Motor system model with reinforcement learning drives virtual arm (Dura-Bernal et al 2017)

 Download zip file   Auto-launch 
Help downloading and running models
"We implemented a model of the motor system with the following components: dorsal premotor cortex (PMd), primary motor cortex (M1), spinal cord and musculoskeletal arm (Figure 1). PMd modulated M1 to select the target to reach, M1 excited the descending spinal cord neurons that drove the arm muscles, and received arm proprioceptive feedback (information about the arm position) via the ascending spinal cord neurons. The large-scale model of M1 consisted of 6,208 spiking Izhikevich model neurons [37] of four types: regular-firing and bursting pyramidal neurons, and fast-spiking and low-threshold-spiking interneurons. These were distributed across cortical layers 2/3, 5A, 5B and 6, with cell properties, proportions, locations, connectivity, weights and delays drawn primarily from mammalian experimental data [38], [39], and described in detail in previous work [29]. The network included 486,491 connections, with synapses modeling properties of four different receptors ..."
1 . Dura-Bernal S, Neymotin SA, Kerr CC, Sivagnanam S, Majumdar A, Francis JT, Lytton WW (2017) Evolutionary algorithm optimization of biological learning parameters in a biomimetic neuroprosthesis. IBM Journal of Research and Development (Computational Neuroscience special issue) 61(2/3):6:1-6:14
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network;
Brain Region(s)/Organism:
Cell Type(s): Abstract Izhikevich neuron;
Gap Junctions:
Receptor(s): GabaA; GabaB; NMDA; AMPA;
Transmitter(s): Glutamate; Gaba;
Simulation Environment: NEURON; Python;
Model Concept(s): Learning; Reinforcement Learning; Reward-modulated STDP; STDP; Motor control; Sensory processing;
Implementer(s): Dura-Bernal, Salvador [salvadordura at]; Kerr, Cliff [cliffk at];
Search NeuronDB for information about:  GabaA; GabaB; AMPA; NMDA; Gaba; Glutamate;
# -- Python code for interfacing the sim with a virtual arm
# Last update: 07/21/14 (

import matplotlib 
matplotlib.use('TkAgg') # for visualization
from socket import *  
import select	# for socket reading
import struct	# to pack messages for socket comms	
import numpy
from pylab import figure, show, ion, pause

# Initialize and setup the sockets and data format 
def setup():
	sendMsgFormat = "dd"  # messages contain 2 doubles = joint angles of shoulder and elbow 
	receiveMsgFormat = "dd"  # messages contain 2 doubles =  velocities
	receiveMsgSize = struct.calcsize(receiveMsgFormat)
	localHostIP =  ""#"localhost"#""# # set IP for local connection 

	print "Setting up connection..." # setup connection to model
	sockReceive = socket(AF_INET, SOCK_DGRAM) # create sockets
	hostPortReceive = 31000 # set port for receiving packets
	sockReceive.bind(('', hostPortReceive)) # bind to port
	sockReceive.setblocking(1) # Set blocking/non-blocking mode
	print ("Created UDP socket to receive packets from NEURON model; binded to port %d"%hostPortReceive)
	sockSend = socket(AF_INET, SOCK_DGRAM) 	# connect to socket for sending packets	
	hostPortSend = 32000 # set port for sending packets
	sockSend.connect((localHostIP, hostPortSend))
	sockSend.setblocking(1) # Set blocking/non-blocking mode
	print ("Created UDP socket to send packets to NEURON model; socket connected to IP %s, port %d" % (localHostIP, hostPortSend))

	return sendMsgFormat, receiveMsgFormat,  sockReceive, sockSend

# Send and receive packets to/from virtual arm
def sendAndReceivePackets(dataSend, sendMsgFormat, receiveMsgFormat,  sockReceive, sockSend):
	dataReceived = [0,0]
		receiveMsgSize = struct.calcsize(receiveMsgFormat)
		dataReceivedPacked = sockReceive.recv(receiveMsgSize) # read packet from socket	
		if len(dataReceivedPacked) == receiveMsgSize:
			dataReceived = struct.unpack(receiveMsgFormat, dataReceivedPacked)
			print "Received packet from model: (%.2f,%.2f)" % (dataReceived[0],dataReceived[1])
		print "Error receiving packet"

	inputready, outputready, e =[] ,[sockSend],[], 0.0) 	# check if other side ready to receive
	if len(outputready)>0:
			sent = sockSend.send(struct.pack(sendMsgFormat, dataSend[0], dataSend[1])) # send packet
			print "Sent packet to virtual arm: (%.2f, %.2f)" % (dataSend[0], dataSend[1])
			print "Sending socket ready but error sending packet"
		print "Sending socket not ready"	
	return dataReceived	  

# Main code for simple virtual arm
duration = 4 # sec
interval = 0.010 # time between packets (sec)
L1 = 1.0 # arm segment 1 length 
L2 = 0.8 # arm segment 2 length
shang = numpy.pi/2 # shoulder angle (rad) 
elang = numpy.pi/2 # elbow angle (rad) 
shvel = 0 # shoulder velocity (rad/s)
elvel = 0 # elbow velocity (rad/s)
friction = 0.1 # friction coefficient

sendMsgFormat, receiveMsgFormat,  sockReceive, sockSend = setup() # setup network comm

fig = figure() # create figure
ax = fig.add_subplot(111, autoscale_on=False, xlim=(-2, 2), ylim=(-2, 2)) # create subplot
line, = ax.plot([], [], 'o-', lw=2)

raw_input("Press Enter to continue...")
for i in numpy.arange(0, duration, interval):
	shang = (shang + shvel * interval) % (2*numpy.pi) # update shoulder angle
	elang = (elang + elvel * interval) % (2*numpy.pi)# update elbow angle
	if shang<0: shang = 2*numpy.pi + shang
	if elang<0: elang = 2*numpy.pi + shang
	shpos = [L1*numpy.sin(shang), L1*numpy.cos(shang)] # calculate shoulder x-y pos
	elpos = [L2*numpy.sin(shang+elang) + shpos[0], L2*numpy.cos(shang+elang) + shpos[1]] # calculate elbow x-y pos

	dataSend = [shang, elang] # set data to send
	dataReceived = sendAndReceivePackets(dataSend, sendMsgFormat, receiveMsgFormat,  sockReceive, sockSend) # send and receive data
	shvel = shvel+dataReceived[0] - (friction * shvel)# update velocities based on incoming commands (accelerations) and friction
	elvel = elvel+dataReceived[1] - (friction * elvel)

	line.set_data([0, shpos[0], elpos[0]], [0, shpos[1], elpos[1]]) # update line in figure
	ax.set_title('Time = %.1f ms, shoulder: pos=%.2f rad, vel=%.2f, acc=%.2f ; elbow: pos = %.2f rad, vel = %.2f, acc=%.2f' % (float(i)*1000, shang, shvel, dataReceived[0] - (friction * shvel), elang, elvel, dataReceived[1] - (friction * elvel) ), fontsize=10)
	pause(0.0001) # pause so that the figure refreshes at every time step

Loading data, please wait...