-
Notifications
You must be signed in to change notification settings - Fork 0
/
paper_gu1_adapt.py
252 lines (207 loc) · 9.08 KB
/
paper_gu1_adapt.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
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
import gurobipy as gp
import matplotlib.pyplot as plt
import numpy as np
from gurobipy import GRB
# Parameters
T = 20 # Prediction horizon (seconds)
dt = 0.1 # Time step (seconds)
N = int(T / dt) # Number of time steps
# State and control bounds
x_min, x_max = 0, np.inf
y_min, y_max = 0, 2.
v_x_min, v_x_max = 0, 20
v_y_min, v_y_max = -1, 1
a_x_min, a_x_max = -2, 2
a_y_min, a_y_max = -0.5, 0.5
j_x_min, j_x_max = -2.5, 2.5
j_y_min, j_y_max = -1, 1
# Other parameters from Table 1
theta_min, theta_max = -0.4, 0.4 # in radians
omega_min, omega_max = -0.26, 0.26 # in radians/s
# Speed bump parameters
x_bump_start, x_bump_end = 30, 35
v_max_bump = 5
# Initial state and reference values
x0, y0 = 0, 0.75
v_x0, v_y0 = 10, 0
a_x0, a_y0 = 0, 0
v_r = 10.0 # Reference speed
y_r = y0 # Reference lateral position (center of the lane)
# wheel base
L = 2.7
# Cost function weights (from Table 1)
q1, q2, q3, q4, q5 = 1, 1, 1, 2, 4
r1, r2 = 40, 40
# Create the model
model = gp.Model("SpeedBump_MIQP")
# Create variables
x = model.addVars(N+1, lb=x_min, ub=x_max, name="x")
y = model.addVars(N+1, lb=y_min, ub=y_max, name="y")
v_x = model.addVars(N+1, lb=v_x_min, ub=v_x_max, name="v_x")
v_y = model.addVars(N+1, lb=v_y_min, ub=v_y_max, name="v_y")
a_x = model.addVars(N+1, lb=a_x_min, ub=a_x_max, name="a_x")
a_y = model.addVars(N+1, lb=a_y_min, ub=a_y_max, name="a_y")
j_x = model.addVars(N, lb=j_x_min, ub=j_x_max, name="j_x")
j_y = model.addVars(N, lb=j_y_min, ub=j_y_max, name="j_y")
# steering angle
steer_max = np.pi/6
delta = model.addVars(N+1, lb=-steer_max, ub=steer_max, name="delta")
epsilon = 1e-6 # Small value for numerical stability
# Binary variables for speed bump logical constraints
delta1 = model.addVars(N+1, vtype=GRB.BINARY, name="delta1")
delta2 = model.addVars(N+1, vtype=GRB.BINARY, name="delta2")
delta3 = model.addVars(N+1, vtype=GRB.BINARY, name="delta3")
# binary variables for steer to left or right
delta4 = model.addVars(N+1, vtype=GRB.BINARY, name="delta4")
# Set initial conditions
model.addConstr(x[0] == x0)
model.addConstr(y[0] == y0)
model.addConstr(v_x[0] == v_x0)
model.addConstr(v_y[0] == v_y0)
model.addConstr(a_x[0] == a_x0)
model.addConstr(a_y[0] == a_y0)
# Add dynamics constraints
for k in range(N):
model.addConstr(x[k+1] == x[k] + v_x[k]*dt + 0.5*a_x[k]*dt**2 + (1/6)*j_x[k]*dt**3)
model.addConstr(y[k+1] == y[k] + v_y[k]*dt + 0.5*a_y[k]*dt**2 + (1/6)*j_y[k]*dt**3)
model.addConstr(v_x[k+1] == v_x[k] + a_x[k]*dt + 0.5*j_x[k]*dt**2)
model.addConstr(v_y[k+1] == v_y[k] + a_y[k]*dt + 0.5*j_y[k]*dt**2)
model.addConstr(a_x[k+1] == a_x[k] + j_x[k]*dt)
model.addConstr(a_y[k+1] == a_y[k] + j_y[k]*dt)
# Add constraints for theta and omega (equations 5 and 6 in the paper)
for k in range(N+1):
model.addConstr(v_y[k] >= v_x[k] * np.tan(theta_min))
model.addConstr(v_y[k] <= v_x[k] * np.tan(theta_max))
model.addConstr(a_y[k] >= -v_x[k] * omega_max)
model.addConstr(a_y[k] <= v_x[k] * omega_max)
# steering angle constraints steer = tan^-1(L * omega / v_x)--> steer = L vy/vx --> L * vy = vx * steer
model.addConstr(v_y[k] * L == v_x[k] * delta[k])
# Speed bump logical constraints using indicator constraints
for k in range(N+1):
# δ1(k) = 1 ⇔ x(k) ≥ x_bump_start
model.addGenConstrIndicator(delta1[k], True, x[k] >= x_bump_start)
model.addGenConstrIndicator(delta1[k], False, x[k] <= x_bump_start + epsilon)
# δ2(k) = 1 ⇔ x(k) ≤ x_bump_end
model.addGenConstrIndicator(delta2[k], True, x[k] <= x_bump_end + epsilon)
model.addGenConstrIndicator(delta2[k], False, x[k] >= x_bump_end)
# δ3(k) = 1 ⇔ v_x(k) ≤ v_max_bump
model.addGenConstrIndicator(delta3[k], True, v_x[k] <= v_max_bump)
model.addGenConstrIndicator(delta3[k], False, v_x[k] >= v_max_bump - epsilon)
# δ4(k) = 1 ⇔ a_y(k) >= a_steer or a_y(k) <= -a_steer
# a_steer_y = 0.1
#model.addGenConstrIndicator(delta4[k], True, a_y[k] >= a_steer_y)
bump_steer = np.pi/90 # 5 degrees steering during bump
model.addGenConstrIndicator(delta4[k], True, delta[k] >= bump_steer)
# Logical implications from equation 7
model.addConstr(-delta1[k] + delta3[k] <= 0)
model.addConstr(-delta2[k] + delta3[k] <= 0)
model.addConstr(delta1[k] + delta2[k] - delta3[k] <= 1)
# Logical implications from steering when in bump
model.addConstr(delta1[k] + delta2[k] - delta4[k] <= 1)
# Objective function
obj = gp.QuadExpr()
for k in range(N+1):
#obj += q1 * (v_x[k] - v_r)**2 + q2 * a_x[k]**2 + q3 * (y[k] - y_r)**2 + q4 * v_y[k]**2 + q5 * a_y[k]**2
obj += q1 * (v_x[k] - v_r)**2 + q3 * (y[k] - y_r)**2 + q4 * v_y[k]**2 + q5 * a_y[k]**2
for k in range(N):
obj += r1 * j_x[k]**2 + r2 * j_y[k]**2
model.setObjective(obj, GRB.MINIMIZE)
# Optimize the model
model.optimize()
# Extract results
x_res = [x[k].X for k in range(N+1)]
y_res = [y[k].X for k in range(N+1)]
v_x_res = [v_x[k].X for k in range(N+1)]
v_y_res = [v_y[k].X for k in range(N+1)]
a_x_res = [a_x[k].X for k in range(N+1)]
a_y_res = [a_y[k].X for k in range(N+1)]
j_x_res = [j_x[k].X for k in range(N)]
j_y_res = [j_y[k].X for k in range(N)]
# Extract steering angles
delta_res = [delta[k].X for k in range(N+1)]
# compute sum of squared jerk
jerk_sum = sum(j_x_res[k]**2 + j_y_res[k]**2 for k in range(N))
print("Sum of squared jerk:", jerk_sum)
# Create the plots
fig, axs = plt.subplots(3, 2, figsize=(15, 10))
# Longitudinal speed plot
axs[0, 0].plot(x_res, v_x_res, 'b-', linewidth=2, label='v_x')
axs[0, 0].set_xlabel('x (m)')
axs[0, 0].set_ylabel('v_x (m/s)')
axs[0, 0].set_title('Longitudinal Speed Profile')
axs[0, 0].axvline(x=x_bump_start, color='r', linestyle='--', label='Speed bump start')
axs[0, 0].axvline(x=x_bump_end, color='r', linestyle='--', label='Speed bump end')
axs[0, 0].axhline(y=v_max_bump, color='g', linestyle=':', label='Max speed in bump')
axs[0, 0].fill_between([x_bump_start, x_bump_end], 0, v_x_max, alpha=0.2, color='r')
axs[0, 0].set_xlim(0, max(x_res))
axs[0, 0].set_ylim(0, v_x_max)
axs[0, 0].legend()
axs[0, 0].grid(True, linestyle=':', alpha=0.7)
# Lateral speed plot
axs[0, 1].plot(x_res, v_y_res, 'b-', linewidth=2, label='v_y')
axs[0, 1].set_xlabel('x (m)')
axs[0, 1].set_ylabel('v_y (m/s)')
axs[0, 1].set_title('Lateral Speed Profile')
axs[0, 1].axvline(x=x_bump_start, color='r', linestyle='--', label='Speed bump start')
axs[0, 1].axvline(x=x_bump_end, color='r', linestyle='--', label='Speed bump end')
axs[0, 1].set_xlim(0, max(x_res))
axs[0, 1].set_ylim(v_y_min, v_y_max)
axs[0, 1].legend()
axs[0, 1].grid(True, linestyle=':', alpha=0.7)
# Longitudinal acceleration plot
axs[1, 0].plot(x_res, a_x_res, 'b-', linewidth=2, label='a_x')
axs[1, 0].set_xlabel('x (m)')
axs[1, 0].set_ylabel('a_x (m/s²)')
axs[1, 0].set_title('Longitudinal Acceleration Profile')
axs[1, 0].axvline(x=x_bump_start, color='r', linestyle='--', label='Speed bump start')
axs[1, 0].axvline(x=x_bump_end, color='r', linestyle='--', label='Speed bump end')
axs[1, 0].set_xlim(0, max(x_res))
axs[1, 0].set_ylim(a_x_min, a_x_max)
axs[1, 0].legend()
axs[1, 0].grid(True, linestyle=':', alpha=0.7)
# Lateral acceleration plot
axs[1, 1].plot(x_res, a_y_res, 'b-', linewidth=2, label='a_y')
axs[1, 1].set_xlabel('x (m)')
axs[1, 1].set_ylabel('a_y (m/s²)')
axs[1, 1].set_title('Lateral Acceleration Profile')
axs[1, 1].axvline(x=x_bump_start, color='r', linestyle='--', label='Speed bump start')
axs[1, 1].axvline(x=x_bump_end, color='r', linestyle='--', label='Speed bump end')
axs[1, 1].set_xlim(0, max(x_res))
axs[1, 1].set_ylim(a_y_min, a_y_max)
axs[1, 1].legend()
axs[1, 1].grid(True, linestyle=':', alpha=0.7)
# Vehicle Trajectory plot with heading angles
axs[2, 0].plot(x_res, y_res, 'b-', linewidth=2, label='Vehicle Path')
axs[2, 0].set_xlabel('x (m)')
axs[2, 0].set_ylabel('y (m)')
axs[2, 0].set_title('Vehicle Trajectory with Heading Angles')
axs[2, 0].axvline(x=x_bump_start, color='r', linestyle='--', label='Speed bump start')
axs[2, 0].axvline(x=x_bump_end, color='r', linestyle='--', label='Speed bump end')
axs[2, 0].set_xlim(0, max(x_res))
axs[2, 0].set_ylim(y_min, y_max)
# Plot arrows to indicate heading angle
arrow_spacing = max(1, N // 20) # Adjust this value to change arrow density
for i in range(0, N+1, arrow_spacing):
axs[2, 0].arrow(x_res[i], y_res[i],
np.cos(delta_res[i]), np.sin(delta_res[i]),
head_width=0.1, head_length=0.5, fc='r', ec='r')
axs[2, 0].legend()
axs[2, 0].grid(True, linestyle=':', alpha=0.7)
# Steering angle plot
axs[2, 1].plot(x_res, delta_res, 'b-', linewidth=2, label='Steering Angle')
axs[2, 1].set_xlabel('x (m)')
axs[2, 1].set_ylabel('delta (rad)')
axs[2, 1].set_title('Steering Angle Profile')
axs[2, 1].axvline(x=x_bump_start, color='r', linestyle='--', label='Speed bump start')
axs[2, 1].axvline(x=x_bump_end, color='r', linestyle='--', label='Speed bump end')
axs[2, 1].set_xlim(0, max(x_res))
axs[2, 1].set_ylim(-steer_max, steer_max)
axs[2, 1].legend()
axs[2, 1].grid(True, linestyle=':', alpha=0.7)
# Adjust layout for better visibility
plt.tight_layout()
plt.show()
print("Optimization status:", model.status)
print("Objective value:", model.objVal)
print("Optimization status:", model.status)
print("Objective value:", model.objVal)