Datasets:
jblitzar
/
github-python

Dataset card Data Studio Files
xet
 Community
text
stringlengths
0
93.6k
left_mm = left_diff / TICK_PER_MM;
right_mm = right_diff / TICK_PER_MM;
distance = (left_mm + right_mm) / 2.0;
dAngle = (right_mm - left_mm) / WHEEL_SPAN
dAngle *= ANGULAR_ERROR
self.thrPose += dAngle
if self.thrPose > 100*math.pi:
self.thrPose -= 101*math.pi
if self.thrPose < -100*math.pi:
self.thrPose += 101*math.pi
self.xPose += distance * math.cos(self.thrPose)
self.yPose += distance * math.sin(self.thrPose)
self.leftEncoder_old = self.leftEncoder
self.rightEncoder_old = self.rightEncoder
def _integrateNextOdometricStepCreate(self, distance, rawAngle):
""" integrateNextOdometricStep adds the reported inputs
distance in mm
rawAngle in degrees
to the estimate of the robot's global pose
"""
# OK, so this _should_ be easy
# distance is, supposedly, the arc length that the center
# of the robot has traveled (the average of
# the two wheel's linear distances)
#
# rawAngle is, supposedly, the (RightWheel-LeftWheel)/2.0
#
# the distance (diameter) between the two wheels is 258mm
# keep in mind that the robot's physical diameter is larger ~
#
# 0.5*258 == 129mm radius
#
# perhaps there's nothing to do...
if distance == 0 and rawAngle == 0:
return
print rawAngle
# then again, maybe there is something to do...
dthr = math.radians(rawAngle) # angle traveled
d = distance # distance traveled
# compute offsets in the local coordinate system,
# with the x-axis pointing in the direction the robot was
# initially facing (for this move) and the y-axis pointing
# perpendicularly to the left (when facing forward)
#
# first, the special case when the angle is zero...
if rawAngle == 0:
dx = float(d)
dy = 0.0
# or if the distance is zero...
elif distance == 0:
dx = 0.0
dy = 0.0
# or when neither is zero...
else:
# finite radius of curvature
ROC = float(d)/dthr # remember, this is signed!
dx = ROC*math.sin(dthr) # because ROC is signed,
dy = ROC-ROC*math.cos(dthr) # we don't need two cases
#
# we need to add dx, dy, and dthr to the global pose
# and so we need to do so in the global direction in
# which the robot was facing at the start of this movement
#
# here is the unit vector describing that direction
unitForwardX = math.cos( self.thrPose )
unitForwardY = math.sin( self.thrPose )
# here is the unit vector perpendicular to the left
unitPerpX = math.cos( self.thrPose + math.pi/2.0 )
unitPerpY = math.sin( self.thrPose + math.pi/2.0 )
# now we compute our global offsets
dx_global = dx*unitForwardX + dy*unitPerpX
dy_global = dx*unitForwardY + dy*unitPerpY
##print 'distance and rawAngle', distance, rawAngle
##print 'local offsets, x, y, thd', dx, dy, math.degrees(dthr)
##print 'global offsets, x, y, thd', dx_global, dy_global, math.degrees(dthr)
# and we add them all in...
self.xPose += dx_global
self.yPose += dy_global
self.thrPose += dthr
#print 'final pose', self.xPose, self.yPose, self.thrPose
return
def setWheelVelocities( self, left_cm_sec, right_cm_sec ):
""" sends velocities of each wheel independently
left_cm_sec: left wheel velocity in cm/sec (capped at +- 50)
right_cm_sec: right wheel velocity in cm/sec (capped at +- 50)
"""
if left_cm_sec < -50: left_cm_sec = -50;
if left_cm_sec > 50: left_cm_sec = 50;
if right_cm_sec < -50: right_cm_sec = -50;
if right_cm_sec > 50: right_cm_sec = 50;
# convert to mm/sec, ensure we have integers
leftHighVal, leftLowVal = _toTwosComplement2Bytes( int(left_cm_sec*10) )