def create_params():
p = DotMap()
p.reachability_map = ReachabilityMap
# Map operation
# Obstacle map resolution=0.05m. To shorten the computation time, we down sample the map by the following step size
p.sample_step_x_clipped = 2
p.sample_step_y_clipped = 2
p.sample_step_x_whole = 1
p.sample_step_y_whole = 1
# In a navigation task, we assume the robot successfully reaches the goal if it is within 0.3m of the goal point.
# However, in the TTR computation, in order to lead the robot to the goal more accurately, we set a smaller
# goal_cutoff_dist
p.goal_cutoff_dist = 0.1
p.obstacle_augment_dist = 0.15
# When computing reach-avoid TTR, the whole map is too large. Thus we first clip the map around the start and the
# goal position of each episode. The clipped map augment the boundary of the start-goal rectangle by clip_extension
p.clip_extension = np.float32(2.0)
# System dynamics (state constraints and discretization)
p.theta_dim = int(30)
p.v_dim = int(9)
p.v_dv = float(.1)
p.wMax = float(1.1)
p.aMax = float(.4)
p.v_high = float(.7)
p.v_low = float(-.1)
# System dynamics (disturbances constraints)
# Additive disturbances
# Disturbances used for modeling prediction errors
p.dMax_xy = float(0.05)
p.dMax_theta = float(0.15)
p.dMax_avoid_xy = float(0.05)
p.dMax_avoid_theta = float(0.15)
# # No disturbance version
# p.dMax_xy = float(0.0)
# p.dMax_theta = float(0.0)
# p.dMax_avoid_xy = float(0.0)
# p.dMax_avoid_theta = float(0.0)
# Set Matlab path
p.MATLAB_PATH = '/home/anjianl/Desktop/project/WayPtNav/reachability'
# Set data path based on threads
p.thread = 'v1'
return p
