Classifications

• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10N—ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10N30/00—Piezoelectric or electrostrictive devices
  • H10N30/01—Manufacture or treatment
  • H10N30/09—Forming piezoelectric or electrostrictive materials
  • H10N30/098—Forming organic materials
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
  • G01R27/00—Arrangements for measuring resistance, reactance, impedance, or electric characteristics derived therefrom
  • G01R27/02—Measuring real or complex resistance, reactance, impedance, or other two-pole characteristics derived therefrom, e.g. time constant
  • G01R27/26—Measuring inductance or capacitance; Measuring quality factor, e.g. by using the resonance method; Measuring loss factor; Measuring dielectric constants ; Measuring impedance or related variables
  • G01R27/2617—Measuring dielectric properties, e.g. constants
• • H—ELECTRICITY
  • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
  • H02N—ELECTRIC MACHINES NOT OTHERWISE PROVIDED FOR
  • H02N2/00—Electric machines in general using piezoelectric effect, electrostriction or magnetostriction
  • H02N2/02—Electric machines in general using piezoelectric effect, electrostriction or magnetostriction producing linear motion, e.g. actuators; Linear positioners ; Linear motors
  • H02N2/06—Drive circuits; Control arrangements or methods
  • H02N2/062—Small signal circuits; Means for controlling position or derived quantities, e.g. for removing hysteresis
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10N—ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10N30/00—Piezoelectric or electrostrictive devices
  • H10N30/20—Piezoelectric or electrostrictive devices with electrical input and mechanical output, e.g. functioning as actuators or vibrators
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10N—ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10N30/00—Piezoelectric or electrostrictive devices
  • H10N30/80—Constructional details
  • H10N30/802—Circuitry or processes for operating piezoelectric or electrostrictive devices not otherwise provided for, e.g. drive circuits

Definitions

• This invention relates to a method and apparatus for sensing electrical parameters of a dielectric elastomer (DE). More particularly, the invention relates to self-sensing by dynamically estimating the electrical parameters of the DE and deriving an estimate of the state of the DE from those parameters.
• DE dielectric elastomer
• a dielectric elastomer can be used to convert electrical energy into mechanical energy, or vice versa.
• DEs can be used as actuators, generators, and/or sensors.
• a DE typically comprises a volumetrically incompressible dielectric elastomer membrane 11 provided between compliant electrodes 12.
• Dielectric Elastomer Actuators deform when a voltage difference is applied across the electrodes 12, as disclosed in "Electrostriction of polymer dielectrics with compliant electrodes as a means of actuation" (Pelrine R.E., ornbluh R.D., & Joseph J. P., Sensors and Actuators, A: Physical 64(1), pp 77-85 (1998)), for example. Electrical charge accumulates on the electrodes 12 in the manner of a capacitor and an electrostatic pressure is generated that results in a through-thickness compression and in-plane expansion of the membrane, from the rest or contracted state as shown by example in Figure 1 (a), to the actuated or expanded state illustrated in Figure 1 (b).
• DEA Dielectric Elastomer Actuators
• the magnitude of the pressure is defined by the equation below, where P is the pressure, ⁇ ⁇ is the relative permittivity of the dielectric material, ⁇ 0 is the permittivity of free space (8.854 x 10 "12 F/m), V is the voltage, and d is the dielectric membrane thickness.
• a dielectric elastomer generator converts mechanical energy into electrical energy.
• the DEG is a variable capacitor, and its capacitance changes with mechanical strain (i.e. deformation of the membrane).
• the DEG generates electrical energy by increasing the electric potential energy stored in it.
• the steps to achieve this are illustrated diagrammatically in Figure 2.
• mechanical energy 20 is initially applied to the DEG 21 by stretching it. This results in a planar expansion of electrodes 22 and an orthogonal compression of the membrane 23, leading to an increased capacitance.
• Electrical energy 24 is then input to the DEG by charging (or priming) from an electric power source (not shown) so that opposing electrodes 22 become oppositely charged. Relaxing the DEG will convert the mechanical energy into electrical energy by forcing apart the opposite charges (+ and -) on opposing electrodes 22, and forcing the like charges on each electrode 22 closer together due to the planar contraction thereof.
• the electrical energy 24 is extracted and the cycle repeats.
• Dielectric elastomers may alternatively, or additionally, be used as a sensor.
• the electrical properties of dielectric elastomer sensors such as capacitance, electrode resistance, and the equivalent parallel resistance of the dielectric membrane change in response to external stimuli such as mechanical deformation or changes in the DE's operating environment, for example. Changes in the electrical parameters of the DE can therefore be used to acquire feedback information.
• the leakage current is negligible, or at low actuation speeds.
• the methods and/or apparatus of the prior art obtain estimates based upon other assumptions which may not always hold true, are not suitable for practical implementation in a system designed for portable use, are sensitive to noise, and/or require unnecessarily complex, numerous, or processor-intensive calculations to be performed.
• the invention may broadly be said to consist in a method for estimating the state of a dielectric elastomer (DE), comprising the steps of:
• the feedback parameters comprise capacitance of the DE, wherein the capacitance is approximated by a first coefficient of the plane equation.
• the first coefficient is the coefficient of a voltage variable of the plane equation.
• the feedback parameters further comprise leakage current through the DE, wherein the leakage current is derived from a second coefficient of the plane equation.
• the second coefficient is the coefficient of a time variable of the plane equation.
• the period of time over which the potential difference across the DE and series current through the DE are measured is short relative to the mechanical and electrical time constants of the DE system.
• the step of deriving an equation for the plane of best fit comprises performing regression analysis of the data derived from the measurements.
• the step of deriving an equation for the plane of best fit comprises solving a set of linear equations.
• the data derived from the measurements comprises potential difference across the DE, series current through the DE, charge supplied to the DE, and time, wherein the plane of best fit comprises a hyperplane.
• the feedback parameters further comprise equivalent series resistance of the DE, wherein the equivalent series resistance is derived from a third coefficient of the hyperplane equation.
• the third coefficient is the coefficient of a current variable of the hyperplane equation.
• the equivalent series resistance is approximated by the third coefficient of the hyperplane equation, divided by the capacitance of the DE.
• the leakage current is derived from the second coefficient of the plane equation by subtracting the nominal combined effects of the rate of change of capacitance of the DE and the rate of change of equivalent series resistance of the DE from the second coefficient.
• the leakage current is derived using a time history of the capacitance and equivalent series resistance of the DE.
• the approximation of the leakage current of the DE is derived using the approximations of capacitance and equivalent series resistance obtained according to the method of the present invention.
• the first coefficient of the plane equation corresponds to a partial derivative of charge supplied to the DE with respect to the potential difference across the DE.
• the second coefficient corresponds to a partial derivative of charge supplied to the DE with respect to time.
• the third coefficient corresponds to a partial derivative of charge supplied to the DE with respect to the series current through the DE.
• the data derived from the measurements comprises the potential difference across the capacitive component of the DE, charge supplied to the DE, and time.
• said time, potential difference across the capacitive component of the DE, and charge supplied to the DE are defined as orthogonal x-, y-, and z-axes, respectively.
• the leakage current is derived from the second coefficient of the plane equation by subtracting the product of the rate of change of the capacitance of the DE and the potential difference across the capacitive component of the DE from the second coefficient.
• the rate of change of the capacitance of the DE is derived using a time history of the capacitance obtained according to the method of the present invention.
• the method further comprises estimating the equivalent series resistance (ESR) of the DE by:
• PWM Pulse Width Modulated
• estimating the ESR of the DE further comprises the steps of:
• the method further includes the step of relating the capacitance of the DE to the physical state of the DE. This may include estimates of one or more of the planar area, thickness, and/or dimensions of the DE, for example.
• the method further includes the step of monitoring the leakage current and controlling inputs to the DE system to limit the leakage current.
• the method further includes the step of relating the equivalent series resistance of the DE to the strain in the DE.
• the method further comprises the step of controlling actuation of the DEA based upon the feedback parameters, thereby forming a closed-loop feedback control system.
• the invention may broadly be said to consist in a method of controlling a dielectric elastomer (DE), comprising the steps of:
• the invention may broadly be said to consist in a system adapted to perform the method of any one of the preceding statements.
• the invention may broadly be said to consist in a system comprising:
• DE dielectric elastomer
• power supply means adapted to supply a small-scale voltage oscillation to the DE; measurement means for measuring the potential difference across the DE and series current through the DE;
• computing means communicatively coupled to the measurement means, the computing means being adapted, over a predetermined period of time, to derive from the measurements data relating to the electrical characteristics of the DE, derive an equation for a plane which best fits the data, and derive one or more feedback parameters from coefficients of the plane equation.
• the measurement means has a sampling frequency higher than a frequency of the predetermined period of time.
• the computing means is further adapted to receive an input signal indicative of the required state of the DE, and to control one or more system inputs to provide closed-loop feedback control of the DE based upon the feedback parameters.
• the computing means is adapted to perform regression analysis on the stored data and thereby derive an equation for a hyperplane which best fits the stored data when defined as orthogonal axes.
• the feedback parameters comprise one or more of the capacitance, equivalent series resistance and/or leakage current of the DE.
• the computing means is further adapted to record the feedback parameters.
• Figure 1 is a diagram of a dielectric elastomer (DE) according to the prior art, shown in (a) rest; and (b) expanded states;
• Figure 2 is a process diagram illustrating the generation of electricity using a dielectric elastomer generator according to the prior art
• Figure 3 is a three dimensional representation of a hyperplane in a four dimensional space
• Figure 4 is an electrical lumped circuit model of a DE.
• Figure 5 is a schematic representation of a DEA as part of a simple self-sensing circuit according to the present invention
• Figure 6 illustrates the effect of the ESR upon the voltage across the DEA ( V DE A); and Figure 7 illustrates the series current (/ s ) through the DEA;
• Figure 8 illustrates the influence of capacitance on a fitted plane derived according to a second embodiment of the invention, with simulated capacitances of (a) 500pF; (b) 1000pF; and (c) 1500pF;
• Figure 9 illustrates the influence of a non-zero dC/dt on the fitted plane derived according to the second embodiment of the invention, with dC/dt values of (a) - 500%/s; (b) 0%/s; and (c) 500%/s;
• Figure 10 illustrates the influence of leakage current on the fitted plane derived according to the second embodiment of the invention, with (a) a leakage current of zero (i.e. infinite EPR); and (b) an equivalent parallel resistance (EPR) of 250 ⁇ ;
• Figure 11 illustrates the combined influence of leakage current and a negative value for dC/dt on the fitted plane derived according to the second embodiment of the invention, with (a) a negative value of dC/dt and a leakage current of zero; and (b) the same negative value of dC/dt, combined with an EPR of 250 ⁇ in parallel with the capacitance of the DEA;
• Figure 12 illustrates the combined influence of leakage current and a positive value for dC/dt on the fitted plane derived according to the second .embodiment of the invention, with (a) a positive value of dC/dt and a leakage current of zero; and (b) the same positive value of dC/dt, combined with an EPR of 250 ⁇ in parallel with the capacitance of the DEA;
• Figure 13 illustrates diagrammatically an expanding dot DEA used to experimentally validate the second embodiment of the method of the present invention
• Figure 14 shows a graph of estimated and actual equivalent series resistance (ESR) versus voltage for a range of values of ESR and EPR, wherein the estimated values are derived according to the second embodiment of the invention
• Figure 15 shows a graph of leakage current, estimated using the second embodiment of the invention, versus voltage for a range of values of ESR and EPR;
• Figure 16 shows a comparison of the capacitance of the test DEA as estimated by the second example embodiment of the self-sensing method of the present invention, with respect to capacitance measured by video extensometry;
• Figure 17 shows the capacitance of the DEA as estimated by the second embodiment of the self-sensing method plotted against the capacitance as estimated by video extensometry;
• Figure 18 shows the capacitance of the DEA as estimated by the second embodiment of the self-sensing method plotted against the capacitance as measured by video extensometry, after the estimated capacitance of the DEA has been scaled according to the relationship shown in Figure 17;
• Figure 19 is a schematic block diagram of an example system for implementing the self-sensing method of the present invention.
• Capacitance, leakage current through the dielectric membrane, and resistance of electrodes are each potentially important feedback parameters for a dielectric elastomer (DE), whether it is used as an actuator (DEA), a generator (DEG), or a sensor.
• DE dielectric elastomer
• DEA actuator
• DEG generator
• sensor a sensor
• Capacitance can be related to the instantaneous area of the electrodes and the thickness of the dielectric membrane. It also allows the electric field within the dielectric to be evaluated, and is essential for measuring the electrical charge stored on the DE at any point of time. Leakage current is an indicator of the overall efficiency of the DE and can be used as part of a health monitoring system.
• the resistance of the electrodes (equivalent series resistance - ESR) changes as the DE deforms, thus the electrodes act in a manner akin to a strain gauge. Knowledge of one or more of the above feedback parameters is therefore desirable for determining or estimating the state of the DEA.
• the "state” of the DEA may refer to size, shape, area, thickness, dimensions, deformation, electric field, electric charge, efficiency, and/or health of the DEA, for example.
• Each of the above feedback parameters (capacitance, leakage current, and equivalent series resistance) is directly or indirectly influenced by the others and thus vary dynamically as the DE is subjected to electrical and mechanical stimuli. It is not uncommon for the relationship between strain and resistance to change oyer time and with the number of strain cycles, for example. Unfortunately, however, the parameters of interest cannot easily be measured directly. According to the prior art, self-sensing has typically been limited to estimating the capacitance of the DEA and the Equivalent Series Resistance (ESR) of the electrodes.
• ESR Equivalent Series Resistance
• the present invention provides a method and apparatus which enables estimation of any or all of these key feedback parameters (capacitance, leakage current, and ESR) substantially in real-time, thereby reducing the need to make assumptions about their behaviour and improving the accuracy of feedback on the state of the DE for improved control and/or monitoring over a wide range of conditions.
• these key feedback parameters capacitor, leakage current, and ESR
• the method of the invention involves deriving data regarding four variables comprising three electrical characteristics of a dielectric elastomer (DE), and time. These four variables are defined as orthogonal axes forming an abstract four-dimensional space, and a plane of best fit through the data therefore comprises a three-dimensional hyperplane.
• the DE feedback parameters can be derived from coefficients of an equation for the hyperplane.
• a hyperplane is an (n-l )-dimensional subspace of an n-dimensional space. It can be described in terms of the co-ordinate system of the n-dimensional space by a single linear equation.
• the term 'hyperplane' as used throughout this description in describing the preferred embodiments of the invention refers to a three-dimensional subspace in a four-dimensional space.
• Figure 3 provides a three-dimensional representation of a hyperplane in a four dimensional space to illustrate the concept.
• the X, Y, and Z dimensions of the hyperplane are shown as orthogonal axes on a three dimensional plot, and the hyperplane has been discretized in the fourth dimension (W).
• FIG. 4 An electrical lumped circuit model of a DE is shown in Figure 4.
• the DE can be represented or modelled as a variable capacitor (C) in parallel with a variable resistance (R E PR, representing the equivalent parallel resistance of the DE), both of which are together connected in series with another variable resistance (RESR, representing the equivalent series resistance of the DE).
• C variable capacitor
• R E PR variable resistance
• RESR variable resistance
• the voltage across the capacitor (V c ) is the voltage across the DE (V D E) minus the product of the series current (/ s ) through the DE and the ESR of the electrodes ⁇ RESR), as shown by Equation 1.
• V c V DE - i s R ESR (1)
• the current through the DE (/ s ) is equal to the sum of the current through the capacitive element (/ c ) and leakage current through the EPR (IEPR) (Equation 2).
• the present invention derives estimates of the capacitance (C), leakage current (IEPR, in the model of the DE, the current through the equivalent parallel resistance R E PR), and equivalent series resistance ESR (RESR) by repeatedly measuring and deriving data from certain electrical characteristics of the DE, deriving an equation for a hyperplane of best fit through the data, and deriving feedback parameters of the DE from the coefficients of the hyperplane.
• a window of feedback data that spans a period of time greater than or equal to the period of the oscillatory component of the signal can be used to determine the DE feedback parameters.
• the present invention is not limited to any particular waveform for the oscillatory signal. Suitable waveforms include, but are not limited to, square waves, sine waves, triangle waves, or saw-tooth waves, for example.
• the first step of the method is obtaining and recording measurements of the potential difference or voltage across the DE (V D E), and series current (/ s ) through the DE during a short time window.
• the time window should be short relative to the mechanical and electrical time constants of the DE. It is convenient to set the length of the window to correspond with the period of the small scale oscillation superimposed on the voltage across the DE.
• Measurements of the voltage V DE and series current / s of the DE are taken at various times in the time period. From the latter measurement, the charge input to the DE (Q put) over the time period can also be derived (i.e. as the integral of the series current / s over the time period) and recorded.
• the measurements are preferably taken at a sampling rate substantially higher than the frequency of the time window, to give a plurality of data points during each time window.
• Equation 4 Because of the short time period, C can be modeled using a first order approximation, i.e., it can be assumed that dC/dt is substantially constant. If the peak-to-peak amplitude of V C for the period of the window is sufficiently small, the leakage current (i£PR, equivalent to dQi eakage dt) can also be assumed to be substantially constant. Applying these assumptions to Equation 3 and expressing the charge in terms of C and V C results in Equation 4.
• V C in the first term on the right hand side of Equation 4 can be replaced by an equivalent term incorporating an ESR dependent term, i.e., V C becomes V DE - I S RESR, resulting in Equation 5.
• V DE is an electrical characteristic of the circuit which can be easily measured, unlike V C .
• regression can be used to find the coefficients of a hyperplane which best represents the recorded data points for t, V DE , i s , and Q in p Ut (i.e. a hyperplane of best fit).
• Other methods for estimating the coefficients of the hyperplane may alternatively be used without departing from the scope of the invention, however.
• the term "hyperplane of best fit" is used herein to refer merely to a hyperplane upon or substantially adjacent to which lie at least a majority of the data points in a 4D plot of potential difference, current, charge, and time, by whatever method that plane may be derived, provided that the gradient or slope of the plane along each axis corresponds substantially with the "slope" of the data points along that same axis.
• the coefficients of the hyperplane formed by the recorded data points are equivalent to the partial derivatives of Q, npU f (Equation 5) with respect to t, i s , and V DE . That is, the coefficients associated with each of the variables (f, V DE , i s , and Qiliens put ) of the hyperplane equation represent the slope of the hyperplane along the respective axis. Given that the data points each fall substantially on a hyperplane, those slopes are equivalent to the aforementioned partial derivatives of Equation 5, and the feedback parameters of the DE can thus be derived from the hyperplane coefficients as follows.
• Equation 5 it can be assumed that dR ES p/dt is constant, as well as dC/dt and dQieakage/dt over the short time period or window of the relevant measurements.
• C(0)V c (0) is a constant, and thus does not appear in the partial derivatives.
• the partial derivative of Qm PU t with respect to V DE (equivalent to a first coefficient of the hyperplane equation) is the nominal capacitance of the DE over the period of the feedback window (C, Equation 6, time indices have been omitted for clarity);
• the partial derivative of Q in p Ut with respect to / s (equivalent to what is referred to as a third coefficient of the hyperplane equation), divided by C, is the nominal ESR of the electrodes for the period of the feedback window (RESR, Equation 7, time indices have been omitted for clarity);
• the partial derivative of Q mput with respect to t (equivalent to what is referred to as a second coefficient of the hyperplane equation) is the nominal combined effects of the rate of change of capacitance, the rate of change of the ESR of the electrodes, and leakage current (Equation 8, time indices have been omitted for clarity).
• An approximation or estimate of the capacitance C of the DE can thus be derived directly from the slope of the hyperplane of best fit along the voltage axis V DE (with respect to the Qmput axis), i.e. the coefficient of the voltage variable in an equation defining the hyperplane of best fit.
• an approximation of the equivalent series resistance R ES R of the DE can be derived from the slope of the hyperplane along the series current axis (with respect to the Qm p ut axis), divided by the capacitance C from the previous approximation.
• a time history of C and RESR calculated using Equations 6 and 7 can be used to calculate dC/dt and dR E sF dt respectively (both assumed to be substantially constant), which can be used to identify the individual terms of Equation 8, and in particular the leakage current I E PR which is equivalent to the final term, dQi eakage dt.
• This embodiment of the invention has several advantages over the second embodiment described below and/or the prior art, including one or more of the following:
• the hyperplane fitting process is achieved using a single stage mathematical process;
• the method does not necessarily require a PWM power supply, thereby providing increased flexibility with regard to hardware design;
• Capacitance is obtained directly from the fitting of the hyperplane to the available data.
• the process has low sensitivity to noise because:
• Integrating the series current acts as a means of low pass filtering the series current feedback data. Where noise has a mean value of zero, integrating acts to attenuate the noise component from the signal;
• Using a least squares regression process also provides a mechanism for evaluating the validity of the constant dR E sF dt, dC/dt and dQi eakage dt assumptions. Where the assumptions are valid, the residual of the regression process will be small. A large residual however indicates the one or more of the assumptions are not satisfied.
• the method involves deriving data regarding three variables comprising two electrical characteristics of a dielectric elastomer (DE), and time.
• DE dielectric elastomer
• These three variables are defined as orthogonal x-, y- and z- axes forming a three-dimensional space, and a plane of best fit through the data therefore comprises a plane.
• the DEA system uses Pulse Width Modulation (PWM) control of the input current (the actuation signal) to simultaneously actuate the DEA and introduce a small-scale oscillation to the voltage difference between the DE electrodes.
• PWM Pulse Width Modulation
• EPR equivalent parallel resistance
• ESR equivalent series- resistance
• PWM is readily compatible with digital computers, and enables a single power supply to be set to a fixed voltage and used to power multiple independent DEA simultaneously.
• controlling the duty cycle of the signal controls the average voltage seen across the DEA, which governs the degree of actuation.
• the rapid switching of the PWM signal introduces small scale oscillations to this voltage. This method thus explicitly accounts for dynamic changes in the resistance of the electrodes, changes to the conductance of the dielectric membrane at high electric fields, and current induced due to the rate of change of the capacitance of the DEA as it is actuated and/or relaxed.
• PWM is not an essential feature of the invention. Any other time varying signal may alternatively be used, such as an amplitude-modulated actuation signal with a high frequency oscillatory signal superimposed to permit sensing of capacitance and leakage current of the DEA. It is to be appreciated that the plane-fitting method of the invention may be used to enhance many existing or future methods of self-sensing, such as that disclosed by at least Keplinger et al., for example. However, a PWM power supply provides a number of advantages in at least this second example embodiment of the invention.
• the DEA can be modelled as a variable capacitor (C) connected in parallel with a variable resistance (R E PR), both of which are connected in series with another variable resistance (R E SR)-
• C variable capacitor
• R E PR variable resistance
• R E SR variable resistance
• the power supply of the DEA system is preferably a high voltage DC-DC converter connected to the DEA via a high voltage, high speed opto-coupler.
• the current flowing from the DC-DC converter to the DEA via a PWM switch is proportional to the input current to the opto-coupler.
• the DC-DC converter and opto- coupler behave as a current source (l SO urce)- PWM signal is used to control the input current to the opto-coupler.
• a 30 ⁇ ⁇ fixed resistor (R S ) is used to measure the series current through the DEA (by measuring the voltage V1 and applying Ohm's law), and a 100MQ: 120kQ ⁇ R P1 and R P2 respectively) resistor ladder (voltage divider) is used to measure the voltage V2 at the positive terminal of the DEA 50 and provide a passive discharge path for the DEA.
• the method of this second example requires data regarding the potential difference across the capacitive component of the DEA ( V c ), charge supplied to the DEA ⁇ Qmput) and time (f) to be derived from the measurements.
• V C V DEA - V ESR (9)
• the equivalent series resistance RESR can be estimated at the points in time where the PWM signal is transitioning between the high and the low states.
• the step change in the series current through the DE as the PWM signal transitions results in a step change in the voltage across the DE.
• the RESR is equal to the magnitude of the step change in the voltage V D E across the DE divided by the magnitude of the step change in the series current is through the DE at the transition of the PWM signal.
• the ESR can be sensed by considering two short time periods about the transition of the PWM signal (Part 1 and Part 2 in Figures 6 and 7).
• Part 1 begins just before the start of a cycle of the PWM input signal, and ends at the start of the PWM cycle.
• Part 2 continues from the end of Part 1 , beginning at the start of the PWM cycle and ending once the PWM input signal reaches its peak value.
• the time at the beginning of Part 1 is fate the time at the end of Part 1 /beginning of Part 2 is and the time at the end of Part 2 is t f .
• the ESR can be estimated by transforming V DE A to eliminate the discontinuity. This requires four curve fit operations to be performed.
• the series current through the DEA (/ s ) and potential differences across the DEA ( V DE A) during Part 1 , and series ' current through the DEA (i s ) during Part 2 are approximately linear as illustrated in Figures 6 and 7 (Equations 10-12). Assuming that both REPR and the derivative with respect to time of the capacitance are substantially constant about a specific point in time (a linearization point), applying a ramp input current (i.e.
• Equation 16 the equation for shown in Equation 16 can be derived.
• V c(Partl) ( m v ⁇ t + c vl ) ⁇ + c i ⁇ ) R ESR
• Vc ⁇ Part2 ( ⁇ 2* 2 + ⁇ ⁇ + ⁇ - + c i2 ) R ESR
• a limited slew rate on the PWM signal provides a smooth transition from discharging to charging and vice versa. This eliminates switching noise, and ensures V c is a continuous function with respect to time, enabling R E SR to be estimated.
• the coefficients of Equation 13 that fit the raw data in a least squares sense can be found by solving simultaneous equations, which can be achieved in a straightforward manner using Cramer's rule, for example. This process is also convenient in a practical sense because many of the necessary sums can be pre-computed, and those that cannot can be updated as each data point becomes available.
• the capacitance of the DEA can be calculated according to the method of this second example embodiment of the invention.
• V C the voltage across the capacitor
• V DE A the voltage across the DEA 50 ⁇ V DE A) minus the product of the series current (/ s ) through the DEA and the ESR of the electrodes (R E SR), as shown by Equation 17 (derived from Equation 9).
• Equation 17 derived from Equation 9).
• the series current through the DEA sub-circuit (4) is equal to the sum of the current through the capacitive element (/ c ) and leakage current through the EPR (i EPR ) (Equation 18).
• a window of feedback data that spans one or more periods of the oscillatory component of the signal can be used to determine the capacitance of the DEA (C) and the leakage current (I ' EPR)- PWM has been used to generate the oscillatory component of the actuation signal, however if V c is known, other methods to generate periodic oscillations can be used to determine C and leakage current without departing from the scope of the invention.
• the width of the window of feedback data is defined by Equations 3 and 4.
• the coefficients of the 2D plane formed by the t, V c , and Qiggi pu t data are equivalent to the partial derivatives with respect to t and V c of Q input in Equation 4. That is, the partial derivative with respect to V c of Qi crea pu t (the first coefficient) is capacitance (C) (Equation 19, time indices have been omitted for clarity).
• the partial derivative with respect to time of Q mput (the second coefficient) is the combined effect of the rate of change of the capacitance ⁇ V c .dC/dt), and the leakage current through the dielectric membrane (dQi ea kag( dt) (Equation 20, time indices have been omitted for clarity).
• Finite differences can be used to evaluate dC/dt (i.e. approximating dC/dt by determining the slope from two calculations of C over a short time period t), and this can be substituted into Equation 20 to separate the current i c induced due to the capacitance changing and the leakage current i EPR from the series current / s through the DEA.
• the leakage current I E PR (equivalent to the derivative with respect to time of the charge Qi ea kage leaking through the dielectric membrane) can thus be derived from the slope of the plane of best fit along the time axis, i.e., the second coefficient of the equation defining the plane of best fit (Equation 20).
• the 2D plane that best fits the available data in the least squares sense can be found using regression, for example, although other methods may alternatively be used without departing from the scope of the invention, and the term "plane of best fit" is used in this context to refer merely to a 2D plane in a 3D space which is determined as defining a plane upon or substantially adjacent to which at least a majority of measured points in a plot of potential difference and charge with respect to time lie, by whatever method that plane may be derived, provided that the gradient or slope of the plane along both axes corresponds substantially with the slopes of the data points, as described below with respect to Figures 8-12.
• the values for ay, a 2 , and a 3 that define the least squares solution can be found by solving the set of linear equations in Equation 21.
• the oscillations in the voltage across the DEA are important to avoid the Qm PU u V c , and t data collapsing substantially into a line in 3D space, whereupon an accurate plane of best fit cannot be found.
• the oscillations two non-collinear lines of data are created, and therefore a unique plane can be fitted to the available data points.
• Figure 8 displays the influence of changes in capacitance on the fitted plane.
• Leakage current and dC/dt were set to zero, and capacitances of 500pF, 1000pF, and
• (dQ/dV) plane is the slope of the fitted plane in the direction of the voltage axis
• C nom is the actual nominal capacitance of the DEA over the 5ms period shown
• (dQ/dt) p ane is the slope of the fitted plane in the direction of the time axis
• (VdC/dt) n0 m is the product of the nominal voltage across the capacitive component of the DEA and the nominal rate of change of the capacitance of the DEA for the 5ms period shown
• (dQi eakage /dt) n om is the nominal leakage current for the 5ms period shown.
• leakage current also acts to change the slope of the fitted plane along the direction of the time axis.
• leakage current i EPR and dC/dt both manifest as changes to the slope of the fitted plane along the direction of the time axis.
• Figure 1 1 (b) the combined influence of leakage current and a negative value for dC/dt is shown next to a plot for the same value of dC/dt with no leakage current in Figure 11 (a) for comparison.
• Figure 12(b) shows the combined influence of leakage current IEPR and a positive value for dC/dt is shown next to a plot for the same value of dC/dt and no leakage current in Figure 12(a). Both phenomena manifest in the same manner on the plane of best fit. However, because capacitance C can be calculated from the slope of the plane along the direction of the voltage axis, dC/dt can be estimated using a time history of C and finite differences. This can therefore be used to separate the effects of leakage current and the capacitance changing.
• the error in the estimated capacitance increases slightly under the combined influence of both a high value for dC/dt and leakage current, but it is still small and the slope of the fitted plane in the direction of the time axis still accurately estimates the combined influence of the currents induced due to the capacitance changing and leakage current.
• the expanding dot DEA illustrated diagrammatically in Figure 13 (not to scale) and generally indicated 80, was constructed from VHB4905, a common DEA membrane material available from 3M Company, pre-stretched to 16 times its original area and bonded to a rigid support frame 81.
• the frame had a circular internal aperture with a diameter of 130mm.
• a circular electrode 82 of diameter 50mm was applied to opposing sides of the membrane using Nyogel 756 electrically conductive carbon loaded grease, available from Nye Lubricants, Inc..
• Nyogel 756 conducting tracks 83 were also applied radially out from the electrode area to the points on the edge of the support frame that could be connected to external circuitry.
• the expanding dot DEA was used to evaluate the estimation of capacitance.
• C 0 rest capacitance
• a 10 ⁇ resistor was connected between the negative terminal of the DEA and ground, and a low voltage triangle wave (20V peak- to-peak at 0.8Hz) was applied to the positive terminal of the DEA.
• Capacitance was calculated using the ratio of the amplitude of the current through the 10 ⁇ resistor to the rate of change of the input voltage (Equation 23).
• a camera was used to rneasure the area of the electrodes when no voltage was applied to the DEA. This measurement served as the reference area that enabled the area stretch ( ⁇ ) of the DEA to be estimated, which is related to the instantaneous capacitance by Equation 24.
• the DEA was connected to the self-sensing circuit of Figure 5.
• the current source was configured to supply 95 ⁇ when the opto-coupler was switched on and substantially ⁇ when switched off.
• a 200Hz PWM signal was used to control the input current to the opto-coupler.
• the DEA was initially subjected to a PWM duty cycle of 10% until a steady state was reached. This was then increased by 1 % every 10 seconds.
• a sinusoidal disturbance signal (0-2000V at 0.1 Hz) was simultaneously applied to the second DEA patterned around the perimeter of the stretched membrane. The experiment was complete after the DEA had spent 10 seconds at a duty cycle of 30%.
• the area stretch of the electrodes was measured using video extensometry and used to estimate the capacitance. This was compared to the capacitance as estimated using the self-sensing process.
• Figure 14 shows the experimental validation of the ESR estimation.
• the lines indicate the true values of the ESR, and the data points indicate the estimated value for the ESR.
• the test DEA circuits used to evaluate the ESR and leakage current estimation capabilities of the self-sensing method used a ceramic capacitor with an unknown voltage dependent capacitance. Despite this, accurate estimates of both the ESR and leakage current were achieved. As shown by Figure 14, estimations for the ESR exhibited negligible voltage dependence for a range of voltages representative of those likely to be applied to a DEA. Furthermore, the estimated ESR exhibited no notable dependence on the magnitude of the leakage current.
• Figure 15 shows the experimental validation of the leakage current estimation.
• the lines indicate the predicted leakage current based on the voltage across the capacitive element of the test DEA circuit and the value of the EPR connected in parallel.
• the data points indicate the estimated value for the leakage current.
• the present invention thus provides accurate feedback on any one or more of three key feedback parameters of a DE, which in turn provide improved feedback on the state of the DE.
• Leakage current can at the very least provide a measure as to the efficiency of the DE, and may provide a mechanism for predicting and preventing dielectric breakdown (a major failure mode of the DE).
• High localized leakage current is invariably the ultimate cause of irreversible damage to the DEA when it undergoes dielectric breakdown.
• the present invention is also particularly useful for characterising the properties of a DE-based device in situ, i.e., feedback regarding the state of the DE can be provided while it is performing the task it was designed to do, whether as an actuator, generator, or simply a sensor.
• capacitance is an excellent predictor of the area of the electrodes, thus providing a basis for a closed-loop feedback system enabling improved control of actuation or generation.
• the present invention is particularly useful in providing feedback on the actuation of a dielectric elastomer actuator (DEA), and in particular may be utilised in forming a closed-loop feedback system using self- sensing to allow improved control of the DEA.
• DEA dielectric elastomer actuator
• Implementation of such a system typically involving some form of computing means or computing device such as a microcontroller, FPGA, or programmable logic controller (PLC), and a control algorithm such as proportional-integral-derivative (PID) control for example, is considered to be within the capabilities of a skilled person with experience in the field of embedded and control systems, and is therefore not described in detail herein.
• Equation 25 Q is the electrical charge on the DE, d is the thickness of the dielectric membrane, and volume is the product of the area of the electrodes and the thickness d of the dielectric membrane. If a constant charge Q is maintained on the DE electrodes, and the thickness of the dielectric membrane remains substantially uniform, a mechanical disturbance that acts to reduce d will also reduce the electrostatic pressure, and vice versa. The electrostatic pressure thus acts to reject the disturbance and thereby prevents the DE from undergoing electromechanical instability. In contrast, when a constant voltage is maintained the electrostatic pressure will act to sustain any mechanical disturbance. Most importantly, however, being capable of controlling both charge and voltage makes it possible to control the effective stiffness of the DE.
• Controlling the stiffness of the DE creates a number of potential opportunities particularly for DEA-based artificial muscle powered devices.
• Compliant actuators are backdrivable. They have the inherent ability to absorb mechanical shocks and dampen vibrations and mechanical disturbances.
• the grip strength can be controlled, and compliant grippers naturally adapt to the form of the object being gripped, making them well suited to the task of manipulating irregular shaped or soft objects, or objects that may shift relative to the manipulator as they are translated through space.
• compliant joints prevent the feet from bouncing upon landing and ensure good contact is maintained between the robot's feet and the ground, particularly where the ground is uneven, which improves the stability of the robot.
• controlling the compliance of the leg joints and letting momentum and kinetic energy govern the trajectory of each segment of the leg as opposed to maintaining close control of all of the joint angles using high gain position control offers significant savings in terms of the input energy required.
• the method of the present invention relates to obtaining measurement signals indicative of electrical characteristics of a DE, and transformation of those measurements to derive a signal or data representative of the physical state and/or feedback parameters of the DE.
• the present invention may therefore also be said to consist in a system adapted to perform self-sensing as described herein with respect to the method of the invention.
• An example of such a system adapted to perform the method of the first embodiment of the invention for closed-loop feedback control of a DEA is illustrated schematically in Figure 19 by way of a non-limiting example.
• the system preferably comprises a controller 190 which is adapted to control actuation of the DEA 191 (i.e. by controlling the actuation signal 192).
• the system also includes computing means or computing device 193 for implementing the self- sensing method of the present invention to derive estimates of one or more of the capacitance, equivalent series resistance, and leakage current of the DEA 191.
• the system also includes an oscillatory voltage source or other means 194 for introducing a small-scale oscillatory signal to the actuation signal 192 supplied to the DEA, and measurement means or sensors, preferably in the form of the series current sensor 195 and voltage sensor 196.
• the data acquisition means or unit 197 derives data representative of the potential difference across the DEA (or the capacitive component thereof, for the second embodiment of the method), series current through the DEA, and charge supplied to the DEA, with respect to time. These measurements are communicated to the computing means 193 for application of the method of the present invention to derive the DEA feedback parameters as described above.
• the controller 190 is adapted to receive an indication of a required state of the DEA in the form of the setpoint 198. This may be communicated to the controller 190 from a human interface device, for example.
• the method of the present invention would typically be implemented at least in part in a computer program to be executed by some form of programmable computing means or digital or mixed-signal computing device, generally an embedded computing device such as a microcontroller.
• an embedded computing device such as a microcontroller.
• the method of the present invention could potentially be performed at least in part by one or more application specific integrated circuits (ASIC).
• ASIC application specific integrated circuits
• the functions of the controller 190, computing means 193, and data acquisition means 197 described above may be performed centrally by a single such computing device (e.g. microcontroller or ASIC), or alternatively distributed between a plurality of such devices.
• the computing means 193 and/or data acquisition means 197 preferably includes a data storage means or memory device to store or record data required to perform the method of the invention.
• the data storage means may be volatile or non-volatile, and may be integrated within the computing means 193 in a single integrated circuit (IC), or may be a discrete IC component communicatively coupled with the computing means in the known way.
• Computer programs implementing the method of this invention may be distributed to users on a distribution medium such as floppy disk, CD-ROM, USB flash memory, or other computer-readable medium. From there, they may be copied to an embedded solid-state memory or a similar intermediate storage medium of the computing means. When the programs are to be run, they will be loaded either from the intermediate storage medium into the execution memory of the computing means, configuring the computing means to act in accordance with the method of this invention. All these operations are well-known to those skilled in the art of embedded and/or computer systems.
• the term "computer-readable medium” encompasses distribution media, intermediate storage media, execution memory of a computer, and any other medium or device capable of storing for later reading by a computer a computer program implementing the method of this invention.
• the invention may thus also be said to consist in a computer program implementing the method of the invention, a computer-readable medium storing the computer program, and/or a system adapted to perform the method and/or execute the software of the invention.
• the computing means is communicatively coupled with measurement means or measurement devices for measuring the electrical potential difference across the DE and the series current through the DE at specific points in time. From the series current measurements, the charge input to the DE can be derived.
• Suitable measurement means/devices are known to those skilled in the art, and may include circuits comprising discrete or integrated resistors, capacitors, and operational amplifiers (op-amps) as unity gain buffer amplifiers, for example, arranged in the known way.
• the data acquisition means 197 and/or computing means 193 preferably include at least one analog-to-digital converter (ADC) to convert the measurements into a digital format suitable for further processing by a digital computing means.
• ADC analog-to-digital converter
• instantaneous in this specification refers to the general state of a component or integer of the invention at any particular point in time. However, it is not intended to exclude a measurement or approximation of that state which is derived from multiple measurements, as might be the case where the measurement is averaged or filtered over a short period of time (with respect to the mechanical or electrical time constant), for example. In fact, such averaging may be preferred to minimise the effects of noise or quantisation error.
• instantaneous depending on the context thus refers to any measurement, derivation or approximation which is at least indicative of the time-varying state of the component at or substantially adjacent a particular point in time.
• the invention provides a method for dynamically estimating the DE state feedback parameters of equivalent series resistance of the electrodes, the capacitance of a DE, and the leakage current through the dielectric membrane whether the DE is being actuated or being used to generate electrical energy, or being used solely as a sensor. This is achieved by self- sensing, meaning that no discrete strain, displacement, velocity, or acceleration sensors are required.
• the invention may also be said broadly to consist in the parts, elements and features referred to or indicated in the specification of the application, individually or collectively, in any or all combinations of two or more of said parts, elements or features. Furthermore, where reference has been made to specific components or integers of the invention having known equivalents, then such equivalents are herein incorporated as if individually set forth. Unless the context clearly requires otherwise, throughout the description, the words “comprise”, “comprising”, and the like, are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense, that is to say, in the sense of "including, but not limited to”. Any discussion of the prior art throughout the specification should in no way be considered as an admission that such prior art is widely known or forms part of common general knowledge in the field.

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Engineering & Computer Science (AREA)
• Manufacturing & Machinery (AREA)
• Measurement Of Resistance Or Impedance (AREA)
• Investigating Or Analyzing Materials By The Use Of Electric Means (AREA)

Priority Applications (5)

Applications Claiming Priority (4)

Publications (1)

Family

ID=45975441

Family Applications (1)

Country Status (6)

Cited By (3)

Families Citing this family (23)

Citations (1)

Family Cites Families (15)

• 2011
  • 2011-10-18 KR KR1020137012782A patent/KR20140009191A/ko not_active Withdrawn
  • 2011-10-18 CN CN201180056493.5A patent/CN103403902B/zh active Active
  • 2011-10-18 JP JP2013534847A patent/JP6207395B2/ja active Active
  • 2011-10-18 WO PCT/NZ2011/000214 patent/WO2012053906A1/en not_active Ceased
  • 2011-10-18 US US13/880,336 patent/US9170288B2/en active Active
  • 2011-10-18 EP EP11834687.3A patent/EP2630672B1/en active Active

Patent Citations (1)

Non-Patent Citations (2)

Also Published As

Similar Documents

Legal Events