class NeuralSimulation(object): """NeuralSimulation class Responsible for initialization of all children objects such as the Open Dynamics Engine environment and OpenGL viewer. Starts the main event loop with real-time constraints. Attributes: env: The Open Dynamics Engine environment. viewer: The OpenGL viewer for the ODE environment. Methods: start: Begins the main event loop. """ env = None viewer = None rnn = None kinematics = None def __init__(self, randomize=False, rnn_xml=None, verbose=False): """Initialize Creates the environment and viewer objects required to run the neural network PA10 robotic arm simulation. Arguments: randomize: Determines if the test article gates will be randomized. (Default: False) rnn_xml: A XML filename containing neural network parameters. If None, a new neural network will be trained until convergence. (Default: None) verbose: Determines the level out debug output generated. (Default: False) """ # Generate the XODE file XODE_FILENAME = 'model' # .xode is appended automatically print('>>> Generating world model') if os.path.exists('./' + XODE_FILENAME + '.xode'): os.remove('./' + XODE_FILENAME + '.xode') xode_model = NeuralSimWorld(name=XODE_FILENAME, randomize_test_article=randomize) xode_model.generate() print('>>> Generating RNN') # Determine if we need to train the neural network if rnn_xml is not None: print('>>> Loading RNN from file') self.rnn = network.PathPlanningNetwork() self.rnn.load_network_from_file(rnn_xml) else: print('>>> Training new RNN') self.rnn = network.train_path_planning_network() self.rnn.save_network_to_file(constants.G_RNN_XML_OUT) print('>>> Starting kinematics engine') self.kinematics = PA10Kinematics() # Start environment print('>>> Starting environment') self.env = EnvironmentInterface( xode_filename='./' + XODE_FILENAME + '.xode', realtime=False, verbose=verbose, gravity=constants.G_ENVIRONMENT_GRAVITY) # Start viewer print('>>> Starting viewer') self.viewer = ViewerInterface(verbose=verbose) self.viewer.start() # Set up all grouped bodies in the environment self.env.groups = { 'pointer': ['tooltip', 'stick'], } return def start(self, fps, fast_step=False): """Start Begin the continuous event loop for the simulation. This event loop can be exited using the ctrl+c keyboard interrupt. Real-time constraints are enforced. [Hz] Arguments: fps: The value of frames per second of the simulation. fast_step: If True, the ODE fast step algorithm will be used. This is faster and requires less memory but is less accurate. (Default: False) """ paused = False stopped = False # Define the total time for the tooltip traversal t_total = 20.0 # Define the simulation frame rate t = 0.0 # [s] dt = 1.0 / float(fps) # [s] # Keep track of time overshoot in the case that simulation time must be # increased in order to maintain real-time constraints t_overshoot = 0.0 # Get the initial path position (center of gate7) pos_start = self.env.get_body_pos('gate7') # [m] # Get the first position of the PA10 at rest pos_init = self.env.get_body_pos('tooltip') # [m] # Calculate the new required joint angles of the PA10 #pa10_joint_angles = self.kinematics.calc_inverse_kinematics(pos_init, pos_start) # TODO: Move the PA10 end-effector to the starting position along the path # TODO: TEMP - Move the temporary end-effector pointer to the starting position self.env.set_group_pos('pointer', pos_start) # Generate long-term path from initial position t_input = np.linspace(start=0.0, stop=1.0, num=t_total / dt) t_input = np.reshape(t_input, (len(t_input), 1)) rnn_path = self.rnn.extrapolate(t_input, [pos_start], len(t_input) - 1) # Add the initial condition point back onto the data rnn_path = np.vstack((pos_start, rnn_path)) # Retrieve one set of standard gate position/orientation data file_path = pathutils.list_data_files(constants.G_TRAINING_DATA_DIR)[0] gate_data = datastore.retrieve(file_path) gate_start_idx = constants.G_GATE_IDX gate_end_idx = gate_start_idx + constants.G_NUM_GATE_INPUTS # Reshape the gate positions data gate_data = gate_data[0:1, gate_start_idx:gate_end_idx] gate_data = np.tile(gate_data, (len(rnn_path), 1)) # Complete the rnn path data rnn_path = np.hstack((t_input, gate_data, rnn_path)) # Save generated path for later examination datastore.store(rnn_path, constants.G_RNN_STATIC_PATH_OUT) # Define a variable to hold the final path (with real-time correction) final_path = rnn_path[:-1].copy() path_saved = False # Detect all path segments between gates in the generated path segments = pathutils._detect_segments(rnn_path) path_idx = 0 x_path_offset = np.array([0.0, 0.0, 0.0]) # [m] v_curr = np.array([0.0, 0.0, 0.0]) # [m/s] a_max = constants.G_MAX_ACCEL # [m/s^2] # Get the static table position x_table = self.env.get_body_pos('table') while not stopped: t_start = time.time() # If the last calculation took too long, catch up dt_warped = dt + t_overshoot self.env.set_dt(dt_warped) # Determine if the viewer is stopped. Then we can quit if self.viewer.is_dead: break # Pause the simulation if we are at the end if path_idx == len(rnn_path) - 1 or paused: self.env.step(paused=True, fast=fast_step) # If we have really hit the end of the simulation, save/plot the path if not paused and not path_saved: # Save the final data to a file datastore.store(final_path, constants.G_RNN_DYNAMIC_PATH_OUT) path_saved = True continue # Not a very elegant solution to pausing at the start, but it works if t <= 1000.0: self.env.step(paused=True, fast=fast_step) t += dt_warped continue # Determine the current path segment curr_segment_idx = 0 for segment_idx, segment_end in enumerate(segments): if path_idx <= segment_end: curr_segment_idx = segment_idx break x_curr = pathutils.get_path_tooltip_pos(rnn_path, path_idx) + x_path_offset x_next = pathutils.get_path_tooltip_pos( rnn_path, path_idx + 1) + x_path_offset # Get the expected gate position x_gate_expected = pathutils.get_path_gate_pos( rnn_path, segments[curr_segment_idx], curr_segment_idx) # Get the actual gate position x_gate_actual = self.env.get_body_pos('gate%d' % curr_segment_idx) # Calculate the new position from change to new gate position dx_gate = x_gate_actual - (x_gate_expected + x_path_offset) x_new = x_next + dx_gate # Calculate the new velocity v_new = (x_new - x_curr) / dt_warped # Calculate the new acceleration a_new = (v_new - v_curr) / dt_warped # Calculate the acceleration vector norm a_new_norm = np.linalg.norm(a_new) # Limit the norm vector a_new_norm_clipped = np.clip(a_new_norm, -a_max, a_max) # Determine the ratio of the clipped norm, protect against divide by zero if a_new_norm != 0.0: ratio_unclipped = a_new_norm_clipped / a_new_norm else: ratio_unclipped = 0.0 # Scale the acceleration vector by this ratio a_new = a_new * ratio_unclipped # Calculate the new change in velocity dv_new = a_new * dt_warped v_new = v_curr + dv_new # Calculate the new change in position dx_new = v_new * dt_warped x_new = x_curr + dx_new # Modify final path data with current tooltip and gate positions pathutils.set_path_time(final_path, path_idx, t) pathutils.set_path_tooltip_pos(final_path, path_idx, x_curr) for gate_idx in range(constants.G_NUM_GATES): gate_name = 'gate%d' % gate_idx x_gate = self.env.get_body_pos(gate_name) pathutils.set_path_gate_pos(final_path, path_idx, gate_idx, x_gate) # Store this velocity for the next time step v_curr = v_new # Recalculate the current offset x_path_offset += x_new - x_next # Perform inverse kinematics to get joint angles pa10_joint_angles = self.kinematics.calc_inverse_kinematics( x_curr, x_new) # TODO: TEMP - MOVE ONLY POINTER, NO PA10 self.env.set_group_pos('pointer', x_new) if constants.G_TABLE_IS_OSCILLATING: # Move the table with y-axis oscillation x_table_next = shaker_table(t, x_table) else: x_table_next = x_table self.env.set_body_pos('table', x_table_next) # Step through the world by 1 time frame and actuate pa10 joints self.env.performAction(pa10_joint_angles, fast=fast_step) # Update current time after this step t += dt_warped path_idx += 1 # Determine the difference in virtual vs actual time t_warped = dt - (time.time() - t_start) # Attempt to enforce real-time constraints if t_warped >= 0.0: # The calculation took less time than the virtual time. Sleep # the rest off time.sleep(t_warped) t_overshoot = 0.0 else: # The calculation took more time than the virtual time. We need # to catch up with the virtual time on the next time step t_overshoot = -t_warped return def __del__(self): """Delete (del) Attempts to shut down any active engines and kills off spawned class objects. """ # Kill the OpenGL viewer process if self.viewer is not None: self.viewer.stop() del self.viewer # Kill the ODE environment objects if self.env is not None: self.env.stop() del self.env return
class TrainingSimulation(object): """TrainingSimulation class Responsible for initialization of all children objects such as the Open Dynamics Engine environment, Phantom Omni controller, and OpenGL viewer. Starts the main event loop with real-time constraints. Attributes: env: The Open Dynamics Engine environment. omni: The Phantom Omni robotic controller connection. viewer: The OpenGL viewer for the ODE environment. saved_data: A list of tuples containing data from the simulation. Methods: start: Begins the main event loop. """ env = None omni = None viewer = None saved_data = None def __init__(self, randomize=False, network=False, verbose=False): """Initialize Creates the environment, viewer, and (Phantom Omni) controller objects required to run the human data capture simulation. Arguments: randomize: Determines if the test article gates will be randomized. (Default: False) network: Determines if the omni connection is local or networked. (Default: False) verbose: Determines the level out debug output generated. (Default: False) """ # Generate the XODE file XODE_FILENAME = 'model' # .xode is appended automatically print('>>> Generating world model') if os.path.exists('./'+XODE_FILENAME+'.xode'): os.remove('./'+XODE_FILENAME+'.xode') xode_model = TrainingSimWorld( name=XODE_FILENAME, randomize_test_article=randomize ) xode_model.generate() # Start environment print('>>> Starting environment') self.env = EnvironmentInterface( xode_filename='./'+XODE_FILENAME+'.xode', realtime=False, verbose=verbose, gravity=constants.G_ENVIRONMENT_GRAVITY ) # Set up all grouped bodies in the environment self.env.groups = { 'pointer': ['tooltip', 'stick'], } # Start viewer print('>>> Starting viewer') self.viewer = ViewerInterface(verbose=verbose) self.viewer.start() # Start controller print('>>> Starting Phantom Omni interface') self.omni = PhantomOmniInterface() if network: ip = raw_input('<<< Enter host ip: ') port = int(raw_input('<<< Enter tcp port: ')) else: ip = constants.G_IP_LOCAL_DEFAULT port = constants.G_PORT_DEFAULT # Try to connect to the Phantom Omni controller self.omni.connect(ip, port) self.saved_data = np.array([]) return def start(self, fps): """Start Begin the continuous event loop for the simulation. This event loop can be exited using the ctrl+c keyboard interrupt. Real-time constraints are enforced. Arguments: fps: The value of frames per second of the simulation. """ paused = False stopped = False # Define the simulation frame rate t = 0.0 dt = 1.0 / float(fps) # [s] # Keep track of time overshoot in the case that simulation time must be # increased in order to maintain real-time constraints t_overshoot = 0.0 while not stopped: t_start = time.time() # If the last calculation took too long, catch up dt_warped = dt + t_overshoot self.env.set_dt(dt_warped) self.omni.set_dt(dt_warped) # Determine if the viewer is stopped. Then we can quit if self.viewer.is_dead: break if paused: self.env.step(paused=True) continue # Populate the controller with the most up-to-date data self.omni.update() # Determine the input data to record sample_input = np.array([ t, ]).flatten() # Capture the gate position at each time step for gate_idx in range(constants.G_NUM_GATES): gate_pos = np.array(self.env.get_body_pos('gate%d'%gate_idx)).flatten() gate_rot = constants.G_GATE_NORM_ROT[gate_idx] sample_input = np.hstack((sample_input, gate_pos, gate_rot)) # Determine the output data to record sample_output = np.array([ self.env.get_body_pos('tooltip'), ]).flatten() # Join the sample input/output data_sample = np.hstack((sample_input, sample_output)) # Save the data self.save_data(data_sample) # Get the updated linear/angular velocities of the tooltip linear_vel = self.omni.get_linear_vel() angular_vel = self.omni.get_angular_vel() # Set the linear and angular velocities of the simulation self.env.set_group_linear_vel('pointer', linear_vel) self.env.set_group_angular_vel('pointer', angular_vel) # Step through the world by 1 time frame self.env.step() t += dt_warped # Determine the difference in virtual vs actual time t_warped = dt - (time.time() - t_start) # Attempt to enforce real-time constraints if t_warped >= 0.0: # The calculation took less time than the virtual time. Sleep # the rest off time.sleep(t_warped) t_overshoot = 0.0 else: # The calculation took more time than the virtual time. We need # to catch up with the virtual time on the next time step t_overshoot = -t_warped return def save_data(self, data_sample): """Save Data Saves a sample of data into the simulation's saved data array. Arguments: data_sample: A sample of input/output data from a single time step. """ if len(self.saved_data): self.saved_data = np.vstack((self.saved_data, data_sample)) else: self.saved_data = data_sample.copy() return def __del__(self): """Delete (del) Attempts to shut down any active engines and kills off spawned class objects. """ # Kill the OpenGL viewer process if self.viewer is not None: self.viewer.stop() del self.viewer # Kill the ODE environment objects if self.env is not None: self.env.stop() del self.env # Kill the Phantom Omni controller process if self.omni is not None: self.omni.disconnect() del self.omni return
class NeuralSimulation(object): """NeuralSimulation class Responsible for initialization of all children objects such as the Open Dynamics Engine environment and OpenGL viewer. Starts the main event loop with real-time constraints. Attributes: env: The Open Dynamics Engine environment. viewer: The OpenGL viewer for the ODE environment. Methods: start: Begins the main event loop. """ env = None viewer = None rnn = None kinematics = None def __init__(self, randomize=False, rnn_xml=None, verbose=False): """Initialize Creates the environment and viewer objects required to run the neural network PA10 robotic arm simulation. Arguments: randomize: Determines if the test article gates will be randomized. (Default: False) rnn_xml: A XML filename containing neural network parameters. If None, a new neural network will be trained until convergence. (Default: None) verbose: Determines the level out debug output generated. (Default: False) """ # Generate the XODE file XODE_FILENAME = 'model' # .xode is appended automatically print('>>> Generating world model') if os.path.exists('./'+XODE_FILENAME+'.xode'): os.remove('./'+XODE_FILENAME+'.xode') xode_model = NeuralSimWorld( name=XODE_FILENAME, randomize_test_article=randomize ) xode_model.generate() print('>>> Generating RNN') # Determine if we need to train the neural network if rnn_xml is not None: print('>>> Loading RNN from file') self.rnn = network.PathPlanningNetwork() self.rnn.load_network_from_file(rnn_xml) else: print('>>> Training new RNN') self.rnn = network.train_path_planning_network() self.rnn.save_network_to_file(constants.G_RNN_XML_OUT) print('>>> Starting kinematics engine') self.kinematics = PA10Kinematics() # Start environment print('>>> Starting environment') self.env = EnvironmentInterface( xode_filename='./'+XODE_FILENAME+'.xode', realtime=False, verbose=verbose, gravity=constants.G_ENVIRONMENT_GRAVITY ) # Start viewer print('>>> Starting viewer') self.viewer = ViewerInterface(verbose=verbose) self.viewer.start() # Set up all grouped bodies in the environment self.env.groups = { 'pointer': ['tooltip', 'stick'], } return def start(self, fps, fast_step=False): """Start Begin the continuous event loop for the simulation. This event loop can be exited using the ctrl+c keyboard interrupt. Real-time constraints are enforced. [Hz] Arguments: fps: The value of frames per second of the simulation. fast_step: If True, the ODE fast step algorithm will be used. This is faster and requires less memory but is less accurate. (Default: False) """ paused = False stopped = False # Define the total time for the tooltip traversal t_total = 20.0 # Define the simulation frame rate t = 0.0 # [s] dt = 1.0 / float(fps) # [s] # Keep track of time overshoot in the case that simulation time must be # increased in order to maintain real-time constraints t_overshoot = 0.0 # Get the initial path position (center of gate7) pos_start = self.env.get_body_pos('gate7') # [m] # Get the first position of the PA10 at rest pos_init = self.env.get_body_pos('tooltip') # [m] # Calculate the new required joint angles of the PA10 #pa10_joint_angles = self.kinematics.calc_inverse_kinematics(pos_init, pos_start) # TODO: Move the PA10 end-effector to the starting position along the path # TODO: TEMP - Move the temporary end-effector pointer to the starting position self.env.set_group_pos('pointer', pos_start) # Generate long-term path from initial position t_input = np.linspace(start=0.0, stop=1.0, num=t_total/dt) t_input = np.reshape(t_input, (len(t_input), 1)) rnn_path = self.rnn.extrapolate(t_input, [pos_start], len(t_input)-1) # Add the initial condition point back onto the data rnn_path = np.vstack((pos_start, rnn_path)) # Retrieve one set of standard gate position/orientation data file_path = pathutils.list_data_files(constants.G_TRAINING_DATA_DIR)[0] gate_data = datastore.retrieve(file_path) gate_start_idx = constants.G_GATE_IDX gate_end_idx = gate_start_idx + constants.G_NUM_GATE_INPUTS # Reshape the gate positions data gate_data = gate_data[0:1,gate_start_idx:gate_end_idx] gate_data = np.tile(gate_data, (len(rnn_path), 1)) # Complete the rnn path data rnn_path = np.hstack((t_input, gate_data, rnn_path)) # Save generated path for later examination datastore.store(rnn_path, constants.G_RNN_STATIC_PATH_OUT) # Define a variable to hold the final path (with real-time correction) final_path = rnn_path[:-1].copy() path_saved = False # Detect all path segments between gates in the generated path segments = pathutils._detect_segments(rnn_path) path_idx = 0 x_path_offset = np.array([0.0, 0.0, 0.0]) # [m] v_curr = np.array([0.0, 0.0, 0.0]) # [m/s] a_max = constants.G_MAX_ACCEL # [m/s^2] # Get the static table position x_table = self.env.get_body_pos('table') while not stopped: t_start = time.time() # If the last calculation took too long, catch up dt_warped = dt + t_overshoot self.env.set_dt(dt_warped) # Determine if the viewer is stopped. Then we can quit if self.viewer.is_dead: break # Pause the simulation if we are at the end if path_idx == len(rnn_path) - 1 or paused: self.env.step(paused=True, fast=fast_step) # If we have really hit the end of the simulation, save/plot the path if not paused and not path_saved: # Save the final data to a file datastore.store(final_path, constants.G_RNN_DYNAMIC_PATH_OUT) path_saved = True continue # Not a very elegant solution to pausing at the start, but it works if t <= 1000.0: self.env.step(paused=True, fast=fast_step) t += dt_warped continue # Determine the current path segment curr_segment_idx = 0 for segment_idx, segment_end in enumerate(segments): if path_idx <= segment_end: curr_segment_idx = segment_idx break x_curr = pathutils.get_path_tooltip_pos(rnn_path, path_idx) + x_path_offset x_next = pathutils.get_path_tooltip_pos(rnn_path, path_idx+1) + x_path_offset # Get the expected gate position x_gate_expected = pathutils.get_path_gate_pos( rnn_path, segments[curr_segment_idx], curr_segment_idx ) # Get the actual gate position x_gate_actual = self.env.get_body_pos('gate%d'%curr_segment_idx) # Calculate the new position from change to new gate position dx_gate = x_gate_actual - (x_gate_expected + x_path_offset) x_new = x_next + dx_gate # Calculate the new velocity v_new = (x_new - x_curr) / dt_warped # Calculate the new acceleration a_new = (v_new - v_curr) / dt_warped # Calculate the acceleration vector norm a_new_norm = np.linalg.norm(a_new) # Limit the norm vector a_new_norm_clipped = np.clip(a_new_norm, -a_max, a_max) # Determine the ratio of the clipped norm, protect against divide by zero if a_new_norm != 0.0: ratio_unclipped = a_new_norm_clipped / a_new_norm else: ratio_unclipped = 0.0 # Scale the acceleration vector by this ratio a_new = a_new * ratio_unclipped # Calculate the new change in velocity dv_new = a_new * dt_warped v_new = v_curr + dv_new # Calculate the new change in position dx_new = v_new * dt_warped x_new = x_curr + dx_new # Modify final path data with current tooltip and gate positions pathutils.set_path_time(final_path, path_idx, t) pathutils.set_path_tooltip_pos(final_path, path_idx, x_curr) for gate_idx in range(constants.G_NUM_GATES): gate_name = 'gate%d' % gate_idx x_gate = self.env.get_body_pos(gate_name) pathutils.set_path_gate_pos(final_path, path_idx, gate_idx, x_gate) # Store this velocity for the next time step v_curr = v_new # Recalculate the current offset x_path_offset += x_new - x_next # Perform inverse kinematics to get joint angles pa10_joint_angles = self.kinematics.calc_inverse_kinematics(x_curr, x_new) # TODO: TEMP - MOVE ONLY POINTER, NO PA10 self.env.set_group_pos('pointer', x_new) if constants.G_TABLE_IS_OSCILLATING: # Move the table with y-axis oscillation x_table_next = shaker_table(t, x_table) else: x_table_next = x_table self.env.set_body_pos('table', x_table_next) # Step through the world by 1 time frame and actuate pa10 joints self.env.performAction(pa10_joint_angles, fast=fast_step) # Update current time after this step t += dt_warped path_idx += 1 # Determine the difference in virtual vs actual time t_warped = dt - (time.time() - t_start) # Attempt to enforce real-time constraints if t_warped >= 0.0: # The calculation took less time than the virtual time. Sleep # the rest off time.sleep(t_warped) t_overshoot = 0.0 else: # The calculation took more time than the virtual time. We need # to catch up with the virtual time on the next time step t_overshoot = -t_warped return def __del__(self): """Delete (del) Attempts to shut down any active engines and kills off spawned class objects. """ # Kill the OpenGL viewer process if self.viewer is not None: self.viewer.stop() del self.viewer # Kill the ODE environment objects if self.env is not None: self.env.stop() del self.env return