区分，并不用于限制本申请的保护范围。上述系统中单元、模块的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述或记载的部分，可以参见其它实施例的相关描述。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
在本申请所提供的实施例中，应该理解到，所揭露的装置/机械臂和方法，可以通过其它的方式实现。例如，以上所描述的装置/机械臂实施例仅仅是示意性的，例如，所述模块或单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个 单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通讯连接可以是通过一些接口，装置或单元的间接耦合或通讯连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的模块/单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读存储介质中。基于这样的理解，本申请实现上述实施例方法中的全部或部分流程，也可以通过计算机程序来指令相关的硬件来完成，所述的计算机程序可存储于一计算机可读存储介质中，该计算机程序在被处理器执行时，可实现上述各个方法实施例的步骤。其中，所述计算机程序包括计算机程序代码，所述计算机程序代码可以为源代码形式、对象代码形式、可执行文件或某些中间形式等。所述计算机可读存储介质可以包括：能够携带所述计算机程序代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、电载波信号、电信信号以及软件分发介质等。需要说明的是，所述计算机可读存储介质包含的内容可以根据司法管辖区内立法和专利实践的要求进行适当的增减，例如在某些司法管辖区，根据立法和专利实践，计算机可读存储介质不包括电载波信号和电信信号。
以上所述实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围，均应包含在本申请的保护范围之内。

Claims (10)

  1. 一种机械臂运动规划方法，其特征在于，包括：
    在机械臂的自由空间中进行随机采样，得到所述机械臂的位姿采样点；
    根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点；所述第一规划树为从预设的初始位姿向预设的目标位姿扩展的规划树；
    根据预设的路径代价函数将所述第一扩展点添加入所述第一规划树中，并对所述第一规划树进行优化处理；
    根据所述第一扩展点在所述自由空间中确定与第二规划树对应的第二扩展点；所述第二规划树为从所述目标位姿向所述初始位姿扩展的规划树；
    根据所述路径代价函数将所述第二扩展点添加入所述第二规划树中，并对所述第二规划树进行优化处理；
    当所述第一规划树与所述第二规划树满足预设的重合条件时，根据所述第一规划树与所述第二规划树确定所述机械臂的运动规划路径。
  2. 根据权利要求1所述的机械臂运动规划方法，其特征在于，所述根据预设的路径代价函数将所述第一扩展点添加入所述第一规划树中，包括：
    将所述第一规划树中与所述第一扩展点的距离小于预设的距离阈值的节点作为所述第一扩展点的临近点；
    根据所述路径代价函数分别计算与各个临近点对应的第一路径代价；所述第一路径代价为从所述初始位姿至临近点的路径代价与从临近点至所述第一扩展点的路径代价之和；
    以第一路径代价最小的临近点作为所述第一扩展点的父节点，将所述第一扩展点添加入所述第一规划树中。
  3. 根据权利要求2所述的机械臂运动规划方法，其特征在于，所述对所述第一规划树进行优化处理，包括：
    对于除所述第一扩展点的父节点之外的各个临近点，分别根据所述路径代价函数计算与该临近点对应的第二路径代价和第三路径代价；所述第二路径代价为从所述初始位姿至该临近点的路径代价；所述第三路径代价为从所述初始位姿至所述第一扩展点的路径代价与从该临近点至所述第一扩展点的路径代价之和；
    若所述第三路径代价小于所述第二路径代价，则将所述第一扩展点作为该临近点的父节点。
  4. 根据权利要求1所述的机械臂运动规划方法，其特征在于，所述路径代价函数为：
    Figure PCTCN2021124619-appb-100001
    其中，path为目标路径，qn为所述目标路径上的第n个节点，1≤n≤N，N为所述目标路径上的节点数目，similarity(qn,qn+1)为qn与qn+1之间的余弦相似度，cost(path)为所述目标路径的路径代价。
  5. 根据权利要求1所述的机械臂运动规划方法，其特征在于，所述路径代价函数为：
    Figure PCTCN2021124619-appb-100002
    其中，path为目标路径，qn为所述目标路径上的第n个节点，1≤n≤N，N为所述目标路径上的节点数目，dist(qn,qn+1)为qn与qn+1之间的欧氏距离，cost(path)为所述目标路径的路径代价。
  6. 根据权利要求1所述的机械臂运动规划方法，其特征在于，所述路径代价函数为：
    Figure PCTCN2021124619-appb-100003
    其中，path为目标路径，qn为所述目标路径上的第n个节点，1≤n≤N，N为所述目标路径上的节点数目，similarity(qn,qn+1)为qn与qn+1之间的余弦相似度，dist(qn,qn+1)为qn与qn+1之间的欧氏距离，Weight1和Weight2均为预设的权重系数，cost(path)为所述目标路径的路径代价。
  7. 根据权利要求1至6中任一项所述的机械臂运动规划方法，其特征在于，所述根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点，包括：
    将所述第一规划树中与所述位姿采样点的距离最短的节点作为待扩展节点；
    计算所述位姿采样点与所述待扩展节点之间的距离；
    若所述位姿采样点与所述待扩展节点之间的距离小于或等于预设的步长，则将所述位姿采样点确定为与所述第一规划树对应的第一扩展点；
    若所述位姿采样点与所述待扩展节点之间的距离大于所述步长，则将扩展连线上与所述待扩展节点之间的距离为所述步长的点确定为候选点；所述扩展连线为所述位姿采样点与所述待扩展节点之间的连线；
    若所述候选点在所述自由空间内，则将所述候选点确定为与所述第一规划树对应的第一扩展点。
  8. 一种机械臂运动规划装置，其特征在于，包括：
    随机采样模块，用于在机械臂的自由空间中进行随机采样，得到所述机械臂的位姿采样点；
    第一扩展点确定模块，用于根据所述位姿采样点在所述自由空间中确定与第一规划树对应的第一扩展点；所述第一规划树为从预设的初始位姿向预设的目标位姿扩展的规划树；
    第一规划树优化模块，用于根据预设的路径代价函数将所述第一扩展点添加入所述第 一规划树中，并对所述第一规划树进行优化处理；
    第二扩展点确定模块，用于根据所述第一扩展点在所述自由空间中确定与第二规划树对应的第二扩展点；所述第二规划树为从所述目标位姿向所述初始位姿扩展的规划树；
    第二规划树优化模块，用于根据所述路径代价函数将所述第二扩展点添加入所述第二规划树中，并对所述第二规划树进行优化处理；
    运动规划路径确定模块，用于当所述第一规划树与所述第二规划树满足预设的重合条件时，根据所述第一规划树与所述第二规划树确定所述机械臂的运动规划路径。
  9. 一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1至7中任一项所述的机械臂运动规划方法的步骤。
  10. 一种机械臂，包括存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，其特征在于，所述处理器执行所述计算机程序时实现如权利要求1至7中任一项所述的机械臂运动规划方法的步骤。
PCT/CN2021/124619 2021-03-22 2021-10-19 一种机械臂运动规划方法、装置、可读存储介质及机械臂 Ceased WO2022198993A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110301129.6 2021-03-22
CN202110301129.6A CN113119115A (zh) 2021-03-22 2021-03-22 一种机械臂运动规划方法、装置、可读存储介质及机械臂

Publications (1)

Publication Number Publication Date
WO2022198993A1 true WO2022198993A1 (zh) 2022-09-29

Family

ID=76773670

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/124619 Ceased WO2022198993A1 (zh) 2021-03-22 2021-10-19 一种机械臂运动规划方法、装置、可读存储介质及机械臂

Country Status (2)

Country Link
CN (1) CN113119115A (zh)
WO (1) WO2022198993A1 (zh)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115648220A (zh) * 2022-11-15 2023-01-31 华侨大学 基于最小代价下降的机械臂关节空间避障路径规划方法
CN116652952A (zh) * 2023-06-09 2023-08-29 清华大学 一种使用双机械臂的受限环境柔性线缆操作方法和系统
CN117734676A (zh) * 2024-02-19 2024-03-22 知行汽车科技(苏州)股份有限公司 一种自动泊车方法、装置、设备及存储介质
CN118787444A (zh) * 2024-06-14 2024-10-18 中国科学院深圳先进技术研究院 一种机器人辅助骨折复位路径规划方法及系统
CN119178447A (zh) * 2024-10-25 2024-12-24 中国航空工业集团公司洛阳电光设备研究所 一种基于实时山谷检测的低空突防快速规划方法
CN119304889A (zh) * 2024-12-03 2025-01-14 重庆大学 一种用于将单臂代理的动作扩展到双臂操作的方法及系统
CN119512236A (zh) * 2024-11-20 2025-02-25 济南恒立软件技术有限公司 基于图像识别的机动车尾气检测智能插管控制方法和系统

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113119115A (zh) * 2021-03-22 2021-07-16 深圳市优必选科技股份有限公司 一种机械臂运动规划方法、装置、可读存储介质及机械臂
CN114021518B (zh) * 2021-10-09 2024-04-09 深圳市紫光同创电子有限公司 可编程逻辑器件的路径扩展方法、装置及相关设备

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100174435A1 (en) * 2009-01-07 2010-07-08 Samsung Electronics Co., Ltd. Path planning apparatus of robot and method thereof
US20110106306A1 (en) * 2009-11-02 2011-05-05 Samsung Electronics Co., Ltd. Path planning apparatus of robot and method and computer-readable medium thereof
CN104154917A (zh) * 2013-11-19 2014-11-19 深圳信息职业技术学院 一种机器人避碰路径的规划方法、装置
CN104155974A (zh) * 2013-07-29 2014-11-19 深圳信息职业技术学院 一种用于机器人快速避碰的路径规划方法及设备
CN108444489A (zh) * 2018-03-07 2018-08-24 北京工业大学 一种改进rrt算法的路径规划方法
CN110497403A (zh) * 2019-08-05 2019-11-26 上海大学 一种改进双向rrt算法的机械臂运动规划方法
CN110986953A (zh) * 2019-12-13 2020-04-10 深圳前海达闼云端智能科技有限公司 路径规划方法、机器人及计算机可读存储介质
CN113119115A (zh) * 2021-03-22 2021-07-16 深圳市优必选科技股份有限公司 一种机械臂运动规划方法、装置、可读存储介质及机械臂

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110032187B (zh) * 2019-04-09 2020-08-28 清华大学 无人摩托静态障碍避障路径规划计算方法
CN109990796B (zh) * 2019-04-23 2022-07-19 成都信息工程大学 基于双向扩展随机树的智能车路径规划方法
CN110228069B (zh) * 2019-07-17 2022-04-01 东北大学 一种机械臂在线避障运动规划方法
CN110509279B (zh) * 2019-09-06 2020-12-08 北京工业大学 一种仿人机械臂的运动路径规划方法及系统

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100174435A1 (en) * 2009-01-07 2010-07-08 Samsung Electronics Co., Ltd. Path planning apparatus of robot and method thereof
US20110106306A1 (en) * 2009-11-02 2011-05-05 Samsung Electronics Co., Ltd. Path planning apparatus of robot and method and computer-readable medium thereof
CN104155974A (zh) * 2013-07-29 2014-11-19 深圳信息职业技术学院 一种用于机器人快速避碰的路径规划方法及设备
CN104154917A (zh) * 2013-11-19 2014-11-19 深圳信息职业技术学院 一种机器人避碰路径的规划方法、装置
CN108444489A (zh) * 2018-03-07 2018-08-24 北京工业大学 一种改进rrt算法的路径规划方法
CN110497403A (zh) * 2019-08-05 2019-11-26 上海大学 一种改进双向rrt算法的机械臂运动规划方法
CN110986953A (zh) * 2019-12-13 2020-04-10 深圳前海达闼云端智能科技有限公司 路径规划方法、机器人及计算机可读存储介质
CN113119115A (zh) * 2021-03-22 2021-07-16 深圳市优必选科技股份有限公司 一种机械臂运动规划方法、装置、可读存储介质及机械臂

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115648220A (zh) * 2022-11-15 2023-01-31 华侨大学 基于最小代价下降的机械臂关节空间避障路径规划方法
CN116652952A (zh) * 2023-06-09 2023-08-29 清华大学 一种使用双机械臂的受限环境柔性线缆操作方法和系统
CN117734676A (zh) * 2024-02-19 2024-03-22 知行汽车科技(苏州)股份有限公司 一种自动泊车方法、装置、设备及存储介质
CN117734676B (zh) * 2024-02-19 2024-05-03 知行汽车科技(苏州)股份有限公司 一种自动泊车方法、装置、设备及存储介质
CN118787444A (zh) * 2024-06-14 2024-10-18 中国科学院深圳先进技术研究院 一种机器人辅助骨折复位路径规划方法及系统
CN119178447A (zh) * 2024-10-25 2024-12-24 中国航空工业集团公司洛阳电光设备研究所 一种基于实时山谷检测的低空突防快速规划方法
CN119512236A (zh) * 2024-11-20 2025-02-25 济南恒立软件技术有限公司 基于图像识别的机动车尾气检测智能插管控制方法和系统
CN119512236B (zh) * 2024-11-20 2025-12-02 济南恒立软件技术有限公司 基于图像识别的机动车尾气检测智能插管控制方法和系统
CN119304889A (zh) * 2024-12-03 2025-01-14 重庆大学 一种用于将单臂代理的动作扩展到双臂操作的方法及系统

Also Published As

Publication number Publication date
CN113119115A (zh) 2021-07-16

Similar Documents

Publication Publication Date Title
WO2022198993A1 (zh) 一种机械臂运动规划方法、装置、可读存储介质及机械臂
WO2022198994A1 (zh) 一种机械臂运动规划方法、装置、可读存储介质及机械臂
WO2022160787A1 (zh) 一种机器人手眼标定方法, 装置, 可读存储介质及机器人
CN114003613B (zh) 高精地图车道线更新方法、装置、电子设备以及存储介质
WO2022193639A1 (zh) 机械臂及其轨迹规划方法和装置
TW202119406A (zh) 資料處理方法、處理器、電子設備和電腦可讀媒體
WO2022227429A1 (zh) 一种步态轨迹规划方法、装置、可读存储介质及机器人
CN111815754A (zh) 一种三维信息确定方法、三维信息确定装置及终端设备
WO2020135608A1 (zh) 工业机器人的示教轨迹复现方法、系统及机器人
WO2024041647A1 (zh) 一种笛卡尔空间的轨迹规划方法及装置
WO2021134807A1 (zh) 一种异构网络中相似度的计算方法及其相关组件
WO2019085709A1 (zh) 一种应用于卷积神经网络的池化处理的方法及系统
WO2023005021A1 (zh) 一种地图的更新方法、机器人及可读存储介质
WO2024041646A1 (zh) 一种多轴设备的关节空间轨迹的规划方法及装置
WO2022198995A1 (zh) 步态轨迹规划方法、装置、计算机可读存储介质及机器人
CN114742967B (zh) 一种基于建筑数字孪生语义图的视觉定位方法及装置
CN101894123A (zh) 基于子图的链接相似度的快速近似计算系统和方法
CN111024082B (zh) 一种规划机器人局部路径的方法、装置及机器人
CN112465908A (zh) 一种物体定位方法、装置、终端设备及存储介质
WO2022198992A1 (zh) 一种机械臂运动规划方法、装置、可读存储介质及机械臂
CN117570991A (zh) 一种路径规划方法、装置、电子设备和存储介质
CN104516939A (zh) 一种用以建构人工智能计算机的平行硬件搜索系统
CN114494943A (zh) 新型视频目标检测评价方法、装置、产品及存储介质
CN118565503A (zh) 路径优化方法、装置及可读存储介质
CN110609854A (zh) 字段名查询方法、系统、电子设备及计算机存储介质

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: 21932602

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: 21932602

Country of ref document: EP

Kind code of ref document: A1