CN110815211B - Method for quadruped robot to dynamically cross convex obstacle - Google Patents

Method for quadruped robot to dynamically cross convex obstacle Download PDF

Info

Publication number
CN110815211B
CN110815211B CN201910933745.6A CN201910933745A CN110815211B CN 110815211 B CN110815211 B CN 110815211B CN 201910933745 A CN201910933745 A CN 201910933745A CN 110815211 B CN110815211 B CN 110815211B
Authority
CN
China
Prior art keywords
obstacle
robot
convex
virtual
information
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.)
Active
Application number
CN201910933745.6A
Other languages
Chinese (zh)
Other versions
CN110815211A (en
Inventor
许�鹏
苏波
江磊
姚其昌
党睿娜
许威
蒋云峰
慕林栋
降晨星
邓秦丹
梁振杰
汪建兵
郭亮
卢玉传
赵建新
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.)
China North Vehicle Research Institute
Original Assignee
China North Vehicle Research Institute
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 China North Vehicle Research Institute filed Critical China North Vehicle Research Institute
Priority to CN201910933745.6A priority Critical patent/CN110815211B/en
Publication of CN110815211A publication Critical patent/CN110815211A/en
Application granted granted Critical
Publication of CN110815211B publication Critical patent/CN110815211B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
    • B25J9/1664—Programme 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
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1674—Programme controls characterised by safety, monitoring, diagnostic
    • B25J9/1676—Avoiding collision or forbidden zones
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D57/00—Vehicles characterised by having other propulsion or other ground- engaging means than wheels or endless track, alone or in addition to wheels or endless track
    • B62D57/02—Vehicles characterised by having other propulsion or other ground- engaging means than wheels or endless track, alone or in addition to wheels or endless track with ground-engaging propulsion means, e.g. walking members
    • B62D57/032—Vehicles characterised by having other propulsion or other ground- engaging means than wheels or endless track, alone or in addition to wheels or endless track with ground-engaging propulsion means, e.g. walking members with alternately or sequentially lifted supporting base and legs; with alternately or sequentially lifted feet or skid

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Robotics (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Transportation (AREA)
  • Manipulator (AREA)

Abstract

The invention discloses a method for a quadruped robot to dynamically cross convex obstacles, and belongs to the technical field of robot motion control. The method comprises the following implementation steps: firstly, detecting a foot falling safety region of the robot facing a convex obstacle, secondly, re-planning a swing leg track, and finally establishing a robot dynamic convex obstacle crossing stable controller. The invention realizes the crossing of the convex obstacle by establishing the robot dynamic obstacle crossing stable controller, and improves the walking efficiency of the quadruped robot to the complex obstacle pavement.

Description

Method for quadruped robot to dynamically cross convex obstacle
Technical Field
The invention relates to the technical field of robot motion control, in particular to a method for a quadruped robot to dynamically cross a convex obstacle.
Background
Compared with the traditional wheeled and tracked vehicles, the quadruped robot can adopt discrete swing legs to fall to foot points, so that the quadruped robot can adapt to complicated terrains such as plateaus and mountains, at present, the quadruped robot mainly focuses on body control based on self sensing, and the stepping height of the swing legs is fixed or artificially set, so that the quadruped robot is used for large-size convex obstacles: 1) the operator is used for remotely controlling the robot to select the detour; 2) the swing track of the swing leg is manually adjusted by judging the size of the convex obstacle by an operator, so that the convex obstacle is crossed. The former can successfully avoid convex obstacles, but the efficiency is low; the latter may lead to a risk of the robot being blocked in its movement and even toppling due to inaccuracies in the judgement of the convex obstacle by the operator. Although research on autonomous walking of the robot is carried out by part of colleges and universities and research institutions, the research is mainly focused on autonomous behaviors such as personnel following and the like, and the research on local foot-falling autonomous planning is relatively deficient.
Disclosure of Invention
In view of the above, the invention provides a method for a quadruped robot to dynamically cross a convex obstacle, which realizes the crossing of the convex obstacle by establishing a robot dynamic obstacle crossing stable controller, and improves the walking efficiency of the quadruped robot on a complex obstacle road surface.
A method for a quadruped robot to dynamically cross convex obstacles comprises the following implementation steps:
the method comprises the steps of firstly, detecting a foot falling safety region of a robot facing a convex obstacle;
step two, re-planning the swing leg track;
and step three, establishing a robot dynamic convex obstacle crossing stable controller by using the terrain attitude angle as input.
Further, the specific implementation process of the step one is as follows:
step 1: establishing a height chart of the foot falling area by adopting multi-sensor fusion, and acquiring height information of the foot falling area;
step 2: calculating topographic relief information of the foot falling area;
and step 3: and setting a topographic relief information threshold value, comparing the topographic relief information with the current relief information, and regarding the region as a foot-falling safe region when the current relief information is smaller than the set threshold value, otherwise, regarding the region as a dangerous region.
Further, the specific implementation process of the second step is as follows:
step 1: establishing a virtual obstacle according to the size information of the actual obstacle;
step 2: deciding to cross the obstacle and fall on the surface of the obstacle according to the size information of the virtual obstacle;
and step 3: and replanning the cubic spline track curve according to the virtual obstacle corner points and the foot falling points.
Further, the third step is realized as follows:
step 1: estimating terrain attitude angles in real time;
step 2: taking the terrain attitude angle as input, establishing a virtual servo controller of the machine body, and distributing the virtual servo controller to the feet through a target optimization method;
and step 3: a bottom compliant servo controller is established.
Has the advantages that:
1. the invention can effectively judge the safety of crossing convex obstacles and falling feet by adopting the topographic relief information and the elevation information.
2. The invention replans the foot drop point according to the position of the virtual convex barrier, can effectively span the convex barrier, and adopts the spline algorithm to plan the swing track to be stable and reliable.
3. The invention adopts a virtual servo and bottom layer compliance control method, and can improve the stability and compliance of the robot in dynamic crossing of convex obstacles.
4. The invention can realize the dynamic self-adaptation of the quadruped robot in a complex environment and improve the walking efficiency of the quadruped robot under a convex obstacle road surface.
Drawings
FIG. 1 is a flow chart of the steps of a quadruped robot dynamic convex obstacle crossing method;
FIGS. 2 and 3 are schematic diagrams of detection of a four-legged robot foot-placement safety zone;
FIG. 4 is a quadruped robot swing leg trajectory re-planning;
fig. 5 is a dynamic obstacle crossing stabilization control of the quadruped robot.
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The invention provides a method for a quadruped robot to dynamically cross convex obstacles, which comprises the following implementation steps as shown in the attached figure 1:
step 1, detecting a foot falling safety region of the robot. The method specifically comprises the following steps:
step 11: the height map of the foot falling area is built through fusion of multiple sensors such as an airborne camera and a laser radar, and height information x of the foot falling area is obtained, as shown in the attached figure 2.
Step 12: and calculating topographic relief information sigma of the foot falling area according to the height information of the foot falling area, wherein the sigma is calculated as follows:
Figure BDA0002220962700000031
where N is the number of elevation grids, xiThe elevation value of the ith grid is taken as the elevation value, and x is the elevation average value of the N grids.
Step 13: setting a topographic relief information threshold σ*And comparing the current fluctuation information sigma with the current fluctuation information sigma, wherein the current fluctuation information sigma is smaller than a set threshold value sigma*When the foot is in the safe area, the foot is in the dangerous area.
Step 2, re-planning the swing leg trajectory, as shown in fig. 3 and 4. The method specifically comprises the following steps:
step 21: establishing a virtual obstacle according to the information such as the size of the actual obstacle, including the height H of the convex obstacle and the length L of the convex obstacle, setting a distance threshold delta H between the virtual convex obstacle and the actual convex obstacle, and acquiring the size of the virtual convex obstacle, including the height H of the convex obstaclevConvex obstacleLength LvAnd key corner information P2v、P4v。
Step 22: according to the information such as the size of the virtual obstacle, the obstacle crossing and the obstacle falling surface are decided, and the length threshold value of the robot allowed to cross the obstacle is set to be L*When the length L of the virtual convex obstaclevLess than L*When the obstacle is crossed, the obstacle can be crossed, and when the obstacle is not crossed, the obstacle falls on the surface of the convex obstacle.
Step 23: replanning cubic spline track curve according to the corner points and the foot falling points of the virtual obstacles, and setting a safety threshold L of the starting pointfPlanning the starting point P of the swing leg1vPlanning the highest point P of the swing track3vPlanning the foot-falling position P according to the obstacle crossing decision and the foot-falling point safety zone5vAnd planning the swing track of the swing leg by adopting a spline interpolation algorithm.
As shown in fig. 5, step 3, the robot dynamic obstacle crossing stability controller is established. The method specifically comprises the following steps:
step 31: estimating terrain attitude angles including a rolling attitude angle alpha and a pitching attitude angle beta in real time, and solving by calculating a plane normal vector n formed by foot supporting points p, wherein the normal vector is solved as follows:
(p-p0)·n＝0
in the formula, p0Is the initial position.
Step 32: the method comprises the following steps of establishing a virtual servo controller of the fuselage by taking a terrain attitude angle as an expected pose input element of the fuselage, distributing the virtual servo controller to feet through a target optimization method, and establishing the virtual servo controller as follows:
Figure BDA0002220962700000041
in the formula, Mm、FmThe moment vector and the virtual resultant force vector of the engine body are taken as the virtual resultant force moment vector and the virtual resultant force vector of the engine body; k is a radical ofpIs a positive definite gain matrix; q is the actual pose vector of the fuselage; q. q.sdAn expected pose vector for the fuselage; k is a radical ofvIs a positive definite differential coefficient matrix;
Figure BDA0002220962700000051
is the actual pose velocity vector of the fuselage,
Figure BDA0002220962700000052
for the desired pose velocity vector, k, of the fuselageffIs a positive constant velocity feedforward coefficient matrix, m is the robot mass, and g is the gravitational acceleration.
By setting constraint conditions, an objective optimization function F ═ F (M) is establishedm,Fm) Solving a set of solutions which satisfy constraint conditions and enable the objective function to obtain a minimum value, wherein the constraint conditions are as follows:
Figure BDA0002220962700000053
Figure BDA0002220962700000054
in the above formula, the first and second carbon atoms are,
Figure BDA0002220962700000055
normal force at the ith contact point;
Figure BDA0002220962700000056
the tangential force at the ith contact point; μ is a sliding friction coefficient.
Step 33: the bottom layer compliant servo controller was built with the force of the distributed foot as input, and the joint controller was designed as follows.
Figure BDA0002220962700000057
In the formula: k is a radical ofpθ、kvθThe rigidity and damping coefficient of the controller are set; thetadA desired joint angle; θ is the actual joint angle;
Figure BDA0002220962700000058
is as desiredJoint angular velocity;
Figure BDA0002220962700000059
actual joint angular velocity; tau isff＝-JTFdIs a joint force feedforward term; fdA desired force for the foot end; j is a joint force Jacobian matrix; u is the joint controller input.
In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (2)

1. A method for a quadruped robot to dynamically cross convex obstacles is characterized by comprising the following implementation steps:
the method comprises the steps of firstly, detecting a foot falling safety region of a robot facing a convex obstacle;
step two, re-planning the swing leg track;
step three, establishing a robot dynamic crossing convex obstacle stabilizing controller by using the terrain attitude angle as input;
the specific implementation process of the first step is as follows:
step 1: establishing a height chart of the foot falling area by adopting multi-sensor fusion, and acquiring height information of the foot falling area;
step 2: calculating topographic relief information of the foot falling area;
and step 3: setting a topographic relief information threshold value, comparing the topographic relief information with the current topographic relief information, and regarding the current topographic relief information as a foot falling safety area when the current topographic relief information is smaller than the set threshold value, otherwise, regarding the current topographic relief information as a dangerous area;
the concrete implementation process of the second step is as follows:
step 1: establishing a virtual obstacle according to the size information of the actual obstacle;
step 2: deciding to cross the obstacle and fall on the surface of the obstacle according to the size information of the virtual obstacle;
and step 3: and replanning the cubic spline track curve according to the virtual obstacle corner points and the foot falling points.
2. The dynamic convex obstacle crossing method of the quadruped robot as claimed in claim 1, wherein the third step is realized by the following steps:
step 1: estimating terrain attitude angles in real time;
step 2: taking the terrain attitude angle as input, establishing a virtual servo controller of the machine body, and distributing the virtual servo controller to the feet through a target optimization method;
and step 3: a bottom compliant servo controller is established.
CN201910933745.6A 2019-09-29 2019-09-29 Method for quadruped robot to dynamically cross convex obstacle Active CN110815211B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910933745.6A CN110815211B (en) 2019-09-29 2019-09-29 Method for quadruped robot to dynamically cross convex obstacle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910933745.6A CN110815211B (en) 2019-09-29 2019-09-29 Method for quadruped robot to dynamically cross convex obstacle

Publications (2)

Publication Number Publication Date
CN110815211A CN110815211A (en) 2020-02-21
CN110815211B true CN110815211B (en) 2021-04-23

Family

ID=69548529

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910933745.6A Active CN110815211B (en) 2019-09-29 2019-09-29 Method for quadruped robot to dynamically cross convex obstacle

Country Status (1)

Country Link
CN (1) CN110815211B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111367327A (en) * 2020-04-02 2020-07-03 杭州云深处科技有限公司 Control method and device for bionic jumping action of quadruped robot
CN113547517B (en) * 2020-04-24 2022-08-02 深圳市优必选科技股份有限公司 Gait planning method and device for dynamic obstacle avoidance, readable storage medium and robot
CN113485405A (en) * 2021-08-05 2021-10-08 Oppo广东移动通信有限公司 Attitude acquisition method, robot and readable storage medium
CN114063490B (en) * 2021-09-28 2024-08-13 北京炎凌嘉业智能科技股份有限公司 Intelligent bionic foot type robot control system and method
CN113867350A (en) * 2021-09-28 2021-12-31 北京炎凌嘉业机电设备有限公司 Intelligent bionic foot type robot stepping behavior autonomous adaptation method
CN114035593B (en) * 2021-09-28 2024-04-12 北京炎凌嘉业机电设备有限公司 Force servo foot type motion control method for multi-mode motion
CN114683289B (en) * 2022-05-09 2024-06-18 中兵智能创新研究院有限公司 Gesture pre-adaptive control method for four-foot robot

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105806312A (en) * 2016-03-11 2016-07-27 中国北方车辆研究所 Terrain parameter measurement method based on foot type robot with three or more supporting legs
CN106695791A (en) * 2017-01-11 2017-05-24 山东科技大学 Generation method of continuous free tread of four-foot biomimetic robot
CN107121128A (en) * 2017-05-27 2017-09-01 中国北方车辆研究所 A kind of measuring method and system of legged type robot terrain parameter
CN108860360A (en) * 2018-08-01 2018-11-23 清华大学深圳研究生院 A kind of system of quadruped robot obstacle jump
US10226870B1 (en) * 2014-11-11 2019-03-12 Boston Dynamics, Inc. Yaw slip handling in a robotic device
CN109807884A (en) * 2018-12-29 2019-05-28 江苏集萃智能制造技术研究所有限公司 A kind of biped robot's gait stability control method based on inertial navigation system

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10226870B1 (en) * 2014-11-11 2019-03-12 Boston Dynamics, Inc. Yaw slip handling in a robotic device
CN105806312A (en) * 2016-03-11 2016-07-27 中国北方车辆研究所 Terrain parameter measurement method based on foot type robot with three or more supporting legs
CN106695791A (en) * 2017-01-11 2017-05-24 山东科技大学 Generation method of continuous free tread of four-foot biomimetic robot
CN107121128A (en) * 2017-05-27 2017-09-01 中国北方车辆研究所 A kind of measuring method and system of legged type robot terrain parameter
CN108860360A (en) * 2018-08-01 2018-11-23 清华大学深圳研究生院 A kind of system of quadruped robot obstacle jump
CN109807884A (en) * 2018-12-29 2019-05-28 江苏集萃智能制造技术研究所有限公司 A kind of biped robot's gait stability control method based on inertial navigation system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
四足机器人地形识别与路径规划算法;张慧等;《机器人》;20150930;第37卷(第5期);第0546-0556页 *

Also Published As

Publication number Publication date
CN110815211A (en) 2020-02-21

Similar Documents

Publication Publication Date Title
CN110815211B (en) Method for quadruped robot to dynamically cross convex obstacle
CN110834685B (en) Method for quadruped robot to dynamically cross concave obstacle
US11548151B2 (en) Robotically negotiating stairs
US11654985B2 (en) Mechanically-timed footsteps for a robotic device
US10226870B1 (en) Yaw slip handling in a robotic device
US9975245B1 (en) Methods and systems for planning a body position of a robotic device
Gehring et al. Dynamic trotting on slopes for quadrupedal robots
US10059392B1 (en) Control of robotic devices with non-constant body pitch
CN114022824A (en) Narrow environment-oriented quadruped robot motion planning method
US10196104B1 (en) Terrain Evaluation for robot locomotion
Zhang et al. Efficient motion planning based on kinodynamic model for quadruped robots following persons in confined spaces
Hopkins et al. Design of a compliant bipedal walking controller for the DARPA Robotics Challenge
Tennakoon et al. Probe-before-step walking strategy for multi-legged robots on terrain with risk of collapse
CN117270551A (en) Method and system for autonomous climbing stairs by quadruped robot based on local environment perception
Lee et al. Whole-body motion and landing force control for quadrupedal stair climbing
CN116619382A (en) Robot motion control method and system and electronic equipment
Lu et al. Whole‐body motion planning and control of a quadruped robot for challenging terrain
Zhang et al. A composite cog trajectory planning method for the quadruped robot walking on rough terrain
Liu et al. Research on path planning of quadruped robot based on globally mapping localization
Freitas et al. Terrain model-based anticipative control for articulated vehicles with low bandwidth actuators
Bałchanowski Modelling and simulation studies on the mobile robot with self-leveling chassis
Lai et al. Hierarchical incremental path planning and situation-dependent optimized dynamic motion planning considering accelerations
Roditis et al. Maintaining static stability and continuous motion in rough terrain hexapod locomotion without terrain mapping
Yin et al. A novel control strategy for quadruped robot walking over irregular terrain
Jordan et al. Real-time model based path planning for wheeled vehicles

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant