101 lines
3.3 KiB
Python
101 lines
3.3 KiB
Python
from queue import PriorityQueue
|
|
|
|
DEFAULT_COST_VALUE = 1
|
|
|
|
def is_border(x, y, max_x, max_y):
|
|
return 0 <= x < max_x and 0 <= y < max_y
|
|
|
|
def is_obstacle(x, y, obstacles):
|
|
return (x, y) in obstacles
|
|
|
|
def succ(current_state, max_x, max_y, obstacles):
|
|
successors = []
|
|
x, y, direction = current_state
|
|
|
|
# Akcja: Do przodu
|
|
direction_x, direction_y = {'N': (0, -1), 'E': (1, 0), 'S': (0, 1), 'W': (-1, 0)}[direction] # Słownik przesunięć w zależności od kierunku
|
|
new_x, new_y = x + direction_x, y + direction_y
|
|
|
|
if is_border(new_x, new_y, max_x, max_y) and not(is_obstacle(new_x, new_y, obstacles)):
|
|
successors.append(((new_x, new_y, direction), 'Go Forward'))
|
|
|
|
# Akcja: Obrót w lewo
|
|
left_turns = {'N': 'W', 'W': 'S', 'S': 'E', 'E': 'N'} # Słownik kierunków po obrocie w lewo
|
|
successors.append(((x, y, left_turns[direction]), 'Turn Left'))
|
|
|
|
# Akcja: Obrót w prawo
|
|
right_turns = {'N': 'E', 'E': 'S', 'S': 'W', 'W': 'N'} # Słownik kierunków po obrocie w prawo
|
|
successors.append(((x, y, right_turns[direction]), 'Turn Right'))
|
|
|
|
return successors
|
|
|
|
def graphsearch(istate, goal, max_x, max_y, obstacles, cost_map):
|
|
fringe = PriorityQueue()
|
|
explored = set()
|
|
|
|
fringe.put((0, (istate, None , None)))
|
|
|
|
while not fringe.empty():
|
|
_, node = fringe.get()
|
|
state, _, _ = node
|
|
|
|
if goaltest(state, goal):
|
|
return build_action_sequence(node)
|
|
|
|
explored.add(state)
|
|
|
|
successors = succ(state, max_x, max_y, obstacles)
|
|
|
|
for new_state, action in successors:
|
|
new_node = (new_state, node, action)
|
|
|
|
p_new_state = current_cost(node, cost_map) + heuristic(state, goal)
|
|
|
|
if not is_state_in_queue(new_state, fringe) and new_state not in explored:
|
|
fringe.put((p_new_state, new_node))
|
|
|
|
elif is_state_in_queue(new_state, fringe):
|
|
for i, (p_existing_state, (existing_state, _, _)) in enumerate(fringe.queue):
|
|
if existing_state == new_state and p_existing_state > p_new_state:
|
|
fringe.queue[i] = (p_new_state, new_node)
|
|
else:
|
|
break
|
|
|
|
return False
|
|
|
|
def is_state_in_queue(state, queue):
|
|
for _, (s, _, _) in queue.queue:
|
|
if s == state:
|
|
return True
|
|
return False
|
|
|
|
def build_action_sequence(node):
|
|
actions = []
|
|
while node[1] is not None: # Dopóki nie dojdziemy do korzenia
|
|
_, parent, action = node
|
|
actions.append(action)
|
|
node = parent
|
|
actions.reverse()
|
|
return actions
|
|
|
|
def goaltest(state, goal):
|
|
x, y, _ = state
|
|
goal_x, goal_y = goal
|
|
return (x,y) == (goal_x, goal_y)
|
|
|
|
def current_cost(node, cost_map):
|
|
cost = 0
|
|
while node[1] is not None: # Dopóki nie dojdziemy do korzenia
|
|
_, parent, action = node
|
|
# Dodaj koszt pola z mapy kosztów tylko jeśli akcja to "Forward"
|
|
# if action == 'Go Forward':
|
|
if True:
|
|
state, _, _ = node
|
|
cost += cost_map.get(state[:2], DEFAULT_COST_VALUE) # Pobiera koszt przejścia przez dane pole, a jeśli koszt nie jest zdefiniowany to bierze wartość domyślną
|
|
node = parent # Przejdź do rodzica
|
|
return cost
|
|
|
|
def heuristic(state, goal):
|
|
x, y, _ = state
|
|
goal_x, goal_y = goal
|
|
return abs(x - goal_x) + abs(y - goal_y) # Odległość Manhattana do celu |