-
Notifications
You must be signed in to change notification settings - Fork 83
/
cassietest_jac.py
135 lines (118 loc) · 4.95 KB
/
cassietest_jac.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
#!/usr/bin/env python3
# Copyright (c) 2018 Dynamic Robotics Laboratory
#
# Permission to use, copy, modify, and distribute this software for any
# purpose with or without fee is hereby granted, provided that the above
# copyright notice and this permission notice appear in all copies.
#
# THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
# WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
# MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
# ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
# WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
# ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
# OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
from cassiemujoco import *
from cassiemujoco_ctypes import joint_filter_t, drive_filter_t
import time
import numpy as np
import math
import matplotlib.pyplot as plt
"""
Test the end-effector Jacobian, don't care
1. dynamically consistent (Mx*J*Mx^-1);
2. closed-loop constraints;
3. rotational components.
Only to get the rough estimate of the Jacobian matrix on motors.
"""
# Initialize cassie simulation
sim = CassieSim("../model/cassie.xml")
vis = CassieVis(sim)
# Set control parameters
u = pd_in_t()
test_jac_control = True
test_random = True
if test_jac_control:
qpos = sim.qpos()
qpos[2] = 1.5
sim.set_qpos(qpos)
sim.hold()
# Record time
t = time.monotonic()
count = 0
ltarget = np.array([0, 0.13, -0.8])
rtarget = np.array([0, -0.13, -0.5])
kp = np.array([70, 70, 100, 100, 50])
kd = np.array([7.0, 7.0, 8.0, 8.0, 5.0])
# Run until window is closed or vis is quit
draw_state = vis.draw(sim)
vel_idx = [6, 7, 8, 12, 18, 19, 20, 21, 25, 31]
pos_idx = [7, 8, 9, 14, 20, 21, 22, 23, 28, 34]
ts_noise_up = np.array([[0, 0, 0.13, 0.13, 0],
[0.13, 0, 0.00, 0.00, 0],
[0, 0, 0.13, 0.13, 0]])
ts_noise = np.block([
[ts_noise_up, np.zeros((3,5))],
[np.zeros((3,5)), ts_noise_up]
])
offset = np.array([0.0045, 0.0, 0.4973, -1.1997, -1.5968, 0.0045, 0.0, 0.4973, -1.1997, -1.5968])
while draw_state:
if not vis.ispaused():
for i in range(60):
jacpl = sim.get_jacobian(name='left-foot').reshape(3, -1)
jacpr = sim.get_jacobian(name='right-foot').reshape(3, -1)
jacp = np.concatenate((jacpl, jacpr))
jacp_motor = jacp.take(vel_idx, axis=1)
jdag = np.linalg.pinv(jacp_motor)
if test_jac_control:
lpos = np.array(sim.foot_pos()[0:3]) - np.array(sim.qpos()[0:3])
rpos = np.array(sim.foot_pos()[3:6]) - np.array(sim.qpos()[0:3])
dxl = ltarget - lpos
dxr = rtarget - rpos
# print(np.dot(jdag[:,0:3], dxl).shape)
# print(np.dot(jdag[:,3:6], dxr).shape)
dq = np.dot(jdag[:,0:3], dxl) + np.dot(jdag[:,3:6], dxr)
# print(dq.shape)
# print(dq)
# print(lpos)
qpos = sim.qpos()
mpos = [qpos[i] for i in pos_idx]
for i in range(5):
u.leftLeg.motorPd.pGain[i] = kp[i] * 0.1
u.rightLeg.motorPd.pGain[i] = kp[i] * 0.1
u.leftLeg.motorPd.dGain[i] = kd[i] * 0.1
u.rightLeg.motorPd.dGain[i] = kd[i] * 0.1
u.leftLeg.motorPd.torque[i] = 0 # Feedforward torque
u.rightLeg.motorPd.torque[i] = 0
u.leftLeg.motorPd.pTarget[i] = dq[i] + mpos[i]
u.rightLeg.motorPd.pTarget[i] = dq[i+5] + mpos[i+5]
u.leftLeg.motorPd.dTarget[i] = 0
u.rightLeg.motorPd.dTarget[i] = 0
y = sim.step_pd(u)
else:
action = np.random.uniform(-10, 10, size=10)
for i in range(5):
u.leftLeg.motorPd.pGain[i] = kp[i]
u.rightLeg.motorPd.pGain[i] = kp[i]
u.leftLeg.motorPd.dGain[i] = kd[i]
u.rightLeg.motorPd.dGain[i] = kd[i]
u.leftLeg.motorPd.torque[i] = 0 # Feedforward torque
u.rightLeg.motorPd.torque[i] = 0
u.leftLeg.motorPd.pTarget[i] = action[i] + offset[i]
u.rightLeg.motorPd.pTarget[i] = action[i+5] + offset[i+5]
u.leftLeg.motorPd.dTarget[i] = 0
u.rightLeg.motorPd.dTarget[i] = 0
y = sim.step_pd(u)
sd_final = np.matmul(jdag, ts_noise)
# print(sd_final.shape)
# print("new js noise matrix")
# for i in range(10):
# for j in range(10):
# print("{: 3.2f}".format(sd_final[i][j]), end=" ")
# print("\n")
# input()
draw_state = vis.draw(sim)
count += 1
while time.monotonic() - t < 60*0.0005:
time.sleep(0.0001)
t = time.monotonic()