mirror of
https://github.com/AtsushiSakai/PythonRobotics.git
synced 2026-02-11 07:45:23 -05:00
133 lines
3.2 KiB
Python
133 lines
3.2 KiB
Python
#! /usr/bin/python
|
|
# -*- coding: utf-8 -*-
|
|
u"""
|
|
|
|
Path tracking simulation with pure pursuit steering control and PID speed control.
|
|
|
|
author: Atsushi Sakai
|
|
|
|
"""
|
|
import numpy as np
|
|
import math
|
|
import matplotlib.pyplot as plt
|
|
import unicycle_model
|
|
|
|
Kp = 1.0 # speed propotional gain
|
|
Lf = 3.0 # look-ahead distance
|
|
|
|
|
|
def PIDControl(target, current):
|
|
a = Kp * (target - current)
|
|
|
|
return a
|
|
|
|
|
|
def pure_pursuit_control(state, cx, cy, pind):
|
|
|
|
if state.v >= 0:
|
|
ind = calc_nearest_index(state, cx[pind:], cy[pind:])
|
|
else:
|
|
ind = calc_nearest_index(state, cx[:pind + 1], cy[:pind + 1])
|
|
|
|
if state.v >= 0:
|
|
ind = ind + pind
|
|
|
|
tx = cx[ind]
|
|
ty = cy[ind]
|
|
|
|
alpha = math.atan2(ty - state.y, tx - state.x) - state.yaw
|
|
|
|
if state.v < 0: # back
|
|
if alpha > 0:
|
|
alpha = math.pi - alpha
|
|
else:
|
|
alpha = math.pi + alpha
|
|
|
|
delta = math.atan2(2.0 * unicycle_model.L * math.sin(alpha) / Lf, 1.0)
|
|
|
|
if state.v < 0: # back
|
|
delta = delta * -1.0
|
|
|
|
return delta, ind
|
|
|
|
|
|
def calc_nearest_index(state, cx, cy):
|
|
dx = [state.x - icx for icx in cx]
|
|
dy = [state.y - icy for icy in cy]
|
|
|
|
d = [abs(math.sqrt(idx ** 2 + idy ** 2) -
|
|
Lf) for (idx, idy) in zip(dx, dy)]
|
|
|
|
ind = d.index(min(d))
|
|
|
|
return ind
|
|
|
|
|
|
def main():
|
|
# target course
|
|
cx = np.arange(0, 50, 0.1)
|
|
cy = [math.sin(ix / 5.0) * ix / 2.0 for ix in cx]
|
|
target_speed = 30.0 / 3.6
|
|
|
|
T = 15.0 # max simulation time
|
|
|
|
state = unicycle_model.State(x=-1.0, y=-5.0, yaw=0.0, v=0.0)
|
|
# state = unicycle_model.State(x=-1.0, y=-5.0, yaw=0.0, v=-30.0 / 3.6)
|
|
# state = unicycle_model.State(x=10.0, y=5.0, yaw=0.0, v=-30.0 / 3.6)
|
|
# state = unicycle_model.State(
|
|
# x=3.0, y=5.0, yaw=math.radians(-40.0), v=-10.0 / 3.6)
|
|
# state = unicycle_model.State(
|
|
# x=3.0, y=5.0, yaw=math.radians(40.0), v=50.0 / 3.6)
|
|
|
|
lastIndex = len(cx) - 1
|
|
time = 0.0
|
|
x = [state.x]
|
|
y = [state.y]
|
|
yaw = [state.yaw]
|
|
v = [state.v]
|
|
t = [0.0]
|
|
target_ind = calc_nearest_index(state, cx, cy)
|
|
|
|
while T >= time and lastIndex > target_ind:
|
|
ai = PIDControl(target_speed, state.v)
|
|
di, target_ind = pure_pursuit_control(state, cx, cy, target_ind)
|
|
state = unicycle_model.update(state, ai, di)
|
|
|
|
time = time + unicycle_model.dt
|
|
|
|
x.append(state.x)
|
|
y.append(state.y)
|
|
yaw.append(state.yaw)
|
|
v.append(state.v)
|
|
t.append(time)
|
|
|
|
# plt.cla()
|
|
# plt.plot(cx, cy, ".r", label="course")
|
|
# plt.plot(x, y, "-b", label="trajectory")
|
|
# plt.plot(cx[target_ind], cy[target_ind], "xg", label="target")
|
|
# plt.axis("equal")
|
|
# plt.grid(True)
|
|
# plt.pause(0.1)
|
|
# input()
|
|
|
|
flg, ax = plt.subplots(1)
|
|
plt.plot(cx, cy, ".r", label="course")
|
|
plt.plot(x, y, "-b", label="trajectory")
|
|
plt.legend()
|
|
plt.xlabel("x[m]")
|
|
plt.ylabel("y[m]")
|
|
plt.axis("equal")
|
|
plt.grid(True)
|
|
|
|
flg, ax = plt.subplots(1)
|
|
plt.plot(t, [iv * 3.6 for iv in v], "-r")
|
|
plt.xlabel("Time[s]")
|
|
plt.ylabel("Speed[km/h]")
|
|
plt.grid(True)
|
|
plt.show()
|
|
|
|
|
|
if __name__ == '__main__':
|
|
print("Pure pursuit path tracking simulation start")
|
|
main()
|