def run_animation(all_sheep, sheep_dict, herd_a, herd_b):
step = 0
target = np.array([560, 560])
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
last_vector = np.zeros((n, 2), dtype=np.float32)
radius = math.sqrt(n) * r_rep
while True:
herd_point_a = herd_a.position2point().copy()
herd_point_b = herd_b.position2point().copy()
# A-聚集 + 驱赶
# B-驱赶
if common.check(all_sheep, radius):
R1.driving(herd_a, all_sheep, speed, target, app_dist)
else:
R1.collecting(herd_a, all_sheep, speed, app_dist, target)
R1.driving(herd_b, all_sheep, speed, target, app_dist)
sheepR.sheep_move(herd_point_a, herd_point_b, all_sheep, r_dist, r_rep,
speed, sheep_dict, last_vector)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 4000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd_a.delete()
herd_b.delete()
break
step += 1
return step
def run_animation(all_sheep, sheep_dict, herd, theta, k):
step = 0
target = np.array([600, 600])
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
fn = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
while True:
herd_point = herd.position2point().copy()
if common.check_sector(all_sheep, theta, target) and common.check_dist(
all_sheep, target, fn):
shepherdR.driving(herd, all_sheep, speed, target, app_dist)
else:
shepherdR.collecting(herd, all_sheep, speed, app_dist, target)
sheepR2.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector, k)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 2000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
break
step += 1
return step
def run_animation(all_sheep, sheep_dict, herd):
step = 0
target = np.array([600, 600])
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = 50
radius = math.sqrt(n) * r_rep
while True:
herd_point = herd.position2point().copy()
if common.check(all_sheep, radius):
shepherdR.driving(herd, all_sheep, speed, target, app_dist)
else:
shepherdR.collecting(herd, all_sheep, speed, app_dist)
boid.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 4000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
break
step += 1
return step
def run_animation(all_sheep, sheep_dict, herd, canvas):
step = 0
target = np.array([550, 550])
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
theta = math.pi/6
fn = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
while True:
herd_point = herd.position2point().copy()
whichmode = ""
if common.check_sector(all_sheep, theta, target) and common.check_dist(all_sheep, target, fn):
shepherdRR.driving(herd, all_sheep, speed, target, app_dist)
text = canvas.create_text(220, 20, text="driving", font=("宋体", 18))
whichmode = "driving"
canvas.pack()
else:
idx = shepherdRR.collecting(herd, all_sheep, speed, app_dist, target)
text = canvas.create_text(220, 20, text="collecting", font=("宋体", 18))
whichmode = "collecting"
canvas.pack()
step += 1
step_txt = canvas.create_text(420, 20, text="steps: {}/2000".format(step), font=("宋体", 18))
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed, sheep_dict, last_vector)
tk.update()
time.sleep(0.01)
canvas.delete(text)
canvas.delete(step_txt)
if step == 100:
print("[", end="")
for per in all_sheep:
print("[{}, {}], ".format(per[0], per[1]), end="")
print("]")
print(herd.position2point())
print(idx)
print(whichmode)
break
if common.is_all_in_target(all_sheep) or step > 4000:
print("[", end="")
for per in all_sheep:
print("[{}, {}], ".format(per[0], per[1]), end="")
print("]")
print(herd.position2point())
print(whichmode)
print(idx)
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
break
return step
def run_animation(all_sheep, sheep_dict, herd, k):
step = 0
target = np.array([550, 550], np.float64)
r_rep = 14
speed = 2
r_dist = 240
n = len(all_sheep)
app_dist = n + 60
fn = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
theta = math.pi / 6
"""设定新旧行为策略"""
herd_instance = shepherdR.Shepherd(cb.OldCBehavior(), k)
# herd_instance = shepherdR.Shepherd(cb.NewCBehavior())
while True:
herd_point = herd.position2point().copy()
if herd_instance.switch(all_sheep, target, fn, theta):
herd_instance.driving(herd, all_sheep, speed, target, app_dist)
text = canvas.create_text(220, 20, text="driving", font=("宋体", 18))
canvas.pack()
else:
herd_instance.collecting(herd, all_sheep, speed, app_dist, target)
text = canvas.create_text(220,
20,
text="collecting",
font=("宋体", 18))
canvas.pack()
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector)
step += 1
step_txt = canvas.create_text(420,
20,
text="steps: {}/2000".format(step),
font=("宋体", 18))
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector)
tk.update()
canvas.delete(text)
canvas.delete(step_txt)
if common.is_all_in_target(all_sheep) or step > 2000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
break
return step
def run_animation(all_sheep, sheep_dict, herd):
step = 0
target = np.array([550, 550], np.float64)
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
fn = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
theta = math.pi / 6
"""设定新旧行为策略"""
herd_instance = shepherdR.Shepherd(cb.OldCBehavior())
# herd_instance = shepherdR.Shepherd(cb.NewCBehavior())
while True:
herd_point = herd.position2point().copy()
if herd_instance.switch(all_sheep, target, fn, theta):
herd_instance.driving(herd, all_sheep, speed, target, app_dist)
else:
herd_instance.collecting(herd, all_sheep, speed, app_dist, target)
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 4000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
break
step += 1
return step
def run_animation(all_sheep, sheep_dict, herd):
step = 0
target = np.array([600, 600])
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
theta = math.pi / 4.5
fn = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
# 处理轨迹
UU = []
VV = []
XX = []
YY = []
Px = {}
Py = {}
for i in range(n):
Px['coor' + str(i)] = []
Py['coor' + str(i)] = []
for i in range(n):
Px['coor' + str(i)].append(all_sheep[i][0])
Py['coor' + str(i)].append(all_sheep[i][1])
dist_center = 0
dist_shepherd = 0
pre_mean = np.array([np.mean(all_sheep[:, 0]), np.mean(all_sheep[:, 1])])
pre_herd = herd.position2point()
while True:
herd_point = herd.position2point().copy()
if common.check_sector(all_sheep, theta, target) and common.check_dist(
all_sheep, target, fn):
shepherdR.driving(herd, all_sheep, speed, target, app_dist)
else:
shepherdR.collecting(herd, all_sheep, speed, app_dist, target)
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 4000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
print("dispersion: ", common.calculate_dispersion(all_sheep))
break
# 处理轨迹
global_mean = np.array(
[np.mean(all_sheep[:, 0]),
np.mean(all_sheep[:, 1])])
dist_center += la.norm(pre_mean - global_mean)
pre_mean = global_mean
dist_shepherd += la.norm(pre_herd - herd.position2point())
pre_herd = herd.position2point()
for i in range(n):
Px['coor' + str(i)].append(all_sheep[i][0])
Py['coor' + str(i)].append(all_sheep[i][1])
XX.append(herd.position2point()[0])
YY.append(herd.position2point()[1])
UU.append(global_mean[0])
VV.append(global_mean[1])
step += 1
xx = np.array(XX)
yy = np.array(YY)
common.print_list(xx)
common.print_list(yy)
return step
def run_animation(all_sheep, sheep_dict, herd):
step = 0
target = np.array([600, 600], np.float64)
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
radius = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
# 处理轨迹
UU = []
VV = []
XX = []
YY = []
Px = {}
Py = {}
for i in range(n):
Px['coor' + str(i)] = []
Py['coor' + str(i)] = []
for i in range(n):
Px['coor' + str(i)].append(all_sheep[i][0])
Py['coor' + str(i)].append(all_sheep[i][1])
dist_center = 0
dist_shepherd = 0
pre_mean = np.array([np.mean(all_sheep[:, 0]), np.mean(all_sheep[:, 1])])
pre_herd = herd.position2point()
while True:
herd_point = herd.position2point().copy()
if common.check(all_sheep, radius):
shepherdR.driving(herd, all_sheep, speed, target, app_dist)
else:
shepherdR.collecting(herd, all_sheep, speed, app_dist)
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 4000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
print("dispersion: ", common.calculate_dispersion(all_sheep))
break
# 处理轨迹
global_mean = np.array(
[np.mean(all_sheep[:, 0]),
np.mean(all_sheep[:, 1])])
dist_center += la.norm(pre_mean - global_mean)
pre_mean = global_mean
dist_shepherd += la.norm(pre_herd - herd.position2point())
pre_herd = herd.position2point()
for i in range(n):
Px['coor' + str(i)].append(all_sheep[i][0])
Py['coor' + str(i)].append(all_sheep[i][1])
XX.append(herd.position2point()[0])
YY.append(herd.position2point()[1])
UU.append(global_mean[0])
VV.append(global_mean[1])
step += 1
# 处理轨迹
fig, ax = plt.subplots()
xx = np.array(XX)
yy = np.array(YY)
uu = np.array(UU)
vv = np.array(VV)
for i in range(n):
plt.plot(np.array(Px['coor' + str(i)]),
600 - np.array(Py['coor' + str(i)]),
linewidth=0.1,
c='silver')
plt.plot(xx, 600 - yy, c='darkkhaki', label='shepherd', linewidth=2)
plt.plot(uu,
600 - vv,
'-.',
c='darkcyan',
label='sheep center',
linewidth=2)
plt.legend()
y = np.arange(0, 600, 100)
plt.xticks(y)
plt.yticks(y)
plt.xlabel("X Position")
plt.ylabel("Y Position")
plt.show()
fig.savefig('E:\\我的坚果云\\latex\\doubleDistSum\\pics\\SPPL_trial60.pdf',
dpi=600,
format='pdf')
## 输出距离
print("dist_center:", dist_center)
print("dist_shepherd:", dist_shepherd)
return step
def run_animation(all_sheep, sheep_dict, herd):
step = 0
target = np.array([600, 600])
r_dist = 250
r_rep = 14
speed = 2
n = len(all_sheep)
app_dist = n + 50
theta = math.pi / 4.5
fn = math.sqrt(n) * r_rep
last_vector = np.zeros((n, 2), dtype=np.float32)
# 处理轨迹
UU = []
VV = []
XX = []
YY = []
Px = {}
Py = {}
for i in range(n):
Px['coor' + str(i)] = []
Py['coor' + str(i)] = []
for i in range(n):
Px['coor' + str(i)].append(all_sheep[i][0])
Py['coor' + str(i)].append(all_sheep[i][1])
dist_center = 0
dist_shepherd = 0
pre_mean = np.array([np.mean(all_sheep[:, 0]), np.mean(all_sheep[:, 1])])
pre_herd = herd.position2point()
while True:
herd_point = herd.position2point().copy()
if common.check_sector(all_sheep, theta, target) and common.check_dist(
all_sheep, target, fn):
shepherdR.driving(herd, all_sheep, speed, target, app_dist)
else:
shepherdR.collecting(herd, all_sheep, speed, app_dist, target)
sheepR.sheep_move(herd_point, all_sheep, r_dist, r_rep, speed,
sheep_dict, last_vector)
tk.update()
time.sleep(0.01)
if common.is_all_in_target(all_sheep) or step > 4000:
for per_sheep in sheep_dict.values():
per_sheep.delete()
herd.delete()
print("dispersion: ", common.calculate_dispersion(all_sheep))
break
# 处理轨迹
global_mean = np.array(
[np.mean(all_sheep[:, 0]),
np.mean(all_sheep[:, 1])])
dist_center += la.norm(pre_mean - global_mean)
pre_mean = global_mean
dist_shepherd += la.norm(pre_herd - herd.position2point())
pre_herd = herd.position2point()
for i in range(n):
Px['coor' + str(i)].append(all_sheep[i][0])
Py['coor' + str(i)].append(all_sheep[i][1])
XX.append(herd.position2point()[0])
YY.append(herd.position2point()[1])
UU.append(global_mean[0])
VV.append(global_mean[1])
step += 1
# 处理轨迹
fig, ax = plt.subplots()
xx = np.array(XX)
yy = np.array(YY)
uu = np.array(UU)
vv = np.array(VV)
for i in range(n):
plt.plot(np.array(Px['coor' + str(i)]),
600 - np.array(Py['coor' + str(i)]),
linewidth=0.1,
c="silver",
zorder=1)
plt.plot(xx,
600 - yy,
c='darkkhaki',
label='shepherd',
linewidth=2,
zorder=2)
# plt.plot(uu, 600 - vv, '-.', c='darkcyan', label='sheep center', linewidth=2)
y = np.arange(0, 700, 100)
plt.xlabel("X Position")
plt.ylabel("Y Position")
n = 50
X = list()
for i in range(n - 1):
np.random.seed(i)
x = np.random.randint(300, 400)
y = np.random.randint(200, 350)
X.append([x, y])
# outlier sheep
Y = np.array([50, 550])
X = np.array(X)
plt.scatter(X[:, 0], X[:, 1], label="normal sheep", zorder=3)
plt.scatter(Y[0], Y[1], label="outlier sheep", zorder=3)
range_x = np.arange(0, 700, 100)
plt.xticks(range_x)
plt.yticks(range_x)
# plt.grid()
plt.legend()
plt.show()
fig.savefig(
'E:\\我的坚果云\\latex\\doubleDistSum\\pics\\outlier_a_mdaf_trajectory.pdf',
dpi=600,
format='pdf')
## 输出距离
print("dist_center:", dist_center)
print("dist_shepherd:", dist_shepherd)
return step
