diff --git a/Sampling-based Planning/RRT*.py b/Sampling-based Planning/RRT*.py new file mode 100644 index 0000000..c4d1167 --- /dev/null +++ b/Sampling-based Planning/RRT*.py @@ -0,0 +1,200 @@ +import env +import plotting + +import numpy as np +import math + + +class Node: + def __init__(self, n): + self.x = n[0] + self.y = n[1] + self.cost = 0.0 + self.parent = None + + +class RRT: + def __init__(self, xI, xG): + self.xI = Node(xI) + self.xG = Node(xG) + self.expand_len = 1 + self.goal_sample_rate = 0.05 + self.connect_dist = 10 + self.iterations = 5000 + self.node_list = [self.xI] + + self.env = env.Env() + self.plotting = plotting.Plotting(xI, xG) + + self.x_range = self.env.x_range + self.y_range = self.env.y_range + self.obs_circle = self.env.obs_circle + self.obs_rectangle = self.env.obs_rectangle + self.obs_boundary = self.env.obs_boundary + + self.path = self.planning() + self.plotting.animation(self.node_list, self.path, False) + + def planning(self): + for k in range(self.iterations): + node_rand = self.random_state() + node_near = self.nearest_neighbor(self.node_list, node_rand) + node_new = self.new_state(node_near, node_rand) + + if not self.check_collision(node_new): + neighbor_index = self.find_near_neighbor(node_new) + node_new = self.choose_parent(node_new, neighbor_index) + if node_new: + self.node_list.append(node_new) + self.rewire(node_new, neighbor_index) + + # if self.dis_to_goal(self.node_list[-1]) <= self.expand_len: + # self.new_state(self.node_list[-1], self.xG) + # return self.extract_path() + + index = self.search_best_goal_node() + self.xG.parent = self.node_list[index] + return self.extract_path() + + def random_state(self): + if np.random.random() > self.goal_sample_rate: + return Node((np.random.uniform(self.x_range[0], self.x_range[1]), + np.random.uniform(self.y_range[0], self.y_range[1]))) + return self.xG + + def nearest_neighbor(self, node_list, n): + return self.node_list[int(np.argmin([math.hypot(nd.x - n.x, nd.y - n.y) + for nd in node_list]))] + + def new_state(self, node_start, node_goal): + node_new = Node((node_start.x, node_start.y)) + dist, theta = self.get_distance_and_angle(node_new, node_goal) + dist = min(self.expand_len, dist) + + node_new.x += dist * math.cos(theta) + node_new.y += dist * math.sin(theta) + node_new.parent = node_start + + return node_new + + def find_near_neighbor(self, node_new): + n = len(self.node_list) + 1 + r = min(self.connect_dist * math.sqrt((math.log(n) / n)), self.expand_len) + + dist_table = [math.hypot(nd.x - node_new.x, nd.y - node_new.y) for nd in self.node_list] + node_index = [dist_table.index(d) for d in dist_table if d <= r] + + return node_index + + def choose_parent(self, node_new, neighbor_index): + if not neighbor_index: + return None + + cost = [] + + for i in neighbor_index: + node_near = self.node_list[i] + node_mid = self.new_state(node_near, node_new) + + if node_mid and not self.check_collision(node_mid): + cost.append(self.update_cost(node_near, node_mid)) + else: + cost.append(float("inf")) + + if min(cost) != float('inf'): + index = int(np.argmin(cost)) + neighbor_min = neighbor_index[index] + node_new = self.new_state(self.node_list[neighbor_min], node_new) + node_new.cost = min(cost) + return node_new + + return None + + def search_best_goal_node(self): + dist_to_goal_list = [self.dis_to_goal(n) for n in self.node_list] + goal_inds = [dist_to_goal_list.index(i) for i in dist_to_goal_list if i <= self.expand_len] + + return goal_inds[0] + # safe_goal_inds = [] + # for goal_ind in goal_inds: + # t_node = self.new_state(self.node_list[goal_ind], self.xG) + # if self.check_collision(t_node): + # safe_goal_inds.append(goal_ind) + # + # if not safe_goal_inds: + # print('hahhah') + # return None + # + # min_cost = min([self.node_list[i].cost for i in safe_goal_inds]) + # for i in safe_goal_inds: + # if self.node_list[i].cost == min_cost: + # self.xG.parent = self.node_list[i] + + def rewire(self, node_new, neighbor_index): + for i in neighbor_index: + node_near = self.node_list[i] + node_edge = self.new_state(node_new, node_near) + if not node_edge: + continue + + node_edge.cost = self.update_cost(node_new, node_near) + collision = self.check_collision(node_edge) + improved_cost = node_near.cost > node_edge.cost + + if not collision and improved_cost: + self.node_list[i] = node_edge + self.propagate_cost_to_leaves(node_new) + + def update_cost(self, node_start, node_end): + dist, theta = self.get_distance_and_angle(node_start, node_end) + return node_start.cost + dist + + def propagate_cost_to_leaves(self, parent_node): + for node in self.node_list: + if node.parent == parent_node: + node.cost = self.update_cost(parent_node, node) + self.propagate_cost_to_leaves(node) + + def extract_path(self): + path = [[self.xG.x, self.xG.y]] + node = self.xG + while node.parent is not None: + path.append([node.x, node.y]) + node = node.parent + path.append([node.x, node.y]) + + return path + + def dis_to_goal(self, node_cal): + return math.hypot(node_cal.x - self.xG.x, node_cal.y - self.xG.y) + + def check_collision(self, node_end): + if node_end is None: + return True + + for (ox, oy, r) in self.obs_circle: + if math.hypot(node_end.x - ox, node_end.y - oy) <= r: + return True + + for (ox, oy, w, h) in self.obs_rectangle: + if 0 <= (node_end.x - ox) <= w and 0 <= (node_end.y - oy) <= h: + return True + + for (ox, oy, w, h) in self.obs_boundary: + if 0 <= (node_end.x - ox) <= w and 0 <= (node_end.y - oy) <= h: + return True + + return False + + @staticmethod + def get_distance_and_angle(node_start, node_end): + dx = node_end.x - node_start.x + dy = node_end.y - node_start.y + return math.hypot(dx, dy), math.atan2(dy, dx) + + +if __name__ == '__main__': + x_Start = (2, 2) # Starting node + x_Goal = (49, 28) # Goal node + + rrt = RRT(x_Start, x_Goal) diff --git a/Sampling-based Planning/RRT.py b/Sampling-based Planning/RRT.py index 15db973..6052cbb 100644 --- a/Sampling-based Planning/RRT.py +++ b/Sampling-based Planning/RRT.py @@ -9,8 +9,6 @@ class Node: def __init__(self, n): self.x = n[0] self.y = n[1] - self.path_x = [] - self.path_y = [] self.parent = None @@ -18,7 +16,7 @@ class RRT: def __init__(self, xI, xG): self.xI = Node(xI) self.xG = Node(xG) - self.expand_len = 0.8 + self.expand_len = 0.4 self.goal_sample_rate = 0.05 self.iterations = 5000 self.node_list = [self.xI] @@ -28,27 +26,49 @@ class RRT: self.x_range = self.env.x_range self.y_range = self.env.y_range - self.obs_circle = self.env.obs - self.obs_rectangle = self.env.obs_boundary + self.obs_circle = self.env.obs_circle + self.obs_rectangle = self.env.obs_rectangle + self.obs_boundary = self.env.obs_boundary self.path = self.planning() self.plotting.animation(self.node_list, self.path) def planning(self): for i in range(self.iterations): - node_rand = self.generate_random_node() - node_near = self.get_nearest_node(self.node_list, node_rand) - node_new = self.new_node(node_near, node_rand, self.expand_len) + node_rand = self.random_state() + node_near = self.nearest_neighbor(self.node_list, node_rand) + node_new = self.new_state(node_near, node_rand) - if not self.check_collision(node_new, self.obs_circle, self.obs_rectangle): + if not self.check_collision(node_new): self.node_list.append(node_new) - if self.cal_dis_to_goal(self.node_list[-1]) <= self.expand_len: - self.new_node(self.node_list[-1], self.xG, self.expand_len) + if self.dis_to_goal(self.node_list[-1]) <= self.expand_len: + self.new_state(self.node_list[-1], self.xG) return self.extract_path(self.node_list) return None + def random_state(self): + if np.random.random() > self.goal_sample_rate: + return Node((np.random.uniform(self.x_range[0], self.x_range[1]), + np.random.uniform(self.y_range[0], self.y_range[1]))) + return self.xG + + def nearest_neighbor(self, node_list, n): + return self.node_list[int(np.argmin([math.hypot(nd.x - n.x, nd.y - n.y) + for nd in node_list]))] + + def new_state(self, node_start, node_end): + node_new = Node((node_start.x, node_start.y)) + dist, theta = self.get_distance_and_angle(node_new, node_end) + + dist = min(self.expand_len, dist) + node_new.x += dist * math.cos(theta) + node_new.y += dist * math.sin(theta) + node_new.parent = node_start + + return node_new + def extract_path(self, nodelist): path = [(self.xG.x, self.xG.y)] node_now = nodelist[-1] @@ -59,62 +79,36 @@ class RRT: return path - def cal_dis_to_goal(self, node_cal): + def dis_to_goal(self, node_cal): return math.hypot(node_cal.x - self.xG.x, node_cal.y - self.xG.y) - def new_node(self, node_start, node_goal, expand_len): - new_node = Node((node_start.x, node_start.y)) - d, theta = self.calc_distance_and_angle(new_node, node_goal) - - new_node.path_x = [new_node.x] - new_node.path_y = [new_node.y] - - if d < expand_len: - expand_len = d - - new_node.x += expand_len * math.cos(theta) - new_node.y += expand_len * math.sin(theta) - new_node.path_x.append(new_node.x) - new_node.path_y.append(new_node.y) - - new_node.parent = node_start - - return new_node - - def generate_random_node(self): - if np.random.random() > self.goal_sample_rate: - return Node((np.random.uniform(self.x_range[0], self.x_range[1]), - np.random.uniform(self.y_range[0], self.y_range[1]))) - return self.xG - - def get_nearest_node(self, node_list, n): - return self.node_list[int(np.argmin([math.hypot(nd.x - n.x, nd.y - n.y) - for nd in node_list]))] - - @staticmethod - def calc_distance_and_angle(from_node, to_node): - dx = to_node.x - from_node.x - dy = to_node.y - from_node.y - return math.hypot(dx, dy), math.atan2(dy, dx) - - @staticmethod - def check_collision(node_end, obs_circle, obs_rectangle): + def check_collision(self, node_end): if node_end is None: return True - for (ox, oy, r) in obs_circle: + for (ox, oy, r) in self.obs_circle: if math.hypot(node_end.x - ox, node_end.y - oy) <= r: return True - for (ox, oy, w, h) in obs_rectangle: + for (ox, oy, w, h) in self.obs_rectangle: + if 0 <= (node_end.x - ox) <= w and 0 <= (node_end.y - oy) <= h: + return True + + for (ox, oy, w, h) in self.obs_boundary: if 0 <= (node_end.x - ox) <= w and 0 <= (node_end.y - oy) <= h: return True return False + @staticmethod + def get_distance_and_angle(node_start, node_end): + dx = node_end.x - node_start.x + dy = node_end.y - node_start.y + return math.hypot(dx, dy), math.atan2(dy, dx) + if __name__ == '__main__': - x_Start = (15, 5) # Starting node - x_Goal = (45, 25) # Goal node + x_Start = (2, 2) # Starting node + x_Goal = (49, 28) # Goal node rrt = RRT(x_Start, x_Goal) diff --git a/Sampling-based Planning/__pycache__/env.cpython-37.pyc b/Sampling-based Planning/__pycache__/env.cpython-37.pyc index e743b73..1c0531a 100644 Binary files a/Sampling-based Planning/__pycache__/env.cpython-37.pyc and b/Sampling-based Planning/__pycache__/env.cpython-37.pyc differ diff --git a/Sampling-based Planning/__pycache__/env3D.cpython-37.pyc b/Sampling-based Planning/__pycache__/env3D.cpython-37.pyc new file mode 100644 index 0000000..49568f8 Binary files /dev/null and b/Sampling-based Planning/__pycache__/env3D.cpython-37.pyc differ diff --git a/Sampling-based Planning/__pycache__/plotting.cpython-37.pyc b/Sampling-based Planning/__pycache__/plotting.cpython-37.pyc index 59dd57e..efb6181 100644 Binary files a/Sampling-based Planning/__pycache__/plotting.cpython-37.pyc and b/Sampling-based Planning/__pycache__/plotting.cpython-37.pyc differ diff --git a/Sampling-based Planning/__pycache__/utils3D.cpython-37.pyc b/Sampling-based Planning/__pycache__/utils3D.cpython-37.pyc new file mode 100644 index 0000000..2541f91 Binary files /dev/null and b/Sampling-based Planning/__pycache__/utils3D.cpython-37.pyc differ diff --git a/Sampling-based Planning/env.py b/Sampling-based Planning/env.py index 85a0aba..f095a59 100644 --- a/Sampling-based Planning/env.py +++ b/Sampling-based Planning/env.py @@ -3,7 +3,8 @@ class Env: self.x_range = (0, 50) self.y_range = (0, 30) self.obs_boundary = self.obs_boundary() - self.obs = self.obs_circle() + self.obs_circle = self.obs_circle() + self.obs_rectangle = self.obs_rectangle() @staticmethod def obs_boundary(): @@ -11,15 +12,25 @@ class Env: (0, 0, 1, 30), (0, 30, 50, 1), (1, 0, 50, 1), - (50, 1, 1, 30), - (20, 1, 1, 15), - (10, 15, 10, 1), - (30, 15, 1, 15), - (40, 1, 1, 15) + (50, 1, 1, 30) + # (20, 1, 1, 15), + # (10, 15, 10, 1), + # (30, 15, 1, 15), + # (40, 1, 1, 15) ] - return obs_boundary + @staticmethod + def obs_rectangle(): + obs_rectangle = [ + (13, 10, 5, 3), + (18, 4, 5, 4), + (22, 13, 6, 3), + (33, 15, 5, 3), + (42, 6, 5, 3) + ] + return obs_rectangle + @staticmethod def obs_circle(): obs_cir = [ diff --git a/Sampling-based Planning/plotting.py b/Sampling-based Planning/plotting.py index de838bc..02b96b7 100644 --- a/Sampling-based Planning/plotting.py +++ b/Sampling-based Planning/plotting.py @@ -8,33 +8,44 @@ class Plotting: self.xI, self.xG = xI, xG self.env = env.Env() self.obs_bound = self.env.obs_boundary - self.obs_circle = self.env.obs + self.obs_circle = self.env.obs_circle + self.obs_rectangle = self.env.obs_rectangle - def animation(self, nodelist, path): + def animation(self, nodelist, path, animation=False): if path is None: print("No path found!") return - - self.plot_visited(nodelist) + self.plot_grid("RRT") + self.plot_visited(nodelist, animation) self.plot_path(path) def plot_grid(self, name): fig, ax = plt.subplots() - for x in self.obs_bound: + for (ox, oy, w, h) in self.obs_bound: ax.add_patch( patches.Rectangle( - (x[0], x[1]), x[2], x[3], + (ox, oy), w, h, edgecolor='black', facecolor='black', fill=True ) ) - for x in self.obs_circle: + for (ox, oy, w, h) in self.obs_rectangle: + ax.add_patch( + patches.Rectangle( + (ox, oy), w, h, + edgecolor='black', + facecolor='gray', + fill=True + ) + ) + + for (ox, oy, r) in self.obs_circle: ax.add_patch( patches.Circle( - (x[0], x[1]), x[2], + (ox, oy), r, edgecolor='black', facecolor='gray', fill=True @@ -47,13 +58,19 @@ class Plotting: plt.axis("equal") @staticmethod - def plot_visited(nodelist): - for node in nodelist: - if node.parent: - plt.plot(node.path_x, node.path_y, "-g") - plt.gcf().canvas.mpl_connect('key_release_event', - lambda event: [exit(0) if event.key == 'escape' else None]) - plt.pause(0.001) + def plot_visited(nodelist, animation): + if animation: + for node in nodelist: + if node.parent: + plt.plot([node.parent.x, node.x], [node.parent.y, node.y], "-g") + plt.gcf().canvas.mpl_connect('key_release_event', + lambda event: [exit(0) if event.key == 'escape' else None]) + plt.pause(0.001) + else: + for node in nodelist: + if node.parent: + plt.plot([node.parent.x, node.x], [node.parent.y, node.y], "-g") + @staticmethod def plot_path(path): diff --git a/Search-based Planning/.idea/workspace.xml b/Search-based Planning/.idea/workspace.xml index ac72159..2045d90 100644 --- a/Search-based Planning/.idea/workspace.xml +++ b/Search-based Planning/.idea/workspace.xml @@ -20,10 +20,6 @@ - - - - @@ -199,10 +195,11 @@