Patent Grant
7908122
The present application is based on, and claims priority from, French Application Number 07 03734, filed May 25, 2007, the disclosure of which is hereby incorporated by reference herein in its entirety.
1. Field of the Invention
The field of the invention is that of signal processing for systems disposed in differential configuration. Differential configuration is understood to mean a configuration where two systems allow a variable representing a parameter and its opposite to be generated simultaneously.
This configuration allows certain errors to be avoided or to be significantly reduced. This method is particularly suited to the processing of signals coming from measurement sensors where the measurement may be both tainted by noise and be subjected to spurious effects like thermal drifts. The correction for thermal drifts is indeed a major problem in measurement systems. There exist numerous physical principles where a variable, an output signal and its opposite may be readily obtained. Resonant mechanical devices, certain optical devices, electrical or electronic devices will notably be mentioned. To give a simple example, if the movements of an object, which may be a plate, a beam or a membrane, are measured, a displacement of +d of a first side of this object and of −d of the other side is obtained.
2. Description of the Prior Art
The general case where a single system processes the information is presented in
The following may thus be written:
X=X0ƒ(y) with ƒ(y) a polynomial equal to 1 for y=0, X0 forming the quiescent output of the processing operation, in other words for a zero input signal.
In order to determine y starting from X, X0 must be known together with the coefficients of f. These parameters are linked to the characteristic physical dimensions of S and may be determined by calibration, but they can vary as a function, for example, of temperature.
If, in order to determine y starting from X, the numerical values for X0 and the coefficients of f determined during the calibration, which was carried out at a temperature that may be different from the effective temperature of the system at the moment when the latter returns the information X, are simply used, an error is committed in the estimation of y which may be incompatible with the degree of precision required. Conventionally, this error is decomposed into two terms:
In addition, the noise present in the system S leads to noise in the estimation of y which may also be incompatible with the required precision.
Finally, in the case where the processing operation allowing X to be determined starting from u is a digital processing operation, the clock noise involved in the system sampling will cause additional noise in X.
The goal of the invention is to provide a method based on the employment of two systems in differential configuration that will allow:
More precisely, the subject of the invention is a signal measurement and processing device comprising at least:
fi,1 being the coefficients of the said polynomial;
fi,2 being the coefficients of the said second polynomial; the processing device comprising:
Advantageously, the coefficients X0,1, X0,2, fi,1 and fi,2 depend on a parameter and can be put in the form of a polynomial function of the said parameter.
Advantageously, when the output signals X1 and X2 vary linearly with the input signal, the value of the parameter is determined by the value of the weighted sum X4 and the value of the first signal y by the value of the weighted difference X3.
Conventionally, the parameter T is the temperature.
Still other objects and advantages of the present invention will become readily apparent to those skilled in the art from the following detailed description, wherein the preferred embodiments of the invention are shown and described, simply by way of illustration of the best mode contemplated of carrying out the invention. As will be realized, the invention is capable of other and different embodiments, and its several details are capable of modifications in various obvious aspects, all without departing from the invention. Accordingly, the drawings and description thereof are to be regarded as illustrative in nature, and not as restrictive.
The invention will be better understood and other advantages will become apparent upon reading the description that follows presented in a non-limiting manner and with reference to the appended figures among which:
The method according to the invention is based on the employment of a differential system. The signal processing according to the invention implements two systems S1 and S2 that respectively receive the input signal y and its opposite −y.
Each system returns a signal u over one channel. Thus, u1 is on the channel 1 and u2 is on the channel 2. A processing operation ST allows u1 and u2 to be demodulated in order to obtain two signals X1 and X2 varying according to polynomial functions f1 and f2 with, using the previous notations:
on the channel 1 X1=X0,1ƒ1(y)=X0,1[1+ƒ1,1y+ƒ2,1y2+ . . . ]
on the channel 2 X2=X0,2ƒ2(−y)=X0,2[1+ƒ1,2y+ƒ2,2y2+ . . . ]
fi1 and fi2 being the coefficients of the polynomial functions f1 and f2. It should be noted that it is practically always possible to decompose a function in this form. The differences between the parameters X0,1 and X0,2 and also between the coefficients of the polynomials f1 and f2 originate from the disparities between the two systems, which may be intentional in order to avoid spurious coupling effects, or otherwise, in which case coming from manufacturing tolerances.
The core of the method consists in combining the two equations in order to obtain two signals X3 and X4 by performing the aforementioned weighted sum and weighted difference:
This then leads to:
The ratios
can be determined by calibration. If X0 and the coefficients of f vary as a function of a parameter, such as, for example, the temperature T, then these parameters may be written according to a polynomial law. In this case, this yields:
X0,1(T)=X0,1(T0)[1+β1(T−T0)+β2(T−T0)2+ . . . ) and
ƒ1,1(T)=ƒ1,1(T0)[1+δ1(T−T0)+δ2(T−T0)2+ . . . ]
with T the effective temperature of the system and T0 the temperature of the system during the calibration.
If they are analogous systems, X0,2 and ƒ1,2 will vary in the same fashion as a function of temperature and this can be expressed as follows:
X0,2(T)=X0,2(T0)[1+β1(T−T0)+β2(T−T0)2+ . . . )
ƒ1,2(T)=ƒ1,2(T0)[1+δ1(T−T0)+δ2(T−T0)2+ . . . ]
The ratios
are therefore independent of the temperature. They may then just be determined during a calibration. X3 and X4 may thus be determined from X1 and X2 without spoiling these expressions with errors due to the thermal variations. The calculation of X3 and X4 starting from X1 and X2 does not pose any particular technical problem and may be carried out by various electronic methods, either digital or analogue.
If f is a polynomial of degree 1, which represents the general case where the output signal X varies linearly with the input signal, then the equations 3 and 4 become:
X4 no longer depends on the input y. X4 therefore only depends on the temperature. The operating temperature of the system can thus be determined using X4. Knowing the temperature, it is easy to determine the coefficients from the equation 3 which are X0,1 and ƒ1,1+ƒ1,2. Indeed, the variation law with temperature for these coefficients can be determined during calibration. Knowing these coefficients, y is able to be determined starting from X3.
In the case where the degree of f is greater than 1, it is considered that there exists a polynomial variation law for the coefficients fi,1 and fi,2 as a function of temperature, which is always true for physical phenomena. The equations 3 and 4 may then be written as being polynomials in y and T, with coefficients that can be determined by calibration. Starting from the values of X3 and X4, y and T are thus determined.
Thus, using the weighted sum and the weighted difference of X1 and X2, the input y and the temperature are able to be determined. The use of the equations 3 and 4 therefore allows the errors due to the temperature to be avoided.
It is observed that X3 is zero in the absence of an input signal. In determining y starting from X3, an error in the estimation of the parameters of the system leads to a scale factor error but no zero-bias error. This would not have been the case if X3 had been the simple difference of X1 and X2. A zero-bias equal to X0,1−X0,2 would then have resulted in the expression for the difference: X1-X2. Using the weighted difference allows this zero-bias error to be eliminated. The use of the weighted sum allows the dependence of X4 as a function of y to be limited since there is no linear dependency on y. Thus, thanks to the weighted sum and difference of X1 and X2, two quantities varying in a substantially different manner with temperature and the signal to be analysed y are obtained.
By determining y starting from X3, the clock noise is avoided. Indeed, since the sampling moments in time of the output signals of the two systems are identical, the sampling clock noise is in common mode. The clock noise therefore has no direct effect.
Because of the noise affecting the system S, the signals X have a noise component that is additive and that may be denoted:
X1={circumflex over (X)}1+b1
X2={circumflex over (X)}2+b2
{circumflex over (X)}1 and {circumflex over (X)}2 being the signals in the absence of noise, and b1 and b2 being the noise components.
By using the weighted sum and difference of X1 and X2, a gain of 3 dB is obtained in terms of noise power in the estimation of y. Indeed, by considering f to be a polynomial of degree 1, the noise in the estimation of y is:
By considering that the noise components have the same power level on the two channels, the spectral power density of this noise is approximately:
with SPD(b1): Spectral Power Density b1.
The power level of the noise has therefore been divided by 2 by employing the weighted difference with respect to the power obtained by determining y directly from X1.
It will be readily seen by one of ordinary skill in the art that the present invention fulfils all of the objects set forth above. After reading the foregoing specification, one of ordinary skill in the art will be able to affect various changes, substitutions of equivalents and various aspects of the invention as broadly disclosed herein. It is therefore intended that the protection granted hereon be limited only by definition contained in the appended claims and equivalents thereof.
| Number | Date | Country | Kind |
|---|---|---|---|
| 07 03734 | May 2007 | FR | national |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5146414 | McKown et al. | Sep 1992 | A |
| 7453923 | Leblond et al. | Nov 2008 | B2 |
| 20080129586 | Martin | Jun 2008 | A1 |
| 20080214212 | Pridmore et al. | Sep 2008 | A1 |
| Number | Date | Country |
|---|---|---|
| WO9424573 | Oct 1994 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20080294385 A1 | Nov 2008 | US |
