Public Member Functions
def	__init__
def	debug
def	output_voltage
def	plot_fz_adc
def	plot_pressure_z_adc
def	plot_rtax_vout
def	pressure2resistance
def	pressure2resistance_2
def	pressure2resistance_3
def	rtax
Public Attributes
	adc_array
	adc_bias
	adc_bits
	adc_plot
	adc_range
	contact_area
	contact_area_percent
	begin: parameters to be specified
	fz_array
	fz_max
	fz_min
	no_contact_area
	pressure_array
	pressure_max
	r1
	r_min
	r_min_percent_of_r_no_contact
	r_no_contact
	rtax_array
	rtax_max
	taxel_area
	vdigi_array
	vdigi_max
	volts_per_adc_unit
	vtot

Detailed Description

Attempts to model the digital signal that results from a normal
force applied to a taxel. It assumes that the force is uniformly
distributed over an area. The contact area is specified as a
percentage of the taxel area.

Definition at line 25 of file tactile_sensor_model.py.

Constructor & Destructor Documentation

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.__init__	(	self,
		contact_area_percent = `50.0`
	)

Definition at line 30 of file tactile_sensor_model.py.

Member Function Documentation

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.debug ( self )

print out many of the key member variables of the TaxelModel object

Definition at line 251 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.output_voltage	(	self,
		r
	)

given the resistance for the entire taxel, this returns the
voltage across the taxel, which is what the analog to digital
converter reads

Definition at line 88 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.plot_fz_adc ( self )

plot the curve relating applied normal force to the analog
to digital converter output. this corresponds with the
empirically generated scatter plots from a real taxel

Definition at line 225 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.plot_pressure_z_adc ( self )

plot the curve relating applied normal force to the analog
to digital converter output. this corresponds with the
empirically generated scatter plots from a real taxel

Definition at line 233 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.plot_rtax_vout ( self )

plot the curve relating the total taxel resistance and the
voltage across the taxel, which corresponds to the voltage
converted to a digital signal

Definition at line 242 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.pressure2resistance	(	self,
		p
	)

given an applied pressure, returns the resistivity of the
contacted region of the taxel. this uses a simple linear
model, where: 
     0 Pascals   -> r_no_contact
     pressure max -> r_min Ohms

Definition at line 95 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.pressure2resistance_2	(	self,
		p
	)

given an applied pressure, returns the resistivity of the
contacted region of the taxel. this uses a logistic model,
where:
     0 Pascals   -> r_no_contact
     pressure max -> r_min Ohms

Definition at line 112 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.pressure2resistance_3	(	self,
		p
	)

given an applied pressure, returns the resistivity of the
contacted region of the taxel. this was a quick attempt to use
a model similar to "The Working Principle of Resistive Tactile
Sensor Cells" by Karsten Weiss and Heinz Worn. It doesn't
work, yet?

Definition at line 131 of file tactile_sensor_model.py.

def hrl_fabric_based_tactile_sensor.tactile_sensor_model.TaxelModel.rtax	(	self,
		p
	)

given the pressure uniformly applied across the contact
area this returns the resistance for the entire taxel. 

it essentially models the taxel as parallel resistors, where
resistors in the contact area have a resistance dependent on
the applied pressure, and resistors in the non-contact area
have the maximum resistance. i started with a discrete model,
and then made a continuous approximation, which appears to
correspond with using two volumes in parallel with different
resistivities.

based on wikipedia, it looks like i've been using something
called resistivity (rho). so, this model can use the equation
r = rho*(length/area). if we assume length is constant, then
this leads to r_contact = rho_contact/area_contact and the
same for not_contact. assuming that they are in parallel. then
R_total = 1/((area_contact/rho_contact) +
(area_non_contact/rho_non_contact)). if we assume length
changes, then we can make rho_contact = resistivity_contact *
length_contact.

this should be more carefully investigated, but
it seems right...

the biggest unknown is the function that converts pressure to
resistivity. there are currently three models of this function
in this code. fitting a parametric model to the data would be
a good next step. probably use an optimizer like Nelder-Mead
that only requires function evaluations and use a cost
function that compares the fz->adc mapping to empirically
collected data.

Definition at line 175 of file tactile_sensor_model.py.