Beruflich Dokumente
Kultur Dokumente
Abstract
A flexible-rigid hopping mechanism which is inspired by the locust jumping was proposed, and its kinematic characteristics were analyzed. A series of experiments were conducted to observe locust morphology and jumping process. According to
classic mechanics, the jumping process analysis was conducted to build the relationship of the locust jumping parameters. The
take-off phase was divided into four stages in detail. Based on the biological observation and kinematics analysis, a mechanical
model was proposed to simulate locust jumping. The forces of the flexible-rigid hopping mechanism at each stage were analyzed. The kinematic analysis using pseudo-rigid-body model was described by D-H method. It is confirmed that the proposed
bionic mechanism has the similar performance as the locust hind leg in hopping. Moreover, the jumping angle which decides the
jumping process was discussed, and its relation with other parameters was established. A calculation case analysis corroborated
the method. The results of this paper show that the proposed bionic mechanism which is inspired by the locust hind limb has an
excellent kinematics performance, which can provide a foundation for design and motion planning of the hopping robot.
Keywords: hopping robot, flexible-rigid mechanism, bionic mechanism, kinematics
Copyright 2011, Jilin University. Published by Elsevier Limited and Science Press. All rights reserved.
doi: 10.1016/S1672-6529(11)60048-6
1 Introduction
Hopping robot belong to a kind of ground mobile
robot. Such robot must have the ability of adaption to
unstructured environment, especially to avoid dangers
and overleap obstacles agilely and precisely. Many researchers conducted a variety of studies on ground mobile robots with different forms of movement. At present,
the terrain adaptability of ground mobile robot has already been largely improved, but the obstacle negotiating performance is still unsatisfactory due to the limitation of the locomotion mode. Biologic hopping locomotion can achieve a certain height and distance in a
moment, which enables animals to overcome obstacles
and uneven terrains. Therefore, the study on hopping
robot, which developed from nature observation, has
received much attention in recent years. The exploration
of hopping movement will expand the application fields
of mobile robot.
Raibert and Pratt developed one-legged, biped, and
quadruped running and hopping robots based on
Raiberts pneumatic model, all of which have telescopCorresponding author: Diansheng Chen
E-mail: chends@buaa.edu.cn
430
Locust
15
23
30
300
Distant (cm)
30
50
70
900
200
12
22
1.2
3.2
Kangaroo
Flea
Parameters
Height (cm)
Locust 2
Mass (g)
0.22
0.21
0.23
33.25
31.29
34.05
Femur
Tibiae
Tarsus
Femur
13.27
11.97
3.10
13.03
5.10
4.61
2.34
4.55
3.98
1.80
Locust 3
Tarsus
Femur
Tibiae
Tarsus
3.64
13.25
11.99
3.55
5.17
Tibiae
12.1
6
4.80
2.91
5.19
4.82
2.85
4.45
3.56
2.49
4.65
4.21
2.41
Chen et al.: Bionic Mechanism and Kinematics Analysis of Hopping Robot Inspired by Locust Jumping
431
200
150
100
50
0
50
100
150
200
x(t) (mm)
250
300
350
400
y (mm)
200
180
160
140
120
100
80
60
40
20
0
50
100
150
200 250
x (mm)
300
350
400
450
1 2
y v0 t sin D 2 gt ,
(1)
ymax
(v02 sin 2D ) g ,
(v0 sin D )2 2 g .
(2)
432
Eq. (2) shows that the takeoff speed and angle affect the jump height and distance. With a certain speed,
it can be proved that the jumping achieves xmax when the
takeoff angle is about 45. As the function of projectile
trajectory is given by
y
x tan D
350
(3)
250
y = y(t)
200
150
100
50
This derivation ignores the air resistance, so it usually applies to the case with a relatively low speed.
However, the takeoff speed of locust may be very high
that it is uncertain whether the locust case adapts the
equations. Therefore, to get a precise result, the derived
curve F(x) = y(x) and second derived curve G(x) = y(x)
of fitting trajectory curve are given in the Fig. 6. It shows
the case that the takeoff angle is 45 and the jumping
process corresponds to the law of particle movement.
Therefore, based on Eq. (3), the conservation of energy
is given by Eq. (4) and energy loss Es reveals the error in
Fig. 6.
(a)
3.0
F(x), G(x)
150
t (ms)
200
250
sqrt(Vx2 , Vy2 )
2.0
Vx = x(t)
1.5
1.0
Vy = y(t)
0.5
(b)
0.0
0.005
Fig. 6 The derived curve F(x) and second derived curve G(x) of
trajectory curve.
100
2.5
(4)
50
3.5
50
100
150
t (ms)
200
250
300
The acceleration in
the x direction
0.000
Acceleration ( mmms2)
1 2
1
mv0 3 mgh ! mv 2 ! Es ,
2
2
x = x(t)
300
s ( mm )
1 gx 2
.
2 v02 cos 2 D
400
0.005
0.010
The acceleration
in the y direction
0.015
0.020
0.025
0.030
The sum acceleration
in the x and y direction
0.035
0.040
50
100
150
t (ms)
(c)
200
250
300
Chen et al.: Bionic Mechanism and Kinematics Analysis of Hopping Robot Inspired by Locust Jumping
433
Body-horizontal
Tarsus-horizontal
Tibiae-tarsus
Femur-Tibiae
Body-femur
0 ms
10.47
0.00
29.39
5.91
146.05
4 ms
10.32
0.00
27.09
3.95
146.55
8 ms
11.10
0.00
26.54
4.16
146.82
12 ms
11.24
0.00
25.89
4.01
146.87
16 ms
11.38
0.00
24.46
3.72
145.88
20 ms
11.37
0.00
26.52
3.87
145.96
24 ms
11.60
0.00
26.45
4.03
145.96
28 ms
12.70
0.00
28.94
5.18
143.53
32 ms
12.70
0.00
40.27
7.09
134.11
36 ms
8.71
0.00
52.22
24.60
143.67
40 ms
3.18
0.00
73.55
71.68
178.06
44 ms
12.44
40.49
147.94
142.63
132.38
48 ms
22.75
46.12
142.75
135.63
118.24
52 ms
11.03
56.03
170.77
143.07
140.64
434
bionic mechanism. Fig. 10 shows the model of the hopping robot considered in this paper. Considering that the
semi-lunar process of knee joint is a flexible structure, it
is replaced with a linear spring to store energy. Moreover a flexible link takes the place of the tarsus.
The analysis of kinematic properties will be described in section 4.
Fe
Fs
OL
Ff
lEF
lDE
D
D
Extensor
muscle
Tendon
Femur
Joint
Extensor
muscle for
pretarsus
Trochanter
lCD
x
o
Tibiae
Coxa
Fg
Pretarsus
A
Tarsus
Fm
Ffr
Fs
Ff
M1
M0
Fg
Ffr
(5)
Fe cos(G M F )
Ff cos(M E G T ) Fg cos(M E G G D ), (6)
where Fs denote the spring force. The torque balance
equation is given by
Fe lEF sin M F Ffr (lDE cos(M E G )
lCD cos(M E G G D )) Ff lDE sin T
M E Z c, (7)
1
Fe lEF sin M F Ffr (lDE cos(M E G )
lDE sin T
lCD cos(M E G G D )) M E Z c .
(8)
Chen et al.: Bionic Mechanism and Kinematics Analysis of Hopping Robot Inspired by Locust Jumping
Ff
(9)
'LGx
't
c
LGx
435
M C0
M C1
'H Gy
'LGx
i
j,
't
't
(11)
'H Gy
lCE M1 sin(
M1
M0 )
2
T T
lEG (T1 T 0 ) sin(G 0 1 0 ),
2
lCE M1 cos(
M1
2
M0 )
T1 T 0
2
),
(13)
where lCE, lEG denote the length of CE, EG correspondingly. When t is small enough, the difference equations
of Eqs. (12) and (13) can be derived as
M
1
M0 ) lCE M1 M1c cos( 1 M0 )
2
2
2
T1 T 0 1
T T
) lEG (T1 T 0 ) (T1 T 0 )c cos(G 0 1 0 ),
lEG (T1 T 0 )c sin(G 0
2
2
2
(12)
M1
(14)
436
'H Gy
c
H Gy
't
M
1
M0 ) lCE M1 M1c sin( 1 M0 )
2
2
2
T1 T 0 1
T T
) lEG (T1 T 0 ) (T1 T 0 )c sin(G 0 1 0 ).
lEG (T1 T 0 )c cos(G 0
2
2
2
M1
DG
actan
c
H Gy
actan
c
LGx
'H Gy
'LGx
k1 k2 sin(G 0
T1 T 0
)
2
actan
,
T T
k1 k2 cos(G 0 1 0 )
2
lCE M1 sin(
M1
M0 ), k2 lEG (T1 T 0 ).
2
By Eq. (10), the center of mass off angle is
where k1
W i 3 actan
vOy
vOx
Fa 2
(3l a ), (0 d a d l ),
6 EI
4 Kinematic analysis
(i 3 0,1, 2,...).
0
1
cos(T1 (t )) sin(T1 (t ))
sin(T (t )) cos(T (t ))
1
1
0
0
0
0
0
0
1
0
1
2
cos(T 2 (t )) sin(T 2 (t ))
sin(T (t )) cos(T (t ))
2
2
0
0
0
0
0 lo12
0
0
,
1
0
0
1
x (t )
y (t )
,
0
1
(16)
where E and I are the elastic modulus and inertia moment; F, a, l denote force, the force position and length
of beam, respectively.
(a)
(b)
(c)
(d)
(15)
Chen et al.: Bionic Mechanism and Kinematics Analysis of Hopping Robot Inspired by Locust Jumping
2
3
cos( T3 (t )) sin( T3 (t ))
sin( T (t )) cos( T (t ))
3
3
0
0
0
0
0 lo 23
0 0
,
1 0
0 1
3
4
cos( T 4 (t )) sin( T 4 (t ))
sin( T (t )) cos( T (t ))
4
4
0
0
0
0
0 lo 34
0 0
,
1 0
0 1
4
5
0
i
0
1
1
0
0 0 lo 45
1 0 0
,
0 1 0
0 0 1
nx ox
n o
y y
0 0
0 0
T
=
3
0
0
0
0
0
4
5T =
0
0 0
0 lo 23 k 'x
0
0
, (18)
1
0
0
1
Fa
(3l a )
.
6 EI
1
lo 45
1 0
0 1
0 0
0 pxi
0 p yi
(i 1, 2,...,5). (17)
1 pzi
0 1
(19)
It is easier to ascertain 01T according to the experimental result in section 2. 01T indicates the jumping
performance such as height, horizontal distance and
body posture. The relationship between the rotation
angle of joint and time series has been measured
437
J l1 J l 2
J
a1 J a 2
T1
" J li T2
(i 1, 2,...,5),
" J ai #
Ti
zi u 0i R i pn
.
zi
(20)
x (q(24))
35.5000
31.6765
19.4318
30.1404
27.0404
2.8000 0
2.0311 0
7.1457 0 ,
1.7972 0
1.7983 0
438
35.5000
31.6765
19.4332
30.1418
27.0418
x c(q (24))
2.8000 0
2.0311 0
7.1451 0 .
1.7967 0
1.7884 0
In consideration of flexibility, the posture with respect to time is depicted in Fig. 15. The flexible-rigid
mechanism has n = 5 joints, and the i pn0 3 0i R i pn of joint
i (i 3 1, 2,..., n) is calculated in 24 ms as below
p = 1 p50
= 10 R 1 p5
p50
0
2
p50
p50
R 2 p5
0
3
R 3 p5
p50
0
4
R 4 p5
0
5
R 5 p5
0
0
0
0
Then, according to Eq. (20), at this time, the velocity Jacobin matrix can be expressed by
1.7803
9.7422 10.3340 2.2907 1.7877
34.2523 28.4451 20.5774 30.1423 27.0423
0
0
0
0
.
0
0
0
0
0
0
0
0
0
1
1
1
1
1
Acknowledgement
This work is financially supported by the National
Natural Science Foundation of China (Grant No.
51075014).
References
[1]
http://www.ai.mit.edu/projects/leglab/robots/robots.html
[2]
t = 52 ms
y (mm)
[5]
15
t = 40 ms
[6]
[7]
t = 36 ms
t = 0, 4, , 32 ms
20
30
40
x (mm)
50
60
t = 44 ms
10
t = 48 ms
20
10
25
30
[3]
Centre of mass
Joint point
Body and limbs
Abdomen
35
70
80
5 Conclusions
By means of experiment, this paper studied the
jumping motion of locust and the effect of takeoff pos-
[9]
Chen et al.: Bionic Mechanism and Kinematics Analysis of Hopping Robot Inspired by Locust Jumping
cricket, Gryllus bimaculatus. Journal of Experimental Biology, 2003, 206, 523534.
439
Switzerland, 2010.