-
Notifications
You must be signed in to change notification settings - Fork 1
/
Main.py
179 lines (149 loc) · 6.88 KB
/
Main.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
'''
Written By: Abhijeet Kulkarni ([email protected])
For: Research Project MEEG678
Title: Autonomous driving using Model Predictive ContouringControl.
'''
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
from matplotlib.transforms import Affine2D
from matplotlib.animation import FuncAnimation
from matplotlib.animation import FFMpegWriter
from functions.Car_Dynamics import *
from functions.HelpingFunctions import *
from functions.MPCCFunctions import *
# %matplotlib qt
fig, ax = plt.subplots(figsize=(15,15)) # You may need to adjust the figsize to fit your screen
# Draw the road
L = 50.0 # length of straight section [m]
R = 5.0 # radius of curved section centerline [m]
w = 2 # width of lane [m]
Ri = R-w # radius of inner boundary [m]
Ro = R+w # radius of outer boundary [m]
ds = 0.5 # step size [m]
# (xc,yc) = coordinates of center dotted line
# (xi,yi) = coordinates of inner solid line
# (xo,yo) = coordinates of outer solid line
# xc = np.hstack((np.arange(0,L,ds), L+R*np.cos(np.arange(-np.pi/2,np.pi/2,ds/R)),
# np.arange(L,0,-ds), R*np.cos(np.arange(np.pi/2,3*np.pi/2,ds/R))))
# yc = np.hstack((np.zeros_like(np.arange(0,L,ds)), R+R*np.sin(np.arange(-np.pi/2,np.pi/2,ds/R)),
# np.full_like(np.arange(L,0,-ds),2*R), R+R*np.sin(np.arange(np.pi/2,3*np.pi/2,ds/R))))
xc = np.hstack((np.arange(0,L,ds), L+R*np.cos(np.arange(-np.pi/2,np.pi/2,ds/R)),
np.arange(L,0,-ds), R*np.cos(np.arange(np.pi/2,3*np.pi/2,ds/R))))
yc = np.hstack((np.zeros_like(np.arange(0,L,ds)), R+R*np.sin(np.arange(-np.pi/2,np.pi/2,ds/R)),
np.full_like(np.arange(L,0,-ds),2*R), R+R*np.sin(np.arange(np.pi/2,3*np.pi/2,ds/R))))
w_i,w_o = w*np.ones_like(xc),-w*np.ones_like(xc)
w_i[50:60] = -0.5
w_o[80:90] = -0.5
w_o[100:120] = -1
w_i[160:170] = -0.75
# w_i[240:250] = 0
w_o[220:225] = 0
xi,yi=GenOffsetTrack(xc,yc,w_i)
xo,yo=GenOffsetTrack(xc,yc,w_o)
plt.plot(xc,yc,color='b',dashes=[20, 6])
plt.plot(xi,yi,color='k')
plt.plot(xo,yo,color='k')
ax.set_aspect('equal')
ax.axis('off')
# plt.show()
# fig.savefig('Obstacles')
t_max = 12.0 # final time of the simulat
dt = 0.005
ini_states = np.array([5,1,0,1,0,0])
CarD=CarDynamics(ini_states,dt)
print('Generated car object.')
# Draw the Duckiebot
Scale = 20 #43
l_car = (CarD.a1 + CarD.a2)*Scale
w_car = (l_car/3)
r_cam = 0.4 # camera range [m]
phi_cam = 45.0 # camera angle half-width [deg]
car = mpatches.Rectangle([-CarD.a2*Scale,-w_car/2],l_car,w_car,linewidth=1,edgecolor='k',facecolor='r')
ax.add_patch(car)
cam_view = mpatches.Wedge((CarD.a1,0),r_cam,-phi_cam,phi_cam,facecolor='yellow',edgecolor=None)
ax.add_patch(cam_view)
text = plt.annotate('',(L/1.5-0.25,R),fontsize=8,animated=True)
path, = ax.plot([],[],color = 'r',linewidth=2,alpha = 1, aa = True,label = 'Local track')
clpoint, = ax.plot([],[],color = 'k',marker='o',linewidth=5,alpha = 1, aa = True,label='Closest point')
nomialpath, = ax.plot([],[],color = 'k',linewidth=1,alpha = 1, aa = True,label='Predicted by MPCC')
bezierpath, = ax.plot([],[],color = 'g',linewidth=1,alpha = 1, aa = True,label='Predicted progress')
truetrajectory, = ax.plot([],[],color = 'm', linewidth = 2, alpha = 0.75, aa = True, label = 'True COM trajectory')
TruetrajX,TruetrajY = [ini_states[0]],[ini_states[1]]
def GetLocalBezierXY(Xcar,Ycar,length,beforelen,degree,numpts):
global xc,yc,ds
Maxidx = len(xc)
Startidx = findclosestONTrack(Xcar,Ycar,xc,yc) - int(beforelen/ds)
Endidx = Startidx + int(length/ds)
if Startidx%Maxidx > Endidx%Maxidx:
pointsxc = np.append(xc[Startidx%Maxidx::],xc[0:Endidx%Maxidx])
pointsyc = np.append(yc[Startidx%Maxidx::],yc[0:Endidx%Maxidx])
else:
pointsxc = xc[Startidx%Maxidx:Endidx%Maxidx]
pointsyc = yc[Startidx%Maxidx:Endidx%Maxidx]
t_points = np.linspace(0,1,len(pointsxc))
alphaX = fitCurvetopoints(pointsxc,t_points,degree)
alphaY = fitCurvetopoints(pointsyc,t_points,degree)
t = np.linspace(0,1,numpts)
tc = ProjectOnBCurveXY(np.array([alphaX,alphaY]),[Xcar,Ycar])
Xvec = GetBezierOut(alphaX,np.append(t,tc))
Yvec = GetBezierOut(alphaY,np.append(t,tc))
return Xvec[0:-1],Yvec[0:-1],Xvec[-1],Yvec[-1],alphaX,alphaY,tc #refpath bez, closestpoint on bez
inputs=[0,0,0]
fps = 30
N_mpcc = 30 #horizon length of the MPC
L_LocalTrack = 20
mpcc = MPCC_Object(N_mpcc,N_mpcc/fps,xc,yc,xi,yi,xo,yo)
mpcc.SetInitialCondition(CarD.states,np.array([0]),[0,2,3,4,5,20],[0,0,0,0,0,0])
print('Initilization Complete.')
print('Starting Simulation.')
# Initialize the axes for animation
def init():
ax.set_aspect('equal')
ax.axis('off')
ax.legend()
return (car,cam_view,text,path,clpoint,nomialpath,bezierpath,truetrajectory)
def update(t):
global CarD,mpcc
x,y,tht,_,_,_=CarD.states
car.set_transform(Affine2D().rotate(tht) + Affine2D().translate(x,y) + ax.transData)
cam_view.set_transform(Affine2D().rotate(tht) + Affine2D().translate(x,y) + ax.transData)
Xbez,Ybez,Xcen,Ycen,alphaX,alphaY,tc= GetLocalBezierXY(x,y,L_LocalTrack,L_LocalTrack/10,7,50)
path.set_data(Xbez,Ybez)
clpoint.set_data(Xcen,Ycen)
mpcc.SetInitialCondition(CarD.states,tc,alphaX,alphaY)
# mpcc.SequentiallySolveMPCC(20,1)
try:
mpcc.SequentiallySolveMPCC(20,1)#(num_iterations,changethreshold)
except:
pass
Xnom,Ynom = mpcc.GetNominalPath()
nomialpath.set_data(Xnom,Ynom)
BXnom,BYnom = mpcc.GetBezierNominalPath()
bezierpath.set_data(BXnom,BYnom)
inputs = list(mpcc._u[0].value.squeeze())
for _ in range(0,int(1//(fps*CarD.dt))):
CarD.sim_step(inputs)
TruetrajX.append(CarD.states[0])
TruetrajY.append(CarD.states[1])
truetrajectory.set_data(TruetrajX,TruetrajY)
text.set_text(f't:{t:.1f} sec\n Steering: {np.rad2deg(inputs[0]):0.1f} deg \n Thrust: {inputs[2]:0.1f} N')
return (car,cam_view,text,path,clpoint,nomialpath,bezierpath,truetrajectory)
# for _ in range(1000):
# update(0)
ani = FuncAnimation(fig, update, frames=np.linspace(0,t_max,int(t_max*fps)), init_func=init,
interval=1000//fps, blit=True, repeat=False)
plt.show()
# plt.close() # hide display of final frame
# plt.rcParams['animation.ffmpeg_path'] = r'C:\Users\amkulk\Downloads\ffmpeg-2021-05-19-git-2261cc6d8a-essentials_build\bin\ffmpeg.exe'
# writer = FFMpegWriter(fps=fps)
# ani.save('T20Obstacle.mp4', writer=writer)
# ##saving final trajectory
# fig1, ax1 = plt.subplots(figsize=((L)//2,(2*R+2*w)//2))
# plt.plot(xc,yc,color='b',dashes=[20, 6])
# plt.plot(xi,yi,color='k')
# plt.plot(xo,yo,color='k')
# plt.plot(TruetrajX,TruetrajY,color = 'm', linewidth = 2, alpha = 0.75, aa = True, label = 'COM Trajectory')
# ax1.set_aspect('equal')
# ax1.axis('off')
# fig1.savefig('T20Obstacle')