InteligentnySaper/classes/bfs.py

235 lines
11 KiB
Python

import heapq # dla utrzymania fringe
from classes import node, minesweeper, system
class BFS:
window: system.Window
agent: minesweeper.Minesweeper
node: node.Node
def __init__(self, agent, window):
self.agent = agent
self.window = window
def successor(self, current_position):
new_nodes = []
neighbours_list = self.agent.sensor(current_position[0], current_position[1])
if current_position[2] == 180: # jesli patrzy na polnoc
if neighbours_list[0][1] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
if neighbours_list[0][1] == 'grass':
cost = 10
elif neighbours_list[0][1] == 'sand':
cost = 2
elif neighbours_list[0][1] == 'mine':
cost = 0
tmp = ('forward', [current_position[0], current_position[1] - 1, 180], cost)
new_nodes.append(tmp)
if neighbours_list[1][0] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[1][0] == 'grass':
# cost = 10
# elif neighbours_list[1][0] == 'sand':
# cost = 2
# elif neighbours_list[1][0] == 'mine':
# cost = 0
tmp = ('left', [current_position[0],current_position[1], 270], 1)
new_nodes.append(tmp)
if neighbours_list[1][2] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[1][2] == 'grass':
# cost = 10
# elif neighbours_list[1][2] == 'sand':
# cost = 2
# elif neighbours_list[1][2] == 'mine':
# cost = 0
tmp = ('right', [current_position[0], current_position[1], 90], 1)
new_nodes.append(tmp)
if current_position[2] == 90: # jesli patrzy na wschod
if neighbours_list[1][2] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
if neighbours_list[1][2] == 'grass':
cost = 10
elif neighbours_list[1][2] == 'sand':
cost = 2
elif neighbours_list[1][2] == 'mine':
cost = 0
tmp = ('forward', [current_position[0] + 1, current_position[1], 90], cost)
new_nodes.append(tmp)
if neighbours_list[0][1] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[0][1] == 'grass':
# cost = 10
# elif neighbours_list[0][1] == 'sand':
# cost = 2
# elif neighbours_list[0][1] == 'mine':
# cost = 0
tmp = ('left', [current_position[0], current_position[1], 180], 1)
new_nodes.append(tmp)
if neighbours_list[2][1] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[2][1] == 'grass':
# cost = 10
# elif neighbours_list[2][1] == 'sand':
# cost = 2
# elif neighbours_list[2][1] == 'mine':
# cost = 0
tmp = ('right', [current_position[0], current_position[1], 0], 1)
new_nodes.append(tmp)
if current_position[2] == 0: # jesli patczy na poludzie
if neighbours_list[2][1] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
if neighbours_list[2][1] == 'grass':
cost = 10
elif neighbours_list[2][1] == 'sand':
cost = 2
elif neighbours_list[2][1] == 'mine':
cost = 0
tmp = ('forward', [current_position[0], current_position[1] + 1, 0], cost)
new_nodes.append(tmp)
if neighbours_list[1][2] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[1][2] == 'grass':
# cost = 10
# elif neighbours_list[1][2] == 'sand':
# cost = 2
# elif neighbours_list[1][2] == 'mine':
# cost = 0
tmp = ('left', [current_position[0], current_position[1], 90], 1)
new_nodes.append(tmp)
if neighbours_list[1][0] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[1][0] == 'grass':
# cost = 10
# elif neighbours_list[1][0] == 'sand':
# cost = 2
# elif neighbours_list[1][0] == 'mine':
# cost = 0
tmp = ('right', [current_position[0],current_position[1], 270], 1)
new_nodes.append(tmp)
if current_position[2] == 270: # jesli patczy na wschod
if neighbours_list[1][0] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
if neighbours_list[1][0] == 'grass':
cost = 10
elif neighbours_list[1][0] == 'sand':
cost = 2
elif neighbours_list[1][0] == 'mine':
cost = 0
tmp = ('forward', [current_position[0] - 1,current_position[1], 270], cost)
new_nodes.append(tmp)
if neighbours_list[2][1] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[2][1] == 'grass':
# cost = 10
# elif neighbours_list[2][1] == 'sand':
# cost = 2
# elif neighbours_list[2][1] == 'mine':
# cost = 0
tmp = ('left', [current_position[0], current_position[1], 0], 1)
new_nodes.append(tmp)
if neighbours_list[0][1] not in ['wall', 'cliff_south', 'cliff_east', 'cliff_north', 'cliff_west']:
# if neighbours_list[0][1] == 'grass':
# cost = 10
# elif neighbours_list[0][1] == 'sand':
# cost = 2
# elif neighbours_list[0][1] == 'mine':
# cost = 0
tmp = ('right', [current_position[0], current_position[1], 180], 1)
new_nodes.append(tmp)
return new_nodes
# fringe = struktura danych przeszowyjąca wierchowki do odwiedzenia
# explored = lista odwiedzonych stanow
# position_at_beginning = stan poczatkowy
# succ = funkcja nastempnika
# goaltest = test spewnienia celu
def graphsearch(self, fringe, explored, succ, goaltest):
def manhattan(state, target_state):
return abs(state[0] - target_state[0]) + abs(state[1] - target_state[1])
self.window.pause(True)
position_at_beginning = [self.agent.position_x, self.agent.position_y, self.agent.rotation_degrees] # x, y, gdzie_patczy
final_action_list = [] # lista co ma robic zeby dojechac do miny
root = node.Node(None, None, position_at_beginning, 0) # parent, action, position, cost
heapq.heappush(fringe, (0, root)) # add first node to fringe
while len(fringe) != 0: # poki sa wezly do odwiedzenia(na fringe)
if len(fringe) == 0:
return False
# get node from fringe
tmp_node = heapq.heappop(fringe) # tuple(priority, node)
tmp_node_position = tmp_node[1].get_position() # x, y , gdzie patczy
# jesli tmp node to goaltest
if tmp_node_position[:2] == goaltest:
print('Find')
while tmp_node[1].get_parent() is not None:
final_action_list.append(tmp_node[1].get_action())
tmp_node = tmp_node[1].get_parent()
final_action_list.reverse()
#print(final_action_list)
self.window.pause(False)
return final_action_list
explored.append(tmp_node[1]) # add node to array of visited nodes
neighbours_list_of_our_node = self.successor(tmp_node_position) # lista możliwych akcij
for node_ in neighbours_list_of_our_node:
notInFringe = True # false if node in fringe
notInExplored = True # false if node in explored
# if node_[1] is None:
# continue
p = manhattan(node_[1], goaltest) + node_[2]
# wyznacza jaki jest priorytet nastempnika
# manchaten from node_ + cost of way from tmp_ode to node_
priority_in_fringe = 0
counter = 0
tmp_position_in_fringe = 0 # zero if node not in fringe
for fringeNode in fringe: # isc po wszystkich wezlach ktore juz sa w fringe
# jesli nasz wezel juz jest w fringe
if fringeNode[1].get_position()[0] == node_[1][0] and fringeNode[1].get_position()[1] == node_[1][1] and fringeNode[1].get_position()[2] == node_[1][2]:
notInFringe = False
priority_in_fringe = fringeNode[0]
# number of element in fringe
tmp_position_in_fringe = counter
counter = counter + 1
for exploredNode in explored: # isc po wszystkich wezlach z listy explored
# jesli nasz wezel juz jest w explored
if exploredNode.get_position()[0] == node_[1][0] and exploredNode.get_position()[1] == node_[1][1] and exploredNode.get_position()[2] == node_[1][2]:
notInExplored = False
# if node not in fringe and not in explored
if notInFringe and notInExplored:
x = node.Node(tmp_node, node_[0], node_[1], node_[2]) # parent, action, state_array, cost
heapq.heappush(fringe, (p, x))
# if node not in fringe
elif notInFringe is False and (priority_in_fringe > p):
x = node.Node(tmp_node, node_[0], node_[1], node_[2]) # parent, action, state_array, cost
tmp = list(fringe[tmp_position_in_fringe])
tmp[0] = p
tmp[1] = x
fringe[tmp_position_in_fringe] = tuple(tmp)
#self.window.draw_search([self.agent.position_x, self.agent.position_y], [node_[1][0], node_[1][1]],self.agent.current_map.tile_size, self.agent.current_map, self.agent)