RU2333523C2 - Transducer static characteristic correction method and device to this effect - Google Patents

Abstract

FIELD: metrology.

SUBSTANCE: invention relates to metrology and can be used for correction of the pressure, force, angular travel transducer static characteristics. The method of correction of the transducer static characteristic of the Z=f(X) form consists in conversion of a physical quantity X by function F₀ into quantity Y, subtraction thereof of a compensating uniform quantity Y_k, conversion of no-compensation quantity ΔY=Y-Y_k, conversion of resulted quantity into quantity Z, conversion by function F₁ of quantity Z into quantity Y_X, and conversion of resulted quantity Yx by parametric conversion Y_k=F₂(Y_X, Θ), conversion of F₂ effected allowing for T factors influencing the X physical quantity measurement representing a mathematical model of converting Y_X into Y_k being expressed as F₂=f[F₀,F₁]. The transducer static characteristic correction device incorporates converter 1 to convert measured quantity X into quantity Y made up of primary converter 2 and first ADC 3, a comparator 4, an integrator 5, DAC 6, ADC 7, mathematical model unit 8 built around a microprocessor and pickups 9 of factors influencing the physical quantity X measurement process.

EFFECT: higher linearity of the transducer static characteristic and its insusceptibility to factors influencing the measurement process.

3 cl, 1 dwg

Description

The invention relates to measuring technique and can be used to correct the static characteristics of measuring transducers of pressure, force, angular displacement and other physical quantities.

Non-linearity of the static characteristic of the measuring transducer, its sensitivity to factors influencing the measurement process can significantly distort the information received from it, therefore, various methods are used to correct the static characteristic.

A known method of correcting the static characteristics of the measuring transducer, based on the approximation of the function inverse to the static characteristic of the transducer, using a power polynomial with an accuracy exceeding the accuracy of the transformation [1].

The disadvantage of this method is the need to take into account non-linearity and sensitivity to influencing factors of all operations, which leads to complication of the approximation of the inverse function and its subsequent implementation.

As a prototype of the proposed method, a method for correcting the non-linearity of the static characteristics of the measuring transducer, having the form Z = f (X), which is based on the use of compensation transformation [2], is selected. It consists in converting the measured physical quantity X by means of the direct function of the measuring transducer F ₀ into the value Y, subtracting the compensating homogeneous quantity Y _k from it, converting the non-compensation value ΔY = YY _k , converting the obtained value into the value Z, using functions F _{1 the} inverse transformation of the magnitude of Z in the value of Y _to , uniform to the value of Y, and compare the values of Y and Y _to .

A device for implementing this method includes a converter of physical quantity X into a value Y connected by its output to a non-inverting input of the comparison device, the output of which is connected to the input of the non-compensation converter, the output of which is connected to the input of the converter of the input signal to the value Z, the output of which is connected to the input of the inverse of the converter Z in the value of Y _to , homogeneous to the value of Y, connected by its output to the inverting input of the comparison device [2].

The disadvantage of this method and device is the dependence of the measurement error on the nonlinearity and sensitivity to the influencing factors of the primary and inverse transformations, which leads to insufficient linearity of the static characteristic of the measuring transducer and its non-invariance to the factors influencing the measurement process.

The problem solved by the invention is to increase the linearity of the static characteristics of the measuring transducer and ensuring its invariance with respect to factors influencing the measurement process.

This problem is solved by the fact that in the method for correcting the static characteristic of the measuring transducer, having the form Z = f (X), in which the physical quantity X is converted into the quantity Y by means of the function F ₀ , the compensating homogeneous quantity Y _{k is} subtracted from it, the non-compensation value is converted ΔY = YY _k , the obtained value is converted to the value Z and, using the function F _1, the value Z is converted to the value Y _x ; the value obtained as a result of the parametric transformation Y _k = F is used as Y _{k 2} (Y _x , Θ), taking into account the factors T that affect the process of measuring the physical quantity X, while F ₂ is a mathematical model of the conversion of Y _x to Y _k , expressed as:

F ₂ = f [F ₀ , F ₁ ],

the parameters Θ of the function F _{2 are} determined by setting the known values of the measured physical quantity X and factors T, measuring the current value of Z, calculating the difference between the current value of Z and the corresponding value from the nominal characteristic Z _n = F _n (X), integrating the difference ΔZ = ZZ _n in time and use ΔZ to adjust the parameters Θ of the function F ₂ in order to reduce the difference ΔZ to zero, and the time integration operation is used as a non-compensation transformation ΔY.

A device for implementing a method for correcting the static characteristic of a measuring transducer, having the form Z = f (X), including a physical quantity X transducer in a Y value, connected by its output to a non-inverting input of the comparison device, the output of which is connected to the input of the non-compensation transformer, the output connected to the transducer input the input signal to the value Z, the output of which is connected to the input of the inverse converter of the value Z to the value Y _x , equipped with a block of the mathematical model the value of Y _x in the value of Y _k , homogeneous to the value of Y, containing n inputs, the first input connected to the output of the inverter, the remaining (n-1) inputs connected to the sensors of the influencing factors T, the output is connected to the inverting input of the comparison device, and the converter non-compensation made in the form of an integrator.

The converter of the physical quantity X to the value Y is made in the form of a primary transducer of the physical quantity X connected by the output to the input of the first analog-to-digital converter; the converter of the input signal to the value Z is made in the form of a digital-to-analog converter, the inverse converter is made in the form of a second analog-to-digital converter, and the mathematical model unit is implemented on a microprocessor.

The drawing schematically shows a device with which the inventive method is implemented. It is an astatic system and includes a transducer 1 of a measured physical quantity X into a Y value, consisting of a primary transducer 2 and a first analog-to-digital converter (ADC) 3, the output of which is connected to a non-inverting input of a comparison device 4, a non-compensation transformer made in the form of an integrator 5, connected to the input with the output of the comparison device 4, and the output to the input of the digital-to-analog converter (DAC) 6, the output of which is the output of the inventive device. The output of the DAC 6 is connected to the input of the second ADC 7, the output of the conversion of the value of Y _x to the value of Y _k connected to the first input of the mathematical model unit, implemented on microprocessor 8, the output of which is connected to the inverting input of the comparison device 4, and the remaining (n-1) inputs 9 factors are connected to the sensor outputs that affect the process of measuring the physical quantity X. Such factors may include, for example, temperature, atmospheric pressure, humidity, etc.

Note that the comparison device 4, the ADC 3 and 7, the integrator 5, the DAC 6 and the temperature sensor can be performed on the basis of the ADUC816 microconverter manufactured by Analog Devices.

The method according to the invention is as follows. The measuring transducer 1 measures a physical quantity X, for example, pressure, and converts it through the direct function of the measuring transducer F ₀ into a digital output signal Y, which comes from the output of the ADC 3 to the non-inverting input of the comparison device 4 and then to the input of the integrator 5. From the output of the integrator 5, the signal arrives at the input of the DAC 6, which converts the digital signal into an analog output signal of the device, for example, current or voltage. From the output of the DAC 6, the signal is fed to the input of the ADC 7, in which, using the function F ₁ , the Z is inverted to a value of Y _x , uniform in nature to the value of Y. Next, the signal (value Y _x ) is fed to the first input of the microprocessor 8, to the rest the inputs of which the output signals of the sensors are supplied 9. The microprocessor 8 algorithm implements a mathematical model for converting Y _x to Y _k — the value supplied to the inverting input of the comparison device 4 and is described by the parametric transformation F ₂ .

In the comparison device 4, the signals Y and Y _k are compared. The value ΔY = YY _k in the static mode when using the integrator as a non-compensation transformer is equal to zero, i.e. Y = Y _to or

In that case, if the function F _{2 is} defined so that the function of converting the quantity Z to Y _k has the same form as the function of converting the input physical quantity X to Y = F ₀ (X), the static characteristic of the measuring transducer Z = f (X) becomes linear Z = KX, where K is the coefficient of proportionality.

The desired function F ₂ is parametric. Its parameters Θ are determined by setting the known values of the measured physical quantity X and additional influencing factors T, measuring the current value of Z, obtaining the difference between the current value of Z and the corresponding value from the nominal characteristic Z _H = F _H (X), integrating the difference ΔZ = ZZ _H in time and using ΔZ to control the adjustment of the parameters of the function F ₂ in order to reduce this difference to zero.

As F _2, you can use a piecewise linear approximating function, an approximation by a power polynomial, a spline approximation, etc. For example, when using a piecewise linear function, the parameters of its ith segment are the multiplicative coefficient A _i and the additive coefficient B _i , i.e. . Θ = {A _i , B _i }. The configuration of the model parameters in this case is as follows. When X = X ₀ = 0, the value of the additive coefficient of the first segment B ₁ is equal to Y ₀ . Next, the known value of the value X = X _{1 is} supplied to the input of the measuring transducer, the current value Z = Z _{1 is} measured with an exemplary device, it is compared with the value calculated by the nominal characteristic, which in most cases uses the linear function Z _н1 = KX ₁ , integrate in time, the obtained value of uncompensation ΔZ ₁ = Z ₁ -Z _n1 and use the value at the output of the integrator as the multiplicative coefficient of the first segment. At the end of the transition process, the stabilized value at the output of the integrator is taken as A ₁ . Next, the following known value of the value X = X _{i is} supplied to the input of the measuring transducer, the current value Z = Z _{i is} measured by an exemplary device, it is compared with the value calculated by the nominal characteristic Z _нi = _{КХ i} , the obtained non-compensation value ΔZ _i = Z is integrated over time _i -Z _нi and use the value at the output of the integrator as A _i , and the current additive coefficient is calculated by the formula:

Upon completion of the transition process at the output of the integrator, the multiplicative and additive coefficients of the i-th segment are considered tuned and then the (i + 1) -th segment is set up.

To ensure the invariance of the measuring transducer to the influence of an influencing factor, such as temperature T, it is necessary to obtain a family of piecewise linear functions, which is used as a model of F ₂ . In this case, the coefficients of the segments of each of them are adjusted according to the algorithm described above for each of a certain number of known values of temperature T _j acting on the measuring transducer, for example, in a heat and cold chamber.

The coefficients of the model at non-nodal temperatures are determined by the microprocessor of the measuring transducer based on the proportion:

,

,

,

- настроенные параметры i-го сегмента на j-й и (j+1)-й температурах.Where

,

,

,

- tuned parameters of the i-th segment at the j-th and (j + 1) -th temperatures.

The determination of parameters Θ according to the above algorithm automatically takes into account the conversion functions F ₀ and F ₁ , which allows you to write F ₂ in the form:

F ₂ = f [F ₀ , F ₁ ].

The necessity of the above procedure for determining the parameters Θ of the function F ₂ is explained by the following.

We represent the function F ₂ in the form

where (•) ^-1 is the symbol of the inverse function.

Substitute (3) in (1):

F ₀ (X) = F ₂ [F ₁ (Z)] = F ₀ ({F ₁ [F ₁ (Z)]} ^-1 ) = F ₀ (Z).

The obtained identity proves that expression (3) determines the form of the function F ₂ .

и

где ΔF₀ и ΔF₁ - погрешности оценивания. Поэтому определение F₂ по формуле (3) на основании оценок вносит большую погрешность в измерение физической величины X по сравнению с определением функции F₂ согласно заявляемому способу.In practice, the functions F ₀ and F ₁ can be determined with errors

and

where ΔF ₀ and ΔF ₁ are the estimation errors. Therefore, the determination of F ₂ by the formula (3) based on the estimates introduces a large error in the measurement of the physical quantity X in comparison with the determination of the function F ₂ according to the claimed method.

The inventive method and device were used to correct the static characteristics of the Sapphire-22 type measuring transducers of absolute, relative and differential pressure in the operating temperature range from -40 ° C to + 80 ° C and significantly increased the linearity of the static characteristics of the pressure transducers and its invariance to various factors affecting the measurement process.

LITERATURE

1. D.Shaponich, A.Zhigich. Correction of a piezoresistive pressure sensor using a microcontroller. Instruments and experimental equipment, 2001, No. 1, S.54-60.

2. P.P. Ornatsky. Automatic measurements and instruments. Kiev: Vishka school, 1980, pp. 107-109 (prototype).

Claims (3)

1. A method for correcting the static characteristic of a measuring transducer, having the form Z = f (X), in which the measured physical quantity X is converted into the value Y by the function of the measuring transducer F0, the homogeneous value Yk is subtracted from it, and the value obtained as a result of the indicated action is converted non-compensation ΔY = Y-Yк in the value of Z, which is fed to the output of the measuring transducer and is carried out using the function F1 analog-to-digital conversion of the value of Z in the value of YX, different yuschiysya in that a direct function of the transmitter F0 converts a physical quantity x to an analog-to-digital signal y, said conversion value non-compensation ΔY in value Z is performed by integrating the values ΔY and digital to analogue conversion of the resulting integral value as Y _to use the value obtained as a result of the parametric transformation Yк = F2 (YX, Θ), where F2 = f [F0, F1], and the parameters Θ of the function F2 are determined by setting the known values of the physical quantity X and (n-1) factors affecting the process of measuring it, measuring the current Z value, calculating the difference between the current Z value and the corresponding value from the nominal characteristic Zн = Fн (X), integrating the difference ΔZ = Z-Zн over time and using ΔZ for adjusting the parameters Θ of the function F2 in order to reduce the difference ΔZ to zero. 2. A device for implementing the method of correcting the static characteristic of a measuring transducer, having the form Z = f (X), including a physical quantity X transducer in a Y value, by means of a measuring transducer function F0, connected by its output to a non-inverting input of the comparison device, the output of which is connected to the input integrator, the output associated with the input of the digital-to-analog converter of the input signal to the value Z, the output of which is connected to the input of the second analog-to-digital converter with converting F1 of magnitude Z to YX, characterized in that a mathematical model block has been introduced for converting YX to Yk, containing n inputs, the first input being connected to the output of the converter of Z to YX, the remaining (n-1) inputs are connected to sensors of factors affecting the measurement process, and the output is connected to the inverting input of the comparison device, while the non-compensation converter is made in the form of an integrator, the converter of the physical quantity X to Y is made in the form of a primary converter atelier of the physical quantity X connected by the output to the input of the first analog-to-digital converter. 3. The device according to claim 2, characterized in that the block of the mathematical model is implemented on a microprocessor.