-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathrm_tp.py
199 lines (163 loc) · 6.14 KB
/
rm_tp.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
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
import numpy as np
from copy import deepcopy
from scipy.io import loadmat
from reachability.tracker import Tracker, Planner, Relative
def normalize_angles(angles):
'''Puts angles in [-pi, pi] range.'''
angles = angles.copy()
if angles.size > 0:
angles = (angles + np.pi) % (2 * np.pi) - np.pi
assert -(np.pi + 1e-6) <= angles.min() and angles.max() <= (np.pi + 1e-6)
return angles
class RmTracker(Tracker):
def __init__(self):
self.size = 5
self.dt = 0.01
def dynamics(self, x, u):
new_x = deepcopy(x)
new_x[0] += (x[1]*np.cos(x[4]) - x[3]*np.sin(x[4])) * self.dt
new_x[1] += u[0] * self.dt
new_x[2] += (x[1]*np.sin(x[4]) + x[3]*np.cos(x[4])) * self.dt
new_x[3] += u[1] * self.dt
new_x[4] += u[2] * self.dt
#new_x[4] = normalize_angles(np.array([new_x[4]]))[0]
return new_x
class RmPlanner(Planner):
def __init__(self):
self.size = 3
def dynamics(self, x, u):
new_x = deepcopy(x)
new_x[0] += u[0]*np.cos(x[2]) - u[1]*np.sin(x[2])
new_x[1] += u[0]*np.sin(x[2]) + u[1]*np.cos(x[2])
new_x[2] += u[2]
return new_x
def project(self, s):
p = [s[0], s[2], s[4]]
return p
def control(self, p):
x = 5
y = 5
theta = 0.7 #p[2] + 4.16 * 0.01
#theta = normalize_angles(np.array([theta]))[0]
return [x, y, theta]
class RmRelative(Relative):
def __init__(self):
self.size = 5
self.uMax = np.array([10.0, 10.0, 6.0, 2.0, 3.0, 4.16])
self.uMin = -self.uMax
def clip_state(self, r):
r[0] = np.clip(r[0], -5, 5)
r[1] = np.clip(r[1], -2, 2)
r[2] = np.clip(r[2], -5, 5)
r[3] = np.clip(r[3], -3, 3)
r[4] = np.clip(r[4], -np.pi, np.pi)
return r
def state(self, s, p):
r = deepcopy(s)
r[0] -= p[0]
r[2] -= p[1]
r[4] -= p[2]
r = self.clip_state(r)
return r
def dynamics(self, r, u, d):
rnext = deepcopy(r)
rnext = np.clip(rnext, [-5, -2, -5, -3, -np.pi], [5, 2, 5, 3, np.pi])
rnext[0] += x[1] - u[3]*np.cos(x[4]) + u[4]*np.sin(x[4]) + d[0]
rnext[1] += u[0]
rnext[2] += x[3] - u[3]*np.sin(x[4]) - u[4]*np.cos(x[4]) + d[1]
rnext[3] += u[1]
rnext[4] += u[2] - u[5] + d[2]
return rnext
def optControl(self, deriv, r, x):
ax = (deriv[1]>=0)*self.uMin[0] + (deriv[1]<0)*self.uMax[0]
ay = (deriv[3]>=0)*self.uMin[1] + (deriv[3]<0)*self.uMax[1]
w = (deriv[4]>=0)*self.uMin[2] + (deriv[4]<0)*self.uMax[2]
# Calculate opt ctrl
uopt = [
ax, ay, w
]
return uopt
class Reach():
def __init__(self):
matlabf = "./reachability/RMAI_g_dt01_t5_medium_quadratic.mat"
fst = loadmat(matlabf)
eps = 0.1
self.eb = np.max([fst['TEB_X1'], fst['TEB_X2'], fst['TEB_X3'], fst['TEB_X4']]) + eps
self.vf_X1 = np.array(fst['data0_X1'])
self.vf_X2 = np.array(fst['data0_X2'])
self.vf_X3 = np.array(fst['data0_X3'])
self.vf_X4 = np.array(fst['data0_X4'])
self.vf_dX1 = fst['derivX1']
self.vf_dX2 = fst['derivX2']
self.vf_dX3 = fst['derivX3']
self.vf_dX4 = fst['derivX4']
def to_grid_index(self, r):
r = deepcopy(r)
r[0] = int(((r[0] + 5) / 10.0) * 60)
r[1] = int(((r[1] + 2) / 4.0) * 60)
r[2] = int(((r[2] + 5) / 10.0) * 60)
r[3] = int(((r[3] + 3) / 6.0) * 60)
r[4] = int(((r[4] + np.pi) / (2*np.pi)) * 60)
return r
def check_on_boundary(self, r):
r_int = [int(i) for i in r]
vf_X1_eb = self.vf_X1[r_int[0], r_int[1], r_int[4]]
vf_X2_eb = self.vf_X2[r_int[0], r_int[3], r_int[4]]
vf_X3_eb = self.vf_X3[r_int[1], r_int[2], r_int[4]]
vf_X4_eb = self.vf_X4[r_int[2], r_int[3], r_int[4]]
vf_eb = np.max([vf_X1_eb, vf_X2_eb, vf_X3_eb, vf_X4_eb])
if vf_eb >= self.eb:
return True
else:
return False
def control(self, s, pnext):
dx = pnext[0] - s[0]
dy = pnext[1] - s[2]
dw = pnext[2] - s[4]
ax = 5.0 * dx
ay = 5.0 * dy
return [ax, ay, dw]
def get_derivs(self, r):
r_int = [int(i) for i in r]
# All x deriv
if self.vf_X1[r_int[0], r_int[1], r_int[4]] >= self.vf_X2[r_int[0], r_int[3], r_int[4]]:
x_deriv = self.vf_dX1[0][0][r_int[0], r_int[1], r_int[4]]
else:
x_deriv = self.vf_dX2[0][0][r_int[0], r_int[3], r_int[4]]
# All vx deriv
if self.vf_X1[r_int[0], r_int[1], r_int[4]] >= self.vf_X3[r_int[1], r_int[2], r_int[4]]:
vx_deriv = self.vf_dX1[1][0][r_int[0], r_int[1], r_int[4]]
else:
vx_deriv = self.vf_dX3[1][0][r_int[1], r_int[2], r_int[4]]
# All y deriv
if self.vf_X3[r_int[1], r_int[2], r_int[4]] >= self.vf_X4[r_int[2], r_int[3], r_int[4]]:
y_deriv = self.vf_dX3[0][0][r_int[1], r_int[2], r_int[4]]
else:
y_deriv = self.vf_dX4[0][0][r_int[2], r_int[3], r_int[4]]
# All vy deriv
if self.vf_X2[r_int[0], r_int[3], r_int[4]] >= self.vf_X4[r_int[2], r_int[3], r_int[4]]:
vy_deriv = self.vf_dX2[1][0][r_int[0], r_int[3], r_int[4]]
else:
vy_deriv = self.vf_dX4[1][0][r_int[2], r_int[3], r_int[4]]
# All theta deriv
theta_idx = np.argmax([
self.vf_X1[r_int[0], r_int[1], r_int[4]],
self.vf_X2[r_int[0], r_int[3], r_int[4]],
self.vf_X3[r_int[1], r_int[2], r_int[4]],
self.vf_X4[r_int[2], r_int[3], r_int[4]]
])
theta_deriv = [
self.vf_dX1[2][0][r_int[0], r_int[1], r_int[4]],
self.vf_dX2[2][0][r_int[0], r_int[3], r_int[4]],
self.vf_dX3[2][0][r_int[1], r_int[2], r_int[4]],
self.vf_dX4[2][0][r_int[2], r_int[3], r_int[4]]
]
theta_deriv = theta_deriv[theta_idx]
deriv = [
x_deriv,
vx_deriv,
y_deriv,
vy_deriv,
theta_deriv
]
return deriv