CN103786806B - A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof - Google Patents

A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof Download PDF

Info

Publication number
CN103786806B
CN103786806B CN201410025261.9A CN201410025261A CN103786806B CN 103786806 B CN103786806 B CN 103786806B CN 201410025261 A CN201410025261 A CN 201410025261A CN 103786806 B CN103786806 B CN 103786806B
Authority
CN
China
Prior art keywords
leg
mode
robot
motion
shank
Prior art date
Application number
CN201410025261.9A
Other languages
Chinese (zh)
Other versions
CN103786806A (en
Inventor
丁希仑
杨帆
徐坤
彭赛金
Original Assignee
北京航空航天大学
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 北京航空航天大学 filed Critical 北京航空航天大学
Priority to CN201410025261.9A priority Critical patent/CN103786806B/en
Publication of CN103786806A publication Critical patent/CN103786806A/en
Application granted granted Critical
Publication of CN103786806B publication Critical patent/CN103786806B/en

Links

Abstract

The invention discloses a kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof, belong to robotics.First described method carries out the Extraction of Topographic Patterns of robot position; Then determine that the landform adaptive capacity of often kind of mode of motion is gone forward side by side line ordering; Determine landform adaptive criteria and the integrated adaptability index of often kind of mode of motion, correction is carried out to integrated adaptability index and obtains optimal movement pattern, finally carry out mode of motion switching.Multi-locomotion mode intelligent switch method provided by the invention, merge different sensor informations, comprehensively be extracted environment geometry and the physical features of restriction different motion pattern, considered switching cost between different motion pattern and speed simultaneously, stability lays particular stress on demand; Changing method is quick, easy, the waste of the possibility be absorbed in when avoiding wheel type movement and legged walking energy, instructs the motion realizing the fast and stable of robot in complex environment.

Description

A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof
Technical field
The invention belongs to robotics, relate to one and there is multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method.
Background technology
Legged mobile robot, by the complex environment such as muddy road surface, rugged mountain region, has stronger adaptive capacity to environment, but has shortcomings such as controlling difficult, energy consumption height.Wheeled robot can be quick, and efficiently by flat road surface, but its obstacle climbing ability is poor.The robot motion that is combined into of wheel, leg mode of motion brings new facility, but influences each other between legged walking and wheel type movement in actual motion, limits speed and the stability of robot motion.Explore better leg, wheel motion combination, the mechanism of design leg, wheel disengaging movement is the current problem needing to solve.
In mobile robot's research, robot only has single movement pattern mostly, limits its field of application.Moveable robot movement pattern is a development tendency of mobile robot to diversified development by simplification.Designing the intelligent switch method between six biped robots and multi-locomotion mode with multi-motion modes is the current problem needing to solve.
Summary of the invention
The present invention, for realizing quick, stable motion in complex environment, avoids leg formula and wheel type movement to influence each other, and provides a kind of multi-functional leg wheel composite machine people, and provides one independently to analyze environment-identification, the method that intelligent decision multi-locomotion mode switches.
First the present invention provides a kind of multi-functional leg wheel composite machine people, and described leg wheel composite machine people evenly arranges six legs around body, and described six legs are circumferentially symmetrical, and every bar leg comprises hip, thigh and shank three sections.In the present invention, shank is set to the shape that curves inwardly, and wheel is set in knee, touch force sensor is set at shank foot end, shank is also provided with infrared distance sensor.Robot body installs binocular stereo camera.
Based on above-mentioned multi-functional leg wheel composite machine people, the present invention also provides a kind of described multi-functional leg wheel composite machine people multi-locomotion mode intelligent switch method to comprise the steps:
The first step, carries out the Extraction of Topographic Patterns of robot position;
Second step, determines the landform adaptive capacity of often kind of mode of motion, line ordering of going forward side by side; Described landform adaptive capacity comprises slope stability, obstacle climbing ability and being absorbed in property;
3rd step, determines the landform adaptive criteria of often kind of mode of motion;
4th step, determines the integrated adaptability index of often kind of mode of motion; For often kind of mode of motion, get the maxim of its landform adaptive criteria and aviation value respectively as this mode of motion integrated adaptability critical for the evaluation;
5th step, revises integrated adaptability index, and decision-making obtains optimal movement pattern;
6th step, carries out mode of motion switching according to optimal movement pattern; Described mode of motion comprises wheeled, imitative mammal, imitative insect and mixed mode; Described mode of motion switch comprise that the walking of imitative mammal switches mutually with wheel type movement pattern, imitative insect walk and mutually switch with wheel type movement, mix walking mode with wheel type movement switches mutually, imitate mammal walk and imitate insect walk mutually switch, imitate mammal walk and mixed mode walk mutually switch, mixed mode walks and imitates insect and walk and mutually switch.
The invention has the advantages that:
(1) multi-functional leg wheel composite machine people provided by the invention adopts semisphere protective case; under same volume, there is larger face area; its surface can cover more solar cells, simultaneously for the built-in system such as sensor, treater of robot provides better protection.The leg distribution of circumference symmetry makes robot can realize multi-motion mode, can realize no-radius simultaneously and turn.Special leg wheel mechanism design meets wheel type movement and legged walking particular configuration separately, and easy switching, avoids wheel simultaneously and is placed in foot end or knee joint place and causes mutual interference between robot wheel type movement and the motion of leg formula.Binocular stereo camera is hidden in the semisphere protective case of robot at ordinary times; stretched out by mast during detection; reduce the loss of valuable sensor; the The Cloud Terrace of mast top has three degree of freedom; multiple visual angle can be realized measure and keep stable; by the fusion of multiple sensor, map structuring and landform hardness test can be completed.
(2) multi-locomotion mode intelligent switch method provided by the invention, merge different sensor informations, comprehensively be extracted environment geometry and the physical features of restriction different motion pattern, considered switching cost between different motion pattern and speed simultaneously, stability lays particular stress on demand.The integrated adaptability critical for the evaluation proposed calculates fast, and form is simple, better can reflect the comformability of different motion pattern to environment.Changing method between the different motion pattern provided is quick, easy.Can high-speed decision switch to the mode of motion of optimum current environment by this intelligent switch method robot, the waste of the possibility be absorbed in when avoiding wheel type movement and legged walking energy, instructs the motion realizing the fast and stable of robot in complex environment.
Accompanying drawing explanation
Figure 1A is the structural representation of multi-functional compound leg wheel robot provided by the invention;
The single leg structural representation of multi-functional compound leg wheel robot of Figure 1B for providing in the present invention;
Fig. 2 is that in the present invention, map structuring DEM schemes;
Robot hardness test configuration schematic diagram when Fig. 3 A is wheel type movement in the present invention;
Robot hardness test configuration schematic diagram when Fig. 3 B is leg formula motion in the present invention;
Fig. 4 A ~ 4D is multi-functional leg wheel composite machine people four kinds of mode of motion schematic diagrams in the present invention, is followed successively by imitative mammal walking mode, imitative insect walking mode, mixing walking mode and wheel type movement pattern;
Fig. 5 is that wheel type movement and imitative insect, imitative mammalian motor mode of motion switch schematic diagram mutually;
Fig. 6 is that wheel type movement and hybrid motion pattern switch schematic diagram mutually;
Fig. 7 is intelligent switch method diagram of circuit provided by the invention.
In figure:
1. main body; 2. semisphere protective case; 3. leg; 4. wheel; 5. touch force sensor;
6. infrared distance sensor; 7. binocular stereo camera; 8. The Cloud Terrace; 9. hip; 10. thigh;
11. shanks; 12. hip steering wheels; 13. thigh steering wheels; 14. shank steering wheels; 15. wheel drive steering wheels;
16. hip steering wheel output shafts; 17. thigh steering wheel output shafts;
18. shank steering wheel output shafts; 19. wheel drive steering wheel output shafts;
301. leg A; 302. leg B; 303. leg C; 304. leg D; 305; Leg E;
306. leg F.401. take turns A; 402. take turns B; 403. take turns C; 404. take turns D;
405. take turns E; 406. take turns F;
Detailed description of the invention
Below in conjunction with drawings and Examples, the present invention is described in detail.
The invention provides a kind of multi-functional compound leg wheel robot; as shown in Figure 1; described robot comprises semisphere protective case 2 and body 1; on body 1, surrounding is evenly provided with the leg 3 of six identical leg wheel composite structures; described leg 3 as shown in Figure 1B; comprise hip 9, thigh 10 and shank 11 3 sections, wherein shank 11 middle bent place is provided with wheel 4.Described robot also comprises the Bumblebee of the binocular stereo camera 7(PointGrey company for terrain detection tM), IMU(XSENSMTi-100) and the touch force sensor 5(FSR4020.5 that installs of calf vola) and shank on the infrared distance sensor 6(SHARPGP2D12 of device).Described semisphere protective case 2 is aluminium alloy structure, and top has inspection door, semisphere protective case surface coverage solar cell.
Described robot body 1 is disc-shaped structure, is connected between body 1 with semisphere protective case 2 by six bolts.Article six, leg 3 is circumferentially symmetrical along robot body 1.On every bar leg 3 of described robot, the length ratio of hip 9, thigh 10 and shank 11 3 sections of legs is 1:9:10.Described hip 9 is connected with body 1 by hip yaw joint, is connected with thigh 10 by hip pitch joint, and described thigh 10 is connected with shank 11 by knee joint; The root of described hip 9 is connected with for driving hip 9 relative to the hip steering wheel 12 of body 1 yaw, and the output shaft 16 of described hip steering wheel 12 is fixedly connected with body 1.
Described hip 9 front end is connected with for driving thigh 10 relative to the thigh steering wheel 13 of body 1 pitch, and described thigh 10 is fixedly connected with the output shaft 17 of thigh steering wheel 13.
Described shank 11 is connected with the shank steering wheel 14 for driving toe raise, gathering, described thigh 10 is fixedly connected with the output shaft 18 of shank steering wheel 14.Described shank 11 is curved shape, and bending angle is 165 degree, bending to body 1 direction.Described shank 11 knee is fixed with 360 ° and rotates steering wheel 15, and the output shaft 19 of steering wheel 15 directly connects wheel 4 by spline, rotates for wheels 4.
Described binocular stereo camera 7 is arranged on the inner liftable The Cloud Terrace 8 of semisphere protective case 2, and the inspection door arranged by semisphere protective case 2 top stretches out and detects, and The Cloud Terrace 8 has three degree of freedom.
Described infrared distance sensor 6 is arranged on the shank 11 of vola and wheel 4 midway location, and the position of general selected distance vola 5cm, is connected with shank 11 by two screws.Described infrared distance sensor 6 is for measuring ground and infrared distance sensor 6 spacing, the distance between indirect calculation vola and ground.
Described touch force sensor 5 is placed in vola, by paste adhesive on the vola of quality of rubber materials, for measuring the contact acted at the bottom of robot foot.
Multifunctional wheel leg composite machine people provided by the invention has multi-motion modes, concrete be divided into wheel type movement and legged walking, and wherein legged walking can be divided into again imitative mammal mode to walk, imitative insect mode is walked, hybrid mode walking etc.The motion of different mode has features and adapt circumstance, by binocular solid camera in the present invention, the data message that touch force sensor and infrared distance sensor gather, modeling is carried out to residing terrain environment, extract the key feature (geometric properties and physical hardness feature) of restriction different motion pattern, analyze the compatible with environment of each mode of motion of contrast, consider the switching cost between different motion pattern and speed simultaneously, stability lays particular stress on demand, the suitable mode of motion of diverse location on decision making package map, and the changing method provided between different motion pattern, for realizing the quick of robot, stable motion provides a kind of effective ways.
The multi-locomotion mode intelligent switch method provided in the present invention, as shown in Figure 7, realizes especially by following steps:
The first step, the Extraction of Topographic Patterns of map structuring and restriction motion;
By binocular solid camera 7, adopt ripe algorithm (as SLAM algorithm, see reference document [1]: A.J.Davison, " Real-timesimultaneouslocalisationandmappingwithasingleca mera, " inComputerVision, 2003.Proceedings.NinthIEEEInternationalConferenceon, 2003, pp.1403-1410.) map structuring is carried out, obtain the landform altitude dot matrix cloud (DEM figure) in certain area residing for robot, as shown in Figure 2, according to resolution requirement, rasterizing is again schemed to this DEM, here the size arranging grid is 2cm × 2cm, robot body center robot system of axes Σ is fixed in foundation rworld coordinate system Σ g.(X i, Y j) represent world coordinate system Σ ga grid on middle X-Y plane, Z ij=f (X i, Y j) represent grid (X i, Y j) corresponding height value.By robot at world coordinate system Σ gx-Y plane inner projection, suppose that robot center projection is positioned at grid (X i, Y j) inner, by the grid region ([x of minimum envelop robot projection i-n, x i+n], [y j-m, y j+m]) as a projection block, the relief feature of the block that projects after moving a grid according to this projection block and Robot sense of motion, the corresponding optimum mode of motion of decision-making, is grid (X i, Y j) optimum mode of motion.Definition robot center is at world coordinate system Σ gmiddle coordinate is
Described relief feature comprises the gradient, obstacle clearing capability, texture and rigidity, and concrete extracting method is as follows:
1, the gradient: roll angle during matching robot different motion pattern and pitch angle, concrete analysis step is as follows,
Robot coordinate system Σ rwith world coordinate system Σ gbetween pose transformation relation can pass through homogeneous matrix T gRrepresent:
Wherein, roll angle ψ is around X-axis angle of rotation, and pitching angle theta is around Y-axis angle of rotation, yaw angle for around Z axis angle of rotation, be respectively robot center at world coordinate system Σ gin x, y, z coordinate, c, s are respectively cos, sin abridge.
Known machine people six foot end is at robot coordinate system Σ rin position be k=1 ... 6, then foot end is at world coordinate system Σ gin position k=1 ... 6, T gRfor above-mentioned homogeneous transform matrix.Simulating each foothold landform value by DEM figure is k=1 ... 6, each foot end of definition robot and earth surface situation c={ Δ 1, Δ 2, Δ 3, Δ 4, Δ 5, Δ 6, wherein, k=1 ... 6.
Work as Δ kwhen=0, the foot end representing robot just in time contacts with ground;
Work as Δ kduring > 0, represent the foot end of robot above ground;
Work as Δ kduring < 0, the foot end representing robot is absorbed in ground.
Existing known device people yaw angle robot center (x, y) coordinate with k=1 ... 6 require the pitching angle theta estimating robot, roll angle ψ, robot center z-axis coordinate each foot end of robot is contacted with ground as far as possible.Be a complicated nonlinear relationship between constrained input, the analytic relationship obtaining display is more difficult, therefore builds BP neural network respectively for often kind of mode of motion here and carries out fitted data and solve.
The BP neural network built comprises two hidden-layer, and input layer is 8, and interlayer comprises 10 respectively, 15 neurons, and output layer node is 3.
Input: foothold place landform value k=1 ... 6, robot yaw angle body central
Export: pitching angle theta, roll angle ψ, body central coordinate
Training sample: build map and three dimensional model for robot, imports in ADAMS simulation software, and the different map location of random selecting carries out simulation calculation, obtains the training sample of neural network.
2, obstacle clearing capability:
Calculating robot moves to the maximum height e crossing over step class obstacle needed for next grid along sense of motion i, j:
e i,j=max p,q(z i+a+p,j+b+q-z i+p,j+q),p∈{-n,…,0,…,n},q∈{-m,…,0,…,m}
Wherein, 2n, 2m are robot projection region length and width, and m, n are integer.
3, texture:
Carry out analysis of texture for the current acquisition of robot map within the vision, concrete step is as follows:
(1) extract N × M grating image and represent texture sample, wherein N, M choose too small, cannot represent textural characteristics, choose excessive then texture analysis consuming time long, get N=18n here, M=18m;
(2) with orthogonal subspaces reflection method, dimension-reduction treatment is carried out to N × M grating image, and the characteristic root of image array after extracting dimensionality reduction, this characteristic root has to a certain degree reacted the textural characteristics of image, the characteristic root that namely different texture characteristic image is corresponding different.
(3) extract different characteristic texture sample individual features root, as rubble, meadow, sand, soil etc., training landform different characteristic texture recognition neural network, completes map zones of different texture recognition.
4, hardness:
Carry out landing stiffness test based on infrared distance sensor 6 and vola touch force sensor 5, when robot detects different texture feature, hardness test is carried out in suggestion, and obtain the stiffness characteristics of landform, prevent ground excessively soft, robot is embedded.During concrete robot wheel type movement pattern, test configuration is as Fig. 3 A, for one of them leg A301, leg A301 is equipped with infrared distance sensor 6 and touch force sensor 5, after shank steering wheel 14 drives shank 11 to be contracted to the limit, thigh steering wheel 13 drives thigh 10 upwards pitch, after adjustment vola contacts with scheduled measurement position, extrude ground at the bottom of robot foot downwards and reach greatest measurement to touch force sensor 5, record infrared distance sensor 6 and vola touch force sensor 5 data value in this process, matching landing stiffness.Leg formula exercise test configuration is as Fig. 3 B, after adjustment vola contacts with scheduled measurement position, extrude ground to touch force sensor 5 at the bottom of robot foot downwards and reach infrared distance sensor 6 and vola touch force sensor 5 data value in this process of greatest measurement record, matching landing stiffness.。
Described matching landing stiffness adopts Bekker equation:
P C=k eqD l
Wherein, P cthe footprint pressure on vola and ground, l and k eqbe landform fitting parameter, D is direction perpendicular to the ground deformation.P cmeasured by touch force sensor and infrared distance sensor respectively with D and obtain.By organizing the P of measurement more c, D, obtains landform fitting parameter l and k eq.
Second step, sorts to the landform comformability of different motion pattern;
Occurring in nature, mammal walks in kicking mode, has an effect mainly through knee joint, and sense of motion is along kicking direction.Insect is moved by the yaw of leg, and position of mainly having an effect is hip joint, and sense of motion is along yaw direction.Multi-functional six sufficient leg wheel composite machine people have joint configuration mode flexibly, and robot has six legs, is followed successively by leg A301, leg B302, leg C303, leg D304, leg E305 and leg F306, bionically can realize multi-motion modes.As shown in Figure 4 A, robot adopts imitative mammal mode to walk, and robot leg B302, D304, F306 are for leading leg, and supporting leg A301, C303, E305 have an effect mainly through shank steering wheel 14, promotes body 1 and moves along leg A301 direction; As shown in Figure 4 B, robot adopts imitative insect mode to walk, and robot leg B302, D304, F306 are for leading leg, and supporting leg A301, C303, E305 have an effect mainly through hip steering wheel 12, promotes body 1 and moves along vertical leg A301 direction.Further, by analysis integrated to nature biotechnology mode of motion, propose a kind of hybrid motion pattern, as shown in Figure 4 C, leg B302, D304, F306 are for leading leg, and supporting leg leg C303, leg E305 are had an effect by hip steering wheel 12, and leg A301 is had an effect by shank steering wheel 14, promote body move along leg A301 direction to mix mammal and insect walking manner, robot adopts hybrid mode 3+3 walking.Multi-functional leg wheel composite machine people can also be switched by wheel leg, flat road surface realizes the motion of wheel row, as shown in Figure 4 D.
Multi-functional leg wheel composite machine people has multi-motion modes, the advantage of each mode of motion of com-parison and analysis and landform adaptive capacity, for appropriate exercise model selection provides foundation.The landform restriction condition of concrete different motion pattern is mainly:
1, slope stability:
Robot different motion pattern has different slope stabilities, here energy stabilization nargin ESM(energystabilitymargin is adopted) index weighs, and concrete robot turn to overcome the minimum value that gravity does work by supporting border along each bar when center of gravity is in directly over this support border and be defined as energy stabilization nargin.
When robot has minimum safe energy stabilization nargin, four kinds of mode of motion robots are analyzed:
For the sequence of pitch angle stability and pitch angle limit:
Imitative mammalian motor (40 °) > mixed mode motion (30 °) > wheel type movement (25 °) > imitates insect motion (20 °)
For the sequence of roll angle stability and roll angle limit:
Imitative insect motion (40 °) > mixed mode motion (30 °) > imitates mammalian motor (20 °) ≈ wheel type movement (20 °)
2, obstacle climbing ability;
The relatively obstacle climbing ability of different motion pattern:
(1) maximum height of surmountable obstacle of wheel type movement is generally wheel radius R;
(2) the imitative maximum obstacle detouring of insect motion pattern is generally body height H;
(3) imitative mammal obstacle clearing capability is seriously by step-length and foothold widths affect, and maximum obstacle detouring is generally less than body height H, and choosing maximum safe obstacle clearing capability is h=H/2, h > R here;
(4) the obstacle climbing ability obstacle performance of hybrid motion pattern is between insect and mammalian motor mode, is restricting along on the axis of sense of motion by mammalian motor mode, and maximum height of surmountable obstacle is h, and other four leg maximum height of surmountable obstacle are H.
Obstacle climbing ability sequence is:
Imitative insect motion (H) > mixed mode motion ([h, H]) > imitates mammalian motor (h) > wheel type movement (R).
3, being absorbed in property:
Different mode of motioies, also different to the requirement of landing stiffness, ground is excessively soft, and as moonscape covers the thick moon ash of one deck, robot may be embedded, and is finally difficult to depart from.From adopting Bekker hardness fit equation in the first step of the present invention, deflection is mainly by landform fitting parameter l and k eqand footprint pressure P cimpact, robot itself can only prevent from being absorbed in by reducing footprint pressure, and under identical weight, footprint pressure mainly affects by robot and contact area of ground, when namely adopting wheel type movement pattern, the quantity of synchronization supporting leg when wheel quantum count and the motion of leg formula.During legged walking, different motion pattern is less on footprint pressure impact, and pressure mainly affects by gait, holds the area of contact with ground contact surface sum wheel and ground according to each foot, calculates footprint pressure sequence to be:
Wheel row (6.7kpa) ≈ 3+3 gait (6.7kpa) >4+2 gait (5kpa) >5+1 gait (4kpa).
Wherein, described 3+3 gait, 4+2 gait, 5+1 gait, the numeral (3,4,5) before plus sige "+" represents the quantity of the supporting leg of any time in gait cycle, after numeral (3,2,1) represent the quantity of leading leg.
3rd step, determines the adaptive criteria of often kind of mode of motion;
The landform comformability of each landform restriction feature and each mode of motion is normalized, obtain the landform adaptive criteria of each mode of motion, wherein, subscript wheel, mammal, insect, hybrid represent wheel type movement, imitative mammal walking, imitative insect walking, the data of mixing walking corresponding to each mode of motion respectively.Concrete:
Obstacle detouring adaptive criteria:
IO wheel = e wheel R IO mammal = e mammal h IO in sec t = e in sec t H IO hybrid = e hybrid h
Wherein, IO is obstacle detouring adaptive criteria, represents the contrast between the obstacle height that robot need be crossed and robot obstacle climbing ability.Wherein e calculates by first step method 2 obstacle height obtained, and R, h, H are respectively the maximum height of surmountable obstacle of corresponding sports pattern (wheeled, imitative mammal and imitative insect motion pattern).
Gradient adaptive criteria:
Wherein, I θ is pitch angle adaptive criteria, and I ψ is roll angle adaptive criteria; Represent the contrast between the attitude angle of robot and extreme attitude angle.θ, ψ are respectively pitch angle and the roll angle of the different motion Mode Robot obtained by the first step, and the corresponding angle of denominator is the stable angle limit being obtained corresponding sports pattern by second step.
Hardness adaptive criteria:
IS wheel = G 6 &times; S wheel &times; k eq &times; D max l IS mammal = G n c &times; S foot &times; k eq &times; D max l IS in sec t = G n c &times; S foot &times; k eq &times; D max l IS hybrid = G n c &times; S foot &times; k eq &times; D max l
Wherein, IS is hardness adaptive criteria, represents the contrast of footprint pressure when the footprint pressure that robot produces and the maximum amount of being absorbed in.N cfor the number of supporting leg, n cdetermined by the gait selected, 3+3 gait n c=3; 4+2 gait n c=4; 5+1 gait n c=5.S wheeland S footbe respectively single-wheel and monopodia and contact area of ground.G is robot weight, D maxfor the maximum deformation quantity of definition, k eq, l is landform fitting parameter.
The above-mentioned each adaptive criteria of integrating representation is as shown in table 1:
Table 1 is without mode of motion adaptive criteria computing formula
If each adaptive criteria in table 1 is greater than 1, expression is not suitable for, and makes this index equal 1 after normalization method.By the size of the numeric representation adaptive criteria between [0,1], 0 represents optimum, and 1 represents least suitable.
4th step, integrated adaptability index calculate;
To get in table 1 each row (i.e. often kind of mode of motion) maxim and this column average value as this mode of motion integrated adaptability critical for the evaluation, maxim characterizes main landform feature constraint, aviation value then characterizes landform on average suitable situation, and concrete example calculation is as shown in table 2.
Table 2: integrated adaptability index calculate
Wheel type movement Imitative mammal walking Imitative insect walking Mixing walking
Obstacle detouring comformability IO 0.2 0.2 0.1 0.1
Pitching comformability I α 0.2 0.3 0.4 0.3
Rolling comformability I β 0.4 0.3 0.3 0.3
Hardness comformability IS 0.1 0.1 0.1 0.1
Integrated adaptability index (0.4,0.25) (0.3,0.25) (0.4,0.25) (0.3,0.2)
5th step, the correction of integrated adaptability index, decision-making obtains optimal movement pattern.
The motion of robot not only restricts by topographic condition, simultaneously the impact of also controlled demand.Below according to switching cost, speed requirement and durability requirements are revised integrated adaptability index, for convenience of representing, below represent wheel type movement pattern with " W ", " M " represents imitative mammal walking mode, and " I " represents imitative insect walking mode, and " H " represents mixing walking mode.1, switching cost correction;
Due to the cost that switching between different mode of motioies can cause energy dissipation etc. certain, for avoiding causing frequent switching, according to the switching cost between different motion pattern, suitable correction is carried out to the integrated adaptability index of existing mode of motion, defines mutual switching cost correction between several mode of motion as shown in table 3:
Mutual switching cost correction between table 3 different motion pattern
Pattern W-M W-I W-H M-I M-H I-H
Correction (0.05,0.2) (0.05,0.2) (0.05,0.2) (0.05,0.1) (0.05,0.1) (0.05,0.1)
Concrete, the state of kinematic motion of such as current robot is W pattern, when carrying out switching cost correction, needs respectively to the integrated adaptability index of M, I, H adding corresponding mode of motion switches correction (W-M), (W-I), (W-H).
2, speed requirement correction;
Different motion pattern has different maximum movement speed, if speed is motion major consideration, then in the close situation of integrated adaptability index, more have a preference for wheel type movement, according to each maximum speed of motion situation, the speed requirement correction defining several mode of motion is as shown in table 4:
Table 4 speed requirement correction
Pattern W pattern M-mode I pattern H pattern
Correction (0,-0.1) (0,-0.05) (0,0) (0,+0.05)
After selecting this demand parameter, the integrated adaptability index of each mode of motion is added corresponding correction, carries out speed requirement correction.
3, durability requirements correction;
As having specific laying particular stress on for stability of motion, in the close situation of integrated adaptability index, more have a preference for the motion of leg formula.According to each motion stabilization implementations, the durability requirements correction defining several mode of motion is as shown in table 5:
Table 5 durability requirements correction
Pattern W pattern M-mode I pattern H pattern
Correction (0,+0.1) (0,0) (0,-0.05) (0,-0.1)
After selecting this demand parameter, the integrated adaptability index of each mode of motion is added corresponding correction, carries out durability requirements correction.
4, decision-making technique;
According to the 4th step, integrated adaptability index comprises two indices part, first is maxim, second is aviation value, first that first compares integrated adaptability index between different motion pattern, choose the mode of motion that first desired value is minimum, such as, if imitative mammal walking mode and mixing walking mode integrated adaptability index first place value are 0.3 in table 2, minimum, as equal in integrated adaptability index first, compare the second of integrated adaptability index again, value minimum in index for selection second, be (0.3 as mixed walking mode comprehensive fitness degree index, 0.2), for optimal movement pattern.
After different demand factor correction, corresponding optimum mode of motion also may change.Such as integrated adaptability index is,
W pattern M-mode I pattern H pattern
Integrated adaptability index (0.4,0.25) (0.3,0.25) (0.4,0.25) (0.3,0.2)
Choosing optimal movement pattern is H pattern.
Suppose that current kinetic pattern is M-mode, search the correction being switched to other three kinds of mode of motioies from M-mode in table 3, (M-W), (M-I), (M-H), and with corresponding W, I, H pattern systhesis adaptive criteria is added, then revised integrated adaptability index is:
W pattern M-mode I pattern H pattern
Integrated adaptability index (0.45,0.45) (0.3,0.25) (0.45,0.35) (0.35,0.3)
After switching cost correction, optimal movement pattern is M-mode.
Integrated adaptability index is after speed requirement correction:
W pattern M-mode I pattern H pattern
Integrated adaptability index (0.4,0.15) (0.3,0.20) (0.4,0.25) (0.3,0.25)
After the correction of speed requirement index, optimal movement pattern is M-mode.
Integrated adaptability index is after durability requirements correction:
W pattern M-mode I pattern H pattern
Integrated adaptability index (0.4,0.35) (0.3,0.25) (0.4,0.20) (0.3,0.1)
After the correction of durability requirements index, optimal movement pattern is H pattern.
By the optimal movement pattern that obtains after revising, compare with current kinetic pattern, if difference switches with regard to execution pattern, turn the 6th step, otherwise do not do and switch.
6th step, carries out the switching between multi-motion modes according to optimal movement pattern;
Multifunctional wheel leg composite machine people provided by the invention has multi-motion modes, and can switch fast according to the result of decision in above-mentioned 5th step between different motion pattern, concrete changing method is:
1, imitative mammal walking and the mutual changing method of wheel type movement pattern;
Robot is switched to the step of wheel type movement as shown in Fig. 5 (a) ~ 5 (d) from imitative mammal walking, and robot initial state is as shown in Fig. 5 (a), and concrete switch step is as follows:
The first step, leg B302, D304, F306 thigh steering wheel 13 drives thigh 10, and downwards pitch, shank steering wheel 14 drive shank 11 to collapse to shank 11 parallel to the ground and shank 11 and body 1 apart from being K, as shown in Fig. 5 (b).
Second step, reduce body 1 highly to taking turns after B402, D404, F406 land, leg A301, C303, E305 thigh steering wheel 13 drives thigh 10 upwards after pitch to maxim, shank steering wheel 14 drives shank 11 to continue to collapse to minimum value, thigh steering wheel 13 drive thigh 10 downwards pitch and shank steering wheel 14 drive shank 11 to extend to shank 11 is parallel to the ground and shank 11 is L (L<K) with body distance, as shown in Fig. 5 (c).
Upper three the thigh steering wheels 13 of 3rd step leg B302, D304, F306 drive thigh 10 upwards pitch, shank steering wheel 14 drives shank 11 to be contracted to the parallel to the ground and shank of shank 11 11 with body 1 apart from being L, so far six wheels 4 completely and earth surface, robot is wheel row state, as shown in Fig. 5 (d).
Robot is switched to the step of imitative mammal walking as shown in Fig. 5 (d) ~ 5 (a) from wheel type movement, and robot initial state is as shown in Fig. 5 (d).Concrete switch step is as follows:
Upper three the thigh steering wheels 13 of first step leg B302, D304, F306 drive thigh 10 pitch, and shank steering wheel 14 drives shank 11 to stretch, and it is K>L that body 1 is increased to ground distance, as shown in Fig. 5 (c).
Second step, upper three the shank steering wheels 14 of leg A301, C303, E305 drive shank 11 to shrink, thigh steering wheel 13 drive thigh 10 extend to each shank 11 perpendicular to the ground after, thigh steering wheel 13 drives thigh 10 downwards to land rear continuations support in pitch to vola, as shown in Fig. 5 (b).
3rd step, raise body 1 height to H, leg B302, D304, F306 move to corresponding foothold, and six legs are parallel, and robot is in imitative mammal walking initial condition, as shown in Fig. 5 (a).
2, imitative insect walking and the mutual changing method of wheel type movement;
Robot imitates step that insect walking is switched to wheel type movement according to the order shown in Fig. 5 (a) ~ 5 (d), consistent with the walk step that is switched to wheel type movement of imitative mammal.
Robot wheel type movement is switched to the step of imitative insect walking according to the order shown in Fig. 5 (d) ~ 5 (a), and to be switched to the step that imitative mammal walks consistent with wheel type movement.
3, walking mode and the mutual changing method of wheel type movement is mixed;
Robot is switched to the step of wheel type movement as shown in Fig. 6 (a) ~ 6 (d) from mixing walking mode, and robot initial state is as shown in Fig. 6 (a).Concrete switch step is as follows:
The first step, upper three the hip steering wheels 12 of leg B302, D304,306F drive hip 9 yaw to parallel with leg A301, thigh steering wheel 13 drives thigh 10 pitch downwards, shank steering wheel 14 drives shank 11 to collapse to the parallel to the ground and shank of shank 11 11 with body 1 apart from being K, as shown in Fig. 6 (b).
Second step, reduce body 1 highly to taking turns after B402, D404, F406 land, upper three the shank steering wheels 14 of leg A301, C303, E305 drive shank 11 to continue to draw in, after hip steering wheel 12 drives hip 9 yaw extremely parallel with axis of movement, thigh steering wheel 13 drive thigh 10 downwards pitch and shank steering wheel 14 drive shank 11 to extend to shank 11 parallel to the ground and shank 11 and body 1 apart from being L (L<K), as shown in Figure 6 (c).
3rd step, upper three the thigh steering wheels 13 of leg B302, D304, F306 drive thigh 10 upwards pitch, shank steering wheel 14 drives shank 11 to extend to the parallel to the ground and shank of shank 11 11 with body 1 apart from being L, so far six wheels 4 completely and earth surface, robot is wheel row state, as shown in Fig. 6 (d).
Robot is switched to the step of mixing walking mode according to the order shown in Fig. 6 (d) ~ 6 (a) from wheel type movement, and robot initial state is as shown in Fig. 6 (d).Concrete switch step is as follows:
Upper three the thigh steering wheels 13 of first step leg B302, D304, F306 drive thigh 10 pitch downwards, and shank steering wheel 14 drives shank 11 to stretch, and it is K>L that body 1 is increased to ground distance, as shown in Fig. 6 (c).
Upper three the shank steering wheels 14 of second step leg A301, C303, E305 drive shank 11 to shrink, thigh steering wheel 13 drive thigh 10 upwards pitch to each shank 11 perpendicular to the ground after, hip steering wheel 12 is regulated to drive hip 9 yaw, leg A301, C303, E305 is made to be 120 ° of rotational symmetry distributions, then thigh steering wheel 13 drives thigh 10 downwards to land rear continuations support in pitch to vola, as shown in Fig. 6 (b).
3rd step, after body 1 is highly increased to H, leg B302, D304, F306 move to corresponding foothold, and six legs distribute axisymmetricly, and robot is in mixing walking mode initial condition, as shown in Fig. 6 (a).
4, imitative mammal walking with imitate insect and to walk mutual changing method
Imitative mammal walking with imitate insect and walk and there is identical initial attitude as shown in Fig. 5 (a).
Switch to imitative insect walking mode from imitative mammal walking mode, as shown in Figure 4 A, might as well suppose that now leg B302, D304, F306 is for leading leg, leg A301, C303, E305 are supporting leg to robot initial.Concrete switch step is as follows:
The first step, adjustment leg A301, C303, E305 supporting body 1 parallel motion 1/2nd stride.
Second step adjusts leg B302, leg D304, the foot point that falls of leg F306 is extremely walked with imitative insect, and initial condition is identical, and now robot is in imitative insect walking mode initial attitude, can start the walking of imitative insect.
Switch to imitative mammal walking mode from imitative insect walking mode, as shown in Figure 4 C, might as well suppose that now leg B302, leg D304, leg F306 are for leading leg, leg A301, leg C303, leg E305 are supporting leg to robot initial.Concrete switch step is as follows:
The first step, adjustment leg A301, leg C303, leg E305 supporting body parallel motion 1/2nd stride.
Second step adjusts leg B302, leg D304, the foot point that falls of leg F306 is extremely walked with imitative mammal, and initial condition is identical, and now robot is in imitative mammal walking mode initial attitude, can start the walking of imitative mammal.
5, imitative mammal walking and mixed mode are walked mutual changing method
Switch to mixed mode walking mode from imitative mammal walking mode, as shown in Figure 4 D, might as well suppose that leg A301, leg C303, leg E305 are supporting leg, leg B302, leg D304, leg F306 are for leading leg for robot initial.Concrete switch step is as follows:
Lead leg leg B302, leg D304, upper three hip steering wheels 12 of leg F306 of the first step, adjustment drive hip 9 yaw, and make it be 120 ° of distributions, be converted to supporting leg after landing, body 1 is highly H.
Second step, leg A301, leg C303, leg E305 are converted to and lead leg, adjustment leg A301, upper three the hip steering wheels 12 of leg C303, leg E305 drive hip 9 yaw circumferentially symmetrical to each leg, now robot is in mixed mode initial attitude, as shown in Figure 6 (a), mixing walking can be started.
Switch to imitative mammal walking mode from mixed mode walking mode, robot initial as shown in Figure 4 D, might as well suppose that leg A301, leg C303, leg E305 are supporting leg, and leg B302, leg D304 leg F306 are for leading leg, and concrete switch step is as follows:
The first step, adjusts upper three the hip steering wheels 12 of leg B302, leg D304 leg F306 of leading leg and drives hip 9 yaw, make it parallel with leg A301, be converted to supporting leg after landing.
Second step, leg A301, leg C303, leg E305 are converted to and lead leg, adjustment leg A301, upper three the hip servo driving hip yaws of leg C303, leg E305 are parallel with leg A301 to each leg, now robot is in imitative mammal walking initial attitude, as shown in Fig. 5 (a), the walking of imitative mammal can be started.
6, mixed mode walking with imitate insect and to walk mutual changing method
Mixed mode walking with imitate walk mutual changing method and imitative mammal of insect and walk identical with mixed mode mutual changing method of walking.

Claims (6)

1. a multi-functional leg wheel composite machine people's multi-locomotion mode intelligent switch method, is characterized in that, comprise the steps:
The first step, carries out the Extraction of Topographic Patterns of robot position;
Second step, determines the landform adaptive capacity of often kind of mode of motion, line ordering of going forward side by side; Described landform adaptive capacity comprises slope stability, obstacle climbing ability and being absorbed in property;
3rd step, determines the landform adaptive criteria of often kind of mode of motion;
4th step, determines the integrated adaptability index of often kind of mode of motion; For often kind of mode of motion, get the maxim of its landform adaptive criteria and aviation value first and second as this mode of motion integrated adaptability index respectively;
5th step, integrated adaptability index is revised, first that first compares integrated adaptability index between different motion pattern, choose the mode of motion that first desired value is minimum, as equal in integrated adaptability index first, compare the second of integrated adaptability index again, minimum in index for selection second is optimal movement pattern, as final optimal movement pattern;
6th step, carries out mode of motion switching according to optimal movement pattern; Described mode of motion comprises wheeled, imitative mammal, imitative insect and mixed mode; Described mode of motion switch comprise that the walking of imitative mammal switches mutually with wheel type movement pattern, imitative insect walk and mutually switch with wheel type movement, mix walking mode with wheel type movement switches mutually, imitate mammal walk and imitate insect walk mutually switch, imitate mammal walk and mixed mode walk mutually switch, mixed mode walks and imitates insect and walk and mutually switch.
2. the multi-locomotion mode intelligent switch method of a kind of multi-functional leg wheel composite machine people according to claim 1, is characterized in that: described Extraction of Topographic Patterns comprises the gradient, obstacle clearing capability, texture and rigidity characteristic and extracts, and concrete extracting method is as follows:
(1), the gradient: roll angle during matching robot different motion pattern and pitch angle, concrete analysis step as follows,
Robot coordinate system Σ rwith world coordinate system Σ gbetween pose transformation relation by homogeneous matrix T gRrepresent:
Wherein, roll angle ψ is around X-axis angle of rotation, and pitching angle theta is around Y-axis angle of rotation, yaw angle for around Z axis angle of rotation, be respectively robot center at world coordinate system Σ gin x, y, z coordinate, c, s are respectively cos, sin abridge;
Known machine people six foot end is at robot coordinate system Σ rin position be k=1 ... 6, then foot end is at world coordinate system Σ gin position k=1 ... 6; Above-mentioned homogeneous matrix simulates each foothold landform value by DEM figure k=1 ... 6, each foot end of definition robot and earth surface situation c={ Δ 1, Δ 2, Δ 3, Δ 4, Δ 5, Δ 6, wherein, k=1 ... 6;
Work as Δ kwhen=0, the foot end representing robot just in time contacts with ground;
Work as Δ kduring >0, represent the foot end of robot above ground;
Work as Δ kduring <0, the foot end representing robot is absorbed in ground;
Existing known device people yaw angle robot center (x, y) coordinate with k=1 ... 6; Require the pitching angle theta estimating robot, roll angle ψ, robot center z-axis coordinate each foot end of robot is contacted with ground as far as possible; Build BP neural network respectively for often kind of mode of motion to carry out fitted data and solve;
(2), obstacle clearing capability:
Calculating robot moves to the maximum height e crossing over step class obstacle needed for next grid along sense of motion i,j:
e i,j=max p,q(z i+a+p,j+b+q-z i+p,j+q),p∈{-n,…,0,…,n},q∈{-m,…,0,…,m}
Wherein, 2n, 2m are robot projection region length and width, and m, n are integer; (3), texture:
Carry out analysis of texture for the current acquisition of robot map within the vision, concrete step is as follows:
I () is extracted N × M grating image and is represented texture sample;
(ii) with orthogonal subspaces reflection method, dimension-reduction treatment is carried out to N × M grating image, and the characteristic root of image array after extracting dimensionality reduction;
(iii) extract different characteristic texture sample individual features root, training landform different characteristic texture recognition neural network, completes map zones of different texture recognition;
(4), hardness:
Matching landing stiffness adopts Bekker equation:
P c=k eqD l
Wherein, P cthe footprint pressure on vola and ground, l and k eqbe landform fitting parameter, D is direction perpendicular to the ground deformation; P cmeasured by touch force sensor and infrared distance sensor respectively with D and obtain; By organizing the P of measurement more c, D, obtains landform fitting parameter l and k eq.
3. the multi-locomotion mode intelligent switch method of a kind of multi-functional leg wheel composite machine people according to claim 1, it is characterized in that: described slope stability, when robot has minimum safe energy stabilization nargin, four kinds of mode of motion robots carry out sorting:
Pitch angle stability is sorted:
Imitative mammalian motor > mixed mode motion > wheel type movement > imitates insect motion
Described four kinds of mode of motion robot pitch angle limits are respectively 40 °, 30 °, 25 °, 20 °;
Roll angle stability is sorted:
Imitative insect motion > mixed mode motion > imitates mammalian motor ≈ wheel type movement
Described four kinds of mode of motion robot roll angle limits are respectively 40 °, 30 °, 20 °, 20 °;
Described obstacle climbing ability, the sequence of four kinds of mode of motioies is:
Imitative insect motion > mixed mode motion > imitates mammalian motor > wheel type movement;
Described being absorbed in property is embodied by the sequence of footprint pressure, and concrete sequence is:
Wheel row mode ≈ 3+3 gait >4+2 gait >5+1 gait.
4. the multi-locomotion mode intelligent switch method of a kind of multi-functional leg wheel composite machine people according to claim 1, it is characterized in that: the landform adaptive criteria of the 3rd often kind of mode of motion described in step, use subscript wheel, mammal, insect, hybrid represents wheel type movement, imitative mammal walking, imitative insect walking, the data of mixing walking corresponding to each mode of motion respectively, is specially:
Obstacle detouring adaptive criteria:
IO w h e e l = e w h e e l R IO m a m m a l = e m a m m a l h IO i n sec t = e i n sec t H IO h y b r i d = e h y b r i d h
Wherein, IO is obstacle detouring adaptive criteria, represents the contrast between the obstacle height that robot need be crossed and robot obstacle climbing ability; Wherein e is obstacle height, and R, h, H are respectively the maximum height of surmountable obstacle of corresponding sports pattern;
Gradient adaptive criteria:
Wherein, I θ is pitch angle adaptive criteria, and I ψ is roll angle adaptive criteria; Represent the contrast between the attitude angle of robot and extreme attitude angle;
Hardness adaptive criteria:
IS w h e e l = G 6 &times; S w h e e l &times; k e q &times; D max l IS m a m m a l = G n c &times; S f o o t &times; k e q &times; D m a x l IS i n sec t = G n c &times; S f o o t &times; k e q &times; D max l IS h y b r i d = G n c &times; S f o o t &times; k e q &times; D max l
Wherein, IS is hardness adaptive criteria, represents the contrast of footprint pressure when the footprint pressure that robot produces and the maximum amount of being absorbed in; n cfor the number of supporting leg, n cdetermined by the gait selected, 3+3 gait n c=3; 4+2 gait n c=4; 5+1 gait n c=5; S wheeland S footbe respectively single-wheel and monopodia and contact area of ground; G is robot weight, D maxfor the maximum deformation quantity of definition, k eq, l is landform fitting parameter.
5. the multi-locomotion mode intelligent switch method of a kind of multi-functional leg wheel composite machine people according to claim 1, it is characterized in that: the correction of integrated adaptability index, comprise switching cost correction, speed requirement correction and durability requirements correction, according to user's request, the basis of integrated adaptability index adds corresponding switching cost correction, speed requirement correction and/or durability requirements correction, obtains optimal movement pattern.
6. the multi-locomotion mode intelligent switch method of a kind of multi-functional leg wheel composite machine people according to claim 1, it is characterized in that: the switching of mode of motion described in the 6th step, suppose that robot body surrounding evenly arranges six legs, be respectively leg A, leg B, leg C, leg D, leg E and leg F, every bar leg has respectively wheel A, wheel B, wheel C, wheel D, wheel E and wheel F, concrete changing method is as follows:
(1), imitative mammal walking and the mutual changing method of wheel type movement pattern;
Robot is switched to wheel type movement from imitative mammal walking, and robot initial state is imitative mammal walking mode, and concrete switch step is as follows:
(1.1) the downward pitch of leg B, D, F upper thigh servo driving thigh, shank servo driving shank collapse to that shank is parallel to the ground and shank and body distance are K;
(1.2) body height is reduced to taking turns after B, D, F land, upper three the thigh servo driving thighs of leg A, C, E are upwards after pitch to maxim, shank servo driving shank continues to collapse to minimum value, the downward pitch of thigh servo driving thigh and shank servo driving toe raise are to shank is parallel to the ground and shank and body distance are L, L<K;
(1.3) upper three the thigh servo driving thighs of leg B, D, F upwards pitch, shank servo driving shank is contracted to the parallel to the ground and shank of shank and body distance is L, and so far six wheels are completely and earth surface, and robot is in wheel row state;
Robot is switched to imitative mammal walking mode from wheel type movement, and robot initial state is wheel type movement pattern, and concrete switch step is as follows:
(1-1) upper three the thigh servo driving thigh pitch of leg B, D, F, shank servo driving toe raise, it is K>L that body is increased to ground distance;
(1-2) upper three the shank servo driving shanks of leg A, C, E shrink, thigh servo driving thigh extend to each shank perpendicular to the ground after, the downward pitch of thigh servo driving thigh lands rear continuations support to vola;
(1-3) raise body height to H, leg B, D, F move to corresponding foothold, and six legs are parallel, and robot is in imitative mammal walking initial condition;
(2), imitative insect walking and the mutual changing method of wheel type movement;
Robot imitates Methods and steps (1.1) ~ (1.3) imitative mammal walking that insect walking is switched to wheel type movement pattern, and to be switched to the step of wheel type movement consistent;
It is consistent that the step that robot wheel type movement is switched to imitative insect walking mode and step (1-1) ~ (1-3) wheel type movement are switched to the step that imitative mammal walks;
(3), walking mode and the mutual changing method of wheel type movement is mixed;
Robot is switched to wheel type movement pattern from mixing walking mode, and robot initial state is mixing walking mode, and concrete switch step is as follows:
(3.1) upper three the hip servo driving hip yaws of leg B, D, F are to parallel with leg A, the downward pitch of thigh servo driving thigh, and shank servo driving shank collapses to that shank is parallel to the ground and shank is K with body distance;
(3.2) body height is reduced to taking turns after B, D, F land, upper three the shank servo driving shanks of leg A, C, E continue to draw in, after hip servo driving hip yaw is extremely parallel with axis of movement, the downward pitch of thigh servo driving thigh and shank servo driving toe raise are to shank is parallel to the ground and shank and body distance are L, L<K;
(3.3) upper three the thigh servo driving thighs of leg B, D, F upwards pitch, shank servo driving toe raise be L to the parallel to the ground and shank of shank and body distance, and so far six wheels are completely and earth surface, and robot is in taking turns row state;
Robot is switched to mixing walking mode from wheel type movement, and concrete switch step is as follows:
(3-1) upper three the downward pitch of thigh servo driving thigh of leg B, D, F, shank servo driving toe raise, it is K>L that body is increased to ground distance;
(3-2) upper three the shank servo driving shanks of leg A, C, E shrink, thigh servo driving thigh upwards pitch to each shank perpendicular to the ground after, regulate hip servo driving hip yaw, make leg A, C, E be the distributions of 120 ° of rotational symmetry, then the downward pitch of thigh servo driving thigh lands rear continuations support to vola;
(3-3) after body height is increased to H, leg B, D, F move to corresponding foothold, and six legs distribute axisymmetricly, and robot is in mixing walking mode initial condition;
(4), imitative mammal walking with imitate insect and to walk mutual changing method;
Imitative mammal walking with imitate insect and walk there is identical initial attitude;
Switch to imitative insect walking mode from imitative mammal walking mode, suppose that now leg B, D, F is for leading leg, leg A, C, E are supporting leg, and concrete switch step is as follows:
(4.1) leg A, C, E supporting body parallel motion 1/2nd stride is adjusted;
(4.2) adjust leg B, leg D, leg F the foot point that falls to identical with imitative insect initial condition of walk, now robot is in imitative insect walking mode initial attitude, can start imitative insect and walk;
Switch to imitative mammal walking mode from imitative insect walking mode, suppose that now leg B, leg D, leg F are for leading leg, leg A, leg C, leg E are supporting leg, and concrete switch step is as follows:
(4-1) leg A, leg C, leg E supporting body parallel motion 1/2nd stride is adjusted;
(4-2) adjust leg B, leg D, leg F the foot point that falls to identical with imitative mammal initial condition of walk, now robot is in imitative mammal walking mode initial attitude, can start imitative mammal and walk;
(5), imitative mammal walking and mixed mode are walked mutual changing method;
Switch to mixed mode walking mode from imitative mammal walking mode, suppose that leg A, leg C, leg E are supporting leg, leg B, leg D, leg F are for leading leg, and concrete switch step is as follows:
(5.1) adjust lead leg leg B, leg D, upper three the hip servo driving hip yaws of leg F, make leg B, leg D, leg F be 120 ° of rotational symmetry distributions, be converted to supporting leg after landing, body height is H;
(5.2) leg A, leg C, leg E are converted to and lead leg, and adjustment leg A, upper three the hip servo driving hip yaws of leg C, leg E are circumferentially symmetrical to each leg, and now robot is in mixed mode initial attitude, can start mixing walking;
Switch to imitative mammal walking mode from mixed mode walking mode, suppose that leg A, leg C, leg E are supporting leg, leg B, leg D leg F are for leading leg, and concrete switch step is as follows:
(5-1) adjust upper three the hip servo driving hip yaws of leg B, leg D leg F of leading leg, make it parallel with leg, after landing, be converted to supporting leg;
(5-2) leg A, leg C, leg E are converted to and lead leg, and adjustment leg A, upper three the hip servo driving hip yaws of leg C, leg E are parallel with leg A to each leg, and now robot is in imitative mammal walking initial attitude, can start the walking of imitative mammal;
(6), mixed mode walking with imitate insect and to walk mutual changing method, same to step (5.1) ~ (5.2) and step (5-1) ~ (5-2).
CN201410025261.9A 2014-01-20 2014-01-20 A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof CN103786806B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410025261.9A CN103786806B (en) 2014-01-20 2014-01-20 A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410025261.9A CN103786806B (en) 2014-01-20 2014-01-20 A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof

Publications (2)

Publication Number Publication Date
CN103786806A CN103786806A (en) 2014-05-14
CN103786806B true CN103786806B (en) 2016-02-03

Family

ID=50662985

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410025261.9A CN103786806B (en) 2014-01-20 2014-01-20 A kind of multi-functional leg wheel composite machine people and multi-locomotion mode intelligent switch method thereof

Country Status (1)

Country Link
CN (1) CN103786806B (en)

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104002889B (en) * 2014-06-10 2016-03-16 北京航空航天大学 A kind of multi-functional leg wheel composite mobile robot
CN104443105B (en) * 2014-10-29 2017-03-15 西南大学 Low energy consumption Hexapod Robot
CN106393094A (en) * 2015-07-30 2017-02-15 苏州宝时得电动工具有限公司 Automatic moving robot
CN105730546A (en) * 2016-02-02 2016-07-06 上海交通大学 3D printing technology-based minitype bionic six-legged robot
CN105843222B (en) * 2016-03-18 2019-04-12 北京航空航天大学 A kind of six-wheel/leg robot complex motion paths planing method
CN106980321B (en) * 2017-05-18 2019-04-16 中国科学院合肥物质科学研究院 A kind of six foot carries out the top bar gait planning system and method for formula Unmanned Ground Vehicle
CN107553487A (en) * 2017-08-22 2018-01-09 广东美的智能机器人有限公司 The mode switching method and its device of robot
CN107813309A (en) * 2017-11-16 2018-03-20 徐立志 A kind of control method of fighter toy, device and fighter toy
CN108089583B (en) * 2017-12-20 2020-12-25 中科新松有限公司 Method and device for motion transition of multi-legged robot
CN108536011A (en) * 2018-03-19 2018-09-14 中山大学 A kind of Hexapod Robot complicated landform adaptive motion control method based on deeply study
CN110444102A (en) * 2018-05-02 2019-11-12 北京京东尚科信息技术有限公司 Map constructing method, device and unmanned equipment
CN109991672A (en) * 2019-03-22 2019-07-09 华南理工大学 Spherical sensor, monitoring system and monitoring method based on infrared detector
CN110919653A (en) * 2019-11-29 2020-03-27 深圳市优必选科技股份有限公司 Stair climbing control method and device for robot, storage medium and robot

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA1289600C (en) * 1987-02-09 1991-09-24 John Clark Mechanical walking vehicle
JP3870257B2 (en) * 2002-05-02 2007-01-17 独立行政法人 宇宙航空研究開発機構 Robot with offset rotary joint
CN1966337A (en) * 2006-11-23 2007-05-23 北京航空航天大学 Gecko-emulated robot
CN102431604A (en) * 2011-11-09 2012-05-02 上海交通大学 Humanoid robot with function of interchanging double-foot walking with wheeled moving
CN202593677U (en) * 2012-03-31 2012-12-12 浙江工业大学 Six-wheel-foot type series-parallel mixed combination robot
CN102849140A (en) * 2012-09-28 2013-01-02 河北工业大学 Multi-moving-mode bionic moving robot
CN202657137U (en) * 2012-04-11 2013-01-09 中国人民解放军国防科学技术大学 Combined wheel and leg type motion platform

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA1289600C (en) * 1987-02-09 1991-09-24 John Clark Mechanical walking vehicle
JP3870257B2 (en) * 2002-05-02 2007-01-17 独立行政法人 宇宙航空研究開発機構 Robot with offset rotary joint
CN1966337A (en) * 2006-11-23 2007-05-23 北京航空航天大学 Gecko-emulated robot
CN102431604A (en) * 2011-11-09 2012-05-02 上海交通大学 Humanoid robot with function of interchanging double-foot walking with wheeled moving
CN202593677U (en) * 2012-03-31 2012-12-12 浙江工业大学 Six-wheel-foot type series-parallel mixed combination robot
CN202657137U (en) * 2012-04-11 2013-01-09 中国人民解放军国防科学技术大学 Combined wheel and leg type motion platform
CN102849140A (en) * 2012-09-28 2013-01-02 河北工业大学 Multi-moving-mode bionic moving robot

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Dynamics and Wheel’s Slip Ratio of a Wheel-legged Robot in Wheeled Motion Considering the Change of Height;DING Xilun;《CHINESE JOURNAL OF MECHANICAL ENGINEERING》;20121231(第5期);第1061页第3段,图1 *

Also Published As

Publication number Publication date
CN103786806A (en) 2014-05-14

Similar Documents

Publication Publication Date Title
Park et al. Variable-speed quadrupedal bounding using impulse planning: Untethered high-speed 3d running of mit cheetah 2
Xiang et al. Physics-based modeling and simulation of human walking: a review of optimization-based and other approaches
Transeth et al. A survey on snake robot modeling and locomotion
Chase et al. A review of active mechanical driving principles of spherical robots
CN102914967B (en) Autonomous navigation and man-machine coordination picking operating system of picking robot
Li et al. Design and performance evaluation of an amphibious spherical robot
Siegwart et al. Introduction to autonomous mobile robots
Roy et al. Effects of turning gait parameters on energy consumption and stability of a six-legged walking robot
US7865267B2 (en) Environment recognizing device, environment recognizing method, route planning device, route planning method and robot
Heverly et al. Traverse performance characterization for the Mars Science Laboratory rover
Silva et al. Sensitivity of the results produced by the inverse dynamic analysis of a human stride to perturbed input data
CN2695188Y (en) Leg-wheel changed structure type detecting robot
Nansai et al. Dynamic analysis and modeling of Jansen mechanism
Ishigami et al. Path planning and evaluation for planetary rovers based on dynamic mobility index
CN106143750A (en) Relate to the system that traffic is moved
Arechavaleta et al. The nonholonomic nature of human locomotion: a modeling study
Edlund et al. A long-tracked bogie design for forestry machines on soft and rough terrain
Ding et al. Topology and kinematic analysis of color-changing ball
CN103116354B (en) A kind of biped robot&#39;s real-time gait orbit generation method
Kashiri et al. Centauro: A hybrid locomotion and high power resilient manipulation platform
CN101948011A (en) Hexapod universal walking multifunctional moonshot robot
CN102749919B (en) Balance control method of multi-leg robot
CN105667622B (en) It is a kind of to have the six of three sections of bodies to take turns sufficient formula mobile robots
CN103279113B (en) A kind of distributed hydraulic pressure quadruped robot control system and control method
He et al. Preliminary mechanical analysis of an improved amphibious spherical father robot

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant