CN114742310A - Terrain trafficability map construction method based on wheel-ground interaction - Google Patents

Terrain trafficability map construction method based on wheel-ground interaction Download PDF

Info

Publication number
CN114742310A
CN114742310A CN202210426426.8A CN202210426426A CN114742310A CN 114742310 A CN114742310 A CN 114742310A CN 202210426426 A CN202210426426 A CN 202210426426A CN 114742310 A CN114742310 A CN 114742310A
Authority
CN
China
Prior art keywords
wheel
terrain
wheeled robot
grid
ground
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.)
Granted
Application number
CN202210426426.8A
Other languages
Chinese (zh)
Other versions
CN114742310B (en
Inventor
陈超
舒明雷
王英龙
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.)
Shandong Institute of Artificial Intelligence
Original Assignee
Shandong Institute of Artificial Intelligence
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 Shandong Institute of Artificial Intelligence filed Critical Shandong Institute of Artificial Intelligence
Priority to CN202210426426.8A priority Critical patent/CN114742310B/en
Publication of CN114742310A publication Critical patent/CN114742310A/en
Application granted granted Critical
Publication of CN114742310B publication Critical patent/CN114742310B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38—Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3804—Creation or updating of map data
    • G01C21/3833—Creation or updating of map data characterised by the source of data
    • G01C21/3841—Data obtained from two or more sources, e.g. probe vehicles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—2D [Two Dimensional] image generation
    • G06T11/20—Drawing from basic elements, e.g. lines or circles
    • G06T11/206—Drawing of charts or graphs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • Strategic Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Artificial Intelligence (AREA)
  • Development Economics (AREA)
  • Data Mining & Analysis (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Educational Administration (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computer Hardware Design (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Automation & Control Theory (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Mathematical Physics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Biophysics (AREA)
  • Computational Mathematics (AREA)

Abstract

A terrain trafficability map construction method based on wheel-ground interaction is characterized in that a mapping relation between visually acquired terrain classification information and tactile characteristics is constructed based on wheel-ground interaction, and a terrain trafficability comprehensive evaluation rule based on visual-tactile characteristics is designed. By introducing the wheel-ground interaction indexes, the influence of different motion states of the wheels on the terrain trafficability is introduced into an evaluation process, the conservatism of the traditional trafficability evaluation result is reduced, and the calculation amount of the complex map representation in the planning process is reduced.

Description

Terrain trafficability map construction method based on wheel-ground interaction
Technical Field
The invention relates to the technical field of outdoor environment perception and map construction of wheeled robots, in particular to a terrain passability map construction method based on wheel-ground interaction.
Background
The trafficability of the wheeled mobile robot in outdoor complex terrains with different physical characteristics is closely related to the geometric characteristics, the terrain characteristics, the wheel-ground contact characteristics, the movement characteristics and the like of the front obstacle. The terrain passability evaluation result is conservative only by means of visual information to obtain the characteristics, and evaluation errors are easy to occur in certain deceptive terrains. The contact characteristics of the terrain are related to terrain types and wheel motion states, the contact characteristics are represented by simply depending on the terrain types, the important function of the motion states cannot be reflected, and meanwhile, the passability is difficult to quantitatively evaluate. Therefore, it is necessary to integrate visually obtained terrain information and tactile characteristics based on ground-wheel interaction analysis, evaluate the passability of the terrain based on the ground-wheel interaction, and create a vectorized passability map.
At present, a single visual sensor is used for terrain evaluation at home and abroad, the research on the contact characteristic of the terrain is introduced, most evaluation rules are relatively rough in design and cannot be quantitatively analyzed, and the evaluation result is still very conservative. Therefore, a new characterization form of the terrain contact characteristic needs to be researched, and the terrain trafficability characteristic is combined with the characteristics of visual information acquisition such as geometry and texture to realize comprehensive evaluation of terrain trafficability and construction of trafficability maps.
Disclosure of Invention
In order to overcome the defects of the technologies, the invention provides a terrain passability comprehensive evaluation and map construction method integrating visual characteristics and contact characteristics.
The technical scheme adopted by the invention for overcoming the technical problems is as follows:
a terrain passability map construction method based on ground-wheel interaction comprises the following steps:
a) hard soil is treated by using cohesion parameter and shearing deformation modulus parameterEquivalent to soft soil, and constructing a wheel-ground interaction prediction model of the wheeled robot [ FN,FDP,FS,MO,MR,MS]T＝Fterrain(s,ω,β|PS,PR,PT,PW) In which F isNIs a normal force of a wheel of the wheeled robot, FDPFor the hitch traction of the wheels of a wheeled robot, FSIs a side force of a wheel of a wheeled robot, MOTurning moment of wheel of wheeled robot, MRForward moment of resistance, M, for wheels of wheeled robotsSIs the steering resistance moment of the wheel of the wheeled robot, s is the slip ratio of the wheel of the wheeled robot, omega is the wheel rotating speed, beta is the wheel slip angle, P is the wheel slip angleSIs a soil physical mechanical property parameter, PRAs an operating state parameter, P, of the wheels of the wheeled robotTFor a topographic angle parameter, PWIs a wheel parameter;
b) inputting the actual slip ratio s ', the wheel rotation speed omega ' and the wheel slip angle beta ' of the wheel of the wheeled robot into a wheel-ground interaction prediction model [ F ] of the wheeled robotN,FDP,FS,MO,MR,MS]T＝Fterrain(s,ω,β|PS,PR,PT,PW) In the method, the actual normal force F of the wheel type robot is obtained through outputN' wheel hook traction force FD′PLateral force F of wheelS' turning moment M ' of wheel 'OForward moment M 'of wheel'RAnd the steering resistance torque M of the wheelS', calculating F separatelyN' and FNError of (2), FD′PAnd FDPError of (2), FS' and FSError of (2), M'OAnd MOError of (2), M'RAnd MRError of (2), MS' and MSConstructing a probability prediction model of uncertainty;
c) combining a wheel-ground interaction prediction model of the wheeled robot with a probability prediction model of uncertainty to obtain a wheel containing model uncertaintyGround effect probability prediction model
Figure BDA0003608994360000021
Xi is estimation prediction of uncertainty, xi obeys CNP, CNP is modeling of deep Gaussian process learning system, mteIs the mean value of the Gaussian function, kteA covariance function which is a gaussian function;
d) acquiring a soil image of the ground in front of the robot by using a visual sensor, carrying out terrain recognition and classification on the soil image based on a graph neural network and a convolution network, and carrying out classification result and a soil physical and mechanical characteristic parameter PSEstablishing an association array;
e) obtaining the terrain gradient S, the bumpiness U and the obstacle height H information of the soil image by calling a visual image processing function in an OpenCV (open vehicle vision library), establishing a terrain trafficability comprehensive evaluation criterion based on the visual-tactile characteristics by combining the wheel action probability prediction model in the step c), and obtaining an evaluation index T of the terrain trafficability in a single griddyn；
f) Evaluation index T for terrain passability in single griddynAnd assigning values to each grid of the grid map, and constructing to obtain the dynamic vectorization map.
Further, a prediction model of the round-to-round interaction is constructed through a closed analytical decoupling model in the step a).
Further, in the step d), the soil image is input into the convolutional neural network CNN to extract image features, the extracted image features are input into the convolutional neural network GNN to characterize the distribution relationship among different terrains, and the image features and the distribution relationship are input into the softmax classifier together to obtain the classification of the terrains of the soil image.
Further, step e) is performed by the formula Tgrid＝α1Sgrid+α2Hgrid+α3UgridCalculating to obtain an evaluation value T of terrain trafficabilitygridBy the formula Tupdate(a)＝α4FNgrid+α5FDPgrid+α6FSgrid+α7MOgrid+α8MRgrid+α9MSgridCalculating to obtain a passability dynamic update value T with vectorityupdate(a) In the formula, α1、α2、α3、α4、α5、α6、α7、α8、α9Are all weights, α1+α2+α3+α4+α5+α6+α7+α8+α9＝1，SgridIs the terrain slope of the normalized soil image, HgridFor normalized obstacle height, UgridIs normalized roughness, FNgridIs normalized normal force of wheel of wheeled robot, FDPgridFor normalized towing force of wheel of wheeled robot, FSgridFor normalized lateral force, M, of wheels of wheeled robotsOgridIs the turning moment of the wheel of the normalized wheeled robot, MRgridIs a normalized advancing resistance moment, M, of the wheels of the wheeled robotSgridThe normalized steering resistance moment of the wheels of the wheeled robot is obtained by the formula Tdyn＝Tgrid+Tupdate(a) Calculating to obtain an evaluation index T of the terrain trafficability in a single griddyn。
The invention has the beneficial effects that: and constructing a mapping relation between visually acquired terrain classification information and tactile characteristics based on wheel-to-ground interaction, and designing a terrain trafficability characteristic comprehensive evaluation rule based on the visual-tactile characteristics. By introducing the wheel-ground interaction indexes, the influence of different motion states of the wheels on the terrain trafficability is introduced into an evaluation process, the conservatism of the traditional trafficability evaluation result is reduced, and the calculation amount of the complex map representation in the planning process is reduced.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
The invention is further described below with reference to fig. 1.
A terrain passability mapping method based on wheel-ground interaction comprises the following steps:
a) the hard soil is equivalent to the soft soil by the cohesion parameter and the shearing deformation modulus parameter, and a wheel-ground interaction prediction model of the wheeled robot is constructed [ FN,FDP,FS,MO,MR,MS]T＝Fterrain(s,ω,β|PS,PR,PT,PW) In which FNIs a normal force of a wheel of the wheeled robot, FDPFor the hitch traction of the wheels of a wheeled robot, FSLateral force of wheels of wheeled robot, MOTurning moment of wheel of wheeled robot, MRForward moment of resistance, M, for a wheel of a wheeled robotSIs the steering resistance moment of the wheel of the wheeled robot, s is the slip ratio of the wheel of the wheeled robot, omega is the wheel rotating speed, beta is the wheel slip angle, P is the wheel slip angleSIs a soil physical mechanical property parameter, PRAs an operating state parameter, P, of the wheels of the wheeled robotTFor a topographic angle parameter, PWIs a wheel parameter. These four parameters, PSThe method can be obtained by referring to research on active suspension Mars vehicle modeling and simulation technology facing to multi-motion working conditions in 2020, P of the thesis of Khatag, Yangyuan, PhdR、PTAnd PWThe method can be obtained in 2010 by referring to a lunar/celestial wheel ground action ground mechanics model and application research thereof in the Kangda Dingbo paper. And will not be described in detail herein.
b) Inputting the actual slip ratio s ', the wheel rotation speed omega ' and the wheel slip angle beta ' of the wheel of the wheeled robot into a wheel-ground interaction prediction model [ F ] of the wheeled robotN,FDP,FS,MO,MR,MS]T＝Fterrain(s,ω,β|PS,PR,PT,PW) Wherein the output is obtained an actual normal force F 'of the wheel of the wheeled robot'NHook traction force F 'of wheel'DPSide force F 'of wheel'SOverturning moment M 'of wheel'OForward moment M 'of wheel'RAnd steering resistance torque M 'of wheel'SSeparately calculate F'NAnd FNError of (2), F'DPAnd FDPError of (2), F'SAnd FSError of (2), M'OAnd MOError of (2), M'RAnd MRError of (2), M'SAnd MSAnd (4) constructing a probabilistic prediction model of uncertainty.
c) Combining a wheel-ground interaction prediction model of the wheeled robot with an uncertainty probability prediction model to obtain a wheel-ground interaction probability prediction model containing model uncertainty
Figure BDA0003608994360000051
Xi is estimation prediction of uncertainty, xi obeys CNP, CNP is modeling of deep Gaussian process learning system, mteIs the mean value of the Gaussian function, kteIs a covariance function of a gaussian function.
d) Acquiring a soil image of the ground in front of the robot by using a visual sensor, carrying out terrain recognition and classification on the soil image based on a graph neural network and a convolution network, and carrying out classification result and a soil physical and mechanical characteristic parameter PSAnd establishing an association array.
e) Obtaining the terrain gradient S, the bumpiness U and the obstacle height H information of the soil image by calling a visual image processing function in an OpenCV (open vehicle vision library), establishing a terrain trafficability comprehensive evaluation criterion based on the visual-tactile characteristics by combining the wheel action probability prediction model in the step c), and obtaining an evaluation index T of the terrain trafficability in a single griddyn。
f) Evaluation index T for terrain passability in single griddynAnd assigning values to each grid of the grid map, and constructing to obtain the dynamic vectorization map. By utilizing the visual characteristics and the contact characteristics, the comprehensive evaluation of the terrain trafficability is developed, and a map is designed to represent trafficability. Traditional grid map forms are simple and computationally convenient, and therefore still employ such maps as passability maps. Considering that the wheels of the mobile robot are units for directly sensing the terrain change during the movement process, the grid size is set according to the wheel diameter of each wheel.
And constructing a mapping relation between visually acquired terrain classification information and tactile characteristics based on wheel-to-ground interaction, and designing a terrain trafficability characteristic comprehensive evaluation rule based on the visual-tactile characteristics. By introducing the wheel-ground interaction indexes, the influence of different motion states of the wheels on the terrain trafficability is introduced into an evaluation process, the conservatism of the traditional trafficability evaluation result is reduced, and the calculation amount of the complex map representation in the planning process is reduced.
Example 1:
constructing a wheel-ground interaction prediction model through a closed analysis decoupling model in the step a).
Example 2:
in the step d), the soil image is input into the convolutional neural network CNN to extract image features, the extracted image features are input into the graph neural network GNN to characterize the distribution relation among different terrains, and the image features and the distribution relation are input into a softmax classifier together to obtain the classification of the terrains of the soil image.
Example 3:
in step e) by the formula Tgrid＝α1Sgrid+α2Hgrid+α3UgridCalculating to obtain an evaluation value T of terrain trafficabilitygridBy the formula Tupdate(a)＝α4FNgrid+α5FDPgrid+α6FSgrid+α7MOgrid+α8MRgrid+α9MSgridCalculating to obtain a passable dynamic update value T with vectorityupdate(a) In the formula, wherein alpha1、α2、α3、α4、α5、α6、α7、α8、α9Are all weights, α1+α2+α3+α4+α5+α6+α7+α8+α9＝1，SgridIs the terrain slope of the normalized soil image, HgridFor normalized obstacle height, UgridIs the normalized roughness, FNgridIs normalized normal force of wheel of wheeled robot, FDPgridFor normalized wheeled machinesHitch tractive force of robot wheel, FSgridFor normalised lateral forces, M, of wheels of wheeled robotsOgridIs the turning moment of the wheel of the normalized wheeled robot, MRgridIs a normalized advancing resistance moment, M, of the wheels of the wheeled robotSgridAnd (4) calculating terrain trafficability based on indexes such as gradient difference and height difference between adjacent grids after trafficability evaluation in a single grid is finished for the normalized steering resistance torque of the wheels of the wheeled robot. Based on the wheel-ground interaction model, terrain with the same gradient has different trafficability due to different motion strategies of climbing, crossing and descending. Therefore, vectorial performance exists between adjacent grids of the map on the aspect of planning strategy, and the influence of the motion strategy on the terrain passability is shown by adding a vector direction on the basis of the grid map. Specifically, by the formula Tdyn＝Tgrid+Tupdate(a) Calculating to obtain an evaluation index T of the terrain trafficability in a single griddyn。
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described above, or equivalents may be substituted for elements thereof. 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 (4)

1. A terrain passability map construction method based on wheel-ground interaction is characterized by comprising the following steps:
a) the hard soil is equivalent to the soft soil by the cohesion parameter and the shearing deformation modulus parameter, and a wheel-ground interaction prediction model of the wheeled robot is constructed [ FN,FDP,FS,MO,MR,MS]T＝Fterrain(s,ω,β|PS,PR,PT,PW) In which FNNormal force of wheel of wheeled robot, FDPTraction of the hooks for the wheels of a wheeled robot, FSIs a side force of a wheel of a wheeled robot, MOTurning moment of wheel of wheeled robot, MRForward moment of resistance, M, for a wheel of a wheeled robotSIs the steering resistance moment of the wheel of the wheeled robot, s is the slip ratio of the wheel of the wheeled robot, omega is the wheel rotating speed, beta is the wheel slip angle, P is the wheel slip angleSIs a soil physical mechanical property parameter, PRAs an operating state parameter, P, of the wheels of the wheeled robotTFor a topographic angle parameter, PWIs a wheel parameter;
b) inputting the actual slip ratio s ', the wheel rotation speed omega ' and the wheel slip angle beta ' of the wheel of the wheeled robot into a wheel-ground interaction prediction model [ F ] of the wheeled robotN,FDP,FS,MO,MR,MS]T＝Fterrain(s,ω,β|PS,PR,PT,PW) Wherein the output is obtained an actual normal force F 'of the wheel of the wheeled robot'NHook traction force F 'of wheel'DPSide force F 'of wheel'SOverturning moment M 'of wheel'OForward resistance moment M 'of wheel'RAnd steering resistance torque M 'of wheel'SSeparately, calculating F'NAnd FNError of (2), F'DPAnd FDPError of (2), F'SAnd FSError of (2), M'OAnd MOError of (2), M'RAnd MRError of (2), M'SAnd MSConstructing a probability prediction model of uncertainty;
c) combining a wheel-ground interaction prediction model of the wheeled robot with an uncertainty probability prediction model to obtain a wheel-ground interaction probability prediction model containing model uncertainty
Figure FDA0003608994350000011
Xi is estimation prediction of uncertainty, xi obeys CNP, CNP is modeling of deep Gaussian process learning system, mteIs the mean of a Gaussian function，kteA covariance function which is a gaussian function;
d) acquiring a soil image of the ground in front of the robot by using a visual sensor, carrying out terrain recognition and classification on the soil image based on a graph neural network and a convolution network, and carrying out classification result and a soil physical and mechanical characteristic parameter PSEstablishing an association array;
e) obtaining the terrain gradient S, the bumpiness U and the obstacle height H information of the soil image by calling a visual image processing function in an OpenCV library, establishing a terrain trafficability comprehensive evaluation criterion based on the visual-tactile characteristics by combining the wheel action probability prediction model in the step c), and obtaining an evaluation index T of terrain trafficability in a single griddyn；
f) Evaluation index T for terrain passability in single griddynAnd assigning values to each grid of the grid map, and constructing to obtain the dynamic vectorization map.
2. The terrain passability mapping method based on ground-wheel interaction according to claim 1, characterized in that: constructing a wheel-ground interaction prediction model through a closed analysis decoupling model in the step a).
3. The terrain passability mapping method based on ground-wheel interaction according to claim 1, characterized in that: in the step d), the soil image is input into the convolutional neural network CNN to extract image features, the extracted image features are input into the graph neural network GNN to characterize the distribution relation among different terrains, and the image features and the distribution relation are input into a softmax classifier together to obtain the classification of the terrains of the soil image.
4. The terrain passability mapping method based on ground-wheel interaction according to claim 1, characterized in that: in step e) by the formula Tgrid＝α1Sgrid+α2Hgrid+α3UgridCalculating to obtain an evaluation value T of terrain trafficabilitygridBy the formula Tupdate(a)＝α4FNgrid+α5FDPgrid+α6FSgrid+α7MOgrid+α8MRgrid+α9MSgridCalculating to obtain a passable dynamic update value T with vectorityupdate(a) In the formula, wherein alpha1、α2、α3、α4、α5、α6、α7、α8、α9Are all weights, α1+α2+α3+α4+α5+α6+α7+α8+α9＝1，SgridIs the terrain slope of the normalized soil image, HgridFor normalized obstacle height, UgridIs normalized roughness, FNgridIs normalized normal force of wheel of wheeled robot, FDPgridFor normalized couple tractive force of wheeled robot wheels, FSgridFor normalised lateral forces, M, of wheels of wheeled robotsOgridIs the turning moment of the wheel of the normalized wheeled robot, MRgridIs the normalized advancing resistance moment of the wheel of the wheeled robot, MSgridThe normalized steering resistance moment of the wheels of the wheeled robot is obtained through a formula Tdyn＝Tgrid+Tupdate(a) Calculating to obtain an evaluation index T of the terrain trafficability in a single griddyn。
CN202210426426.8A 2022-04-22 2022-04-22 Terrain trafficability map construction method based on wheel-ground interaction Active CN114742310B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210426426.8A CN114742310B (en) 2022-04-22 2022-04-22 Terrain trafficability map construction method based on wheel-ground interaction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210426426.8A CN114742310B (en) 2022-04-22 2022-04-22 Terrain trafficability map construction method based on wheel-ground interaction

Publications (2)

Publication Number Publication Date
CN114742310A true CN114742310A (en) 2022-07-12
CN114742310B CN114742310B (en) 2022-09-16

Family

ID=82284307

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210426426.8A Active CN114742310B (en) 2022-04-22 2022-04-22 Terrain trafficability map construction method based on wheel-ground interaction

Country Status (1)

Country Link
CN (1) CN114742310B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160003792A1 (en) * 2014-05-21 2016-01-07 Purdue Research Foundation Functional soil maps
CN106154287A (en) * 2016-09-28 2016-11-23 深圳市普渡科技有限公司 A kind of map constructing method based on two-wheel speedometer Yu laser radar
US20180238690A1 (en) * 2017-02-17 2018-08-23 The Charles Stark Draper Laboratory, Inc. Stellar-Landscape/Horizon Image Navigation
CN110264572A (en) * 2019-06-21 2019-09-20 哈尔滨工业大学 A kind of terrain modeling method and system merging geometrical property and mechanical characteristic
CN110956651A (en) * 2019-12-16 2020-04-03 哈尔滨工业大学 Terrain semantic perception method based on fusion of vision and vibrotactile sense
CN111598142A (en) * 2020-04-26 2020-08-28 南京工业大学 Outdoor terrain classification method for wheeled mobile robot
CN111812299A (en) * 2020-07-17 2020-10-23 哈尔滨工业大学 Soil parameter identification method and device based on wheeled robot and storage medium
CN112389440A (en) * 2020-11-07 2021-02-23 吉林大学 Vehicle driving risk prediction method in off-road environment based on vehicle-road action mechanism

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160003792A1 (en) * 2014-05-21 2016-01-07 Purdue Research Foundation Functional soil maps
CN106154287A (en) * 2016-09-28 2016-11-23 深圳市普渡科技有限公司 A kind of map constructing method based on two-wheel speedometer Yu laser radar
US20180238690A1 (en) * 2017-02-17 2018-08-23 The Charles Stark Draper Laboratory, Inc. Stellar-Landscape/Horizon Image Navigation
CN110264572A (en) * 2019-06-21 2019-09-20 哈尔滨工业大学 A kind of terrain modeling method and system merging geometrical property and mechanical characteristic
CN110956651A (en) * 2019-12-16 2020-04-03 哈尔滨工业大学 Terrain semantic perception method based on fusion of vision and vibrotactile sense
CN111598142A (en) * 2020-04-26 2020-08-28 南京工业大学 Outdoor terrain classification method for wheeled mobile robot
CN111812299A (en) * 2020-07-17 2020-10-23 哈尔滨工业大学 Soil parameter identification method and device based on wheeled robot and storage medium
CN112389440A (en) * 2020-11-07 2021-02-23 吉林大学 Vehicle driving risk prediction method in off-road environment based on vehicle-road action mechanism

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
丁亮等: "基于应力分布的月球车轮地相互作用地面力学模型", 《机械工程学报》 *
丁亮等: "基于月球车轮地作用地面力学积分模型的月壤力学参数辨识方法", 《航空学报》 *
邓宗全等: "月壤特性对月球车轮地相互作用力的影响", 《哈尔滨工业大学学报》 *

Also Published As

Publication number Publication date
CN114742310B (en) 2022-09-16

Similar Documents

Publication Publication Date Title
CN112356830B (en) Intelligent parking method based on model reinforcement learning
CN110264572A (en) A kind of terrain modeling method and system merging geometrical property and mechanical characteristic
CN111598142B (en) Outdoor terrain classification method for wheeled mobile robot
CN106780458B (en) Point cloud framework extraction method and device
Triest et al. Tartandrive: A large-scale dataset for learning off-road dynamics models
Wiberg et al. Control of rough terrain vehicles using deep reinforcement learning
CN105034986A (en) On-line identification method and on-line identification device for steering characteristics of drivers
Ewen et al. These maps are made for walking: Real-time terrain property estimation for mobile robots
CN116972865A (en) Complex off-road environment navigation method and system based on vehicle maneuverability evaluation
CN115639823A (en) Terrain sensing and movement control method and system for robot under rugged and undulating terrain
CN114742310B (en) Terrain trafficability map construction method based on wheel-ground interaction
Han et al. A new open-source off-road environment for benchmark generalization of autonomous driving
Sivaprakasam et al. TartanDrive 2.0: More Modalities and Better Infrastructure to Further Self-Supervised Learning Research in Off-Road Driving Tasks
CN115374498B (en) Road scene reconstruction method and system considering road attribute characteristic parameters
Inotsume et al. Adaptive terrain traversability prediction based on multi-source transfer gaussian processes
CN114153216B (en) Lunar surface path planning system and method based on deep reinforcement learning and block planning
CN115205478A (en) Method and system for realizing ground physical characteristic inference framework based on vision
Shirkhodaie et al. Soft computing for visual terrain perception and traversability assessment by planetary robotic systems
Bouraine et al. NDT-PSO, a new NDT based SLAM approach using particle swarm optimization
Castejón et al. Compact modeling technique for outdoor navigation
Shirkhodaie et al. Traversable terrain modeling and performance measurement of mobile robots
Katiyar et al. Path Planning and Obstacle Avoidance in CG Space of a 10 DOF Rover using RRT
Spenko Hazard avoidance for high-speed rough-terrain unmanned ground vehicles
Zhou et al. Navigation Approach For Lunar Rover Based On Slip Prediction.
Adiyatov Path planning framework for unmanned ground vehicles on uneven terrain

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