Method and analyzer for determining a measured value of a measured variable of process automation technology
A method and analyzer for determining a measured value of a measured variable of process automation technology in a liquid or gaseous medium by means of an optical sensor includes taking a sample of the medium; mixing the sample with one or more reagents; supplying an emitter of the optical sensor with an exciter signal for producing sent light, wherein the sent light by interaction with the mixed sample is converted into received light as a function of the measured variable; producing a receiver signal by means of a receiver of the optical sensor from the converted, received light; and determining the measured value based on the receiver signal and a calibration function, which includes a term which takes aging of the reagents into consideration.
1. A method for determining a measured value of a measured variable of process automation technology comprising:
taking a sample of a liquid or gaseous medium;
mixing the sample with one or more reagents to generate a mixed sample;
supplying an emitter of an optical sensor with an exciter signal to produce sent light, the emitter structured to produce sent light of at least two wavelengths, and the optical sensor including a receiver associated with the emitter and structured to receive received light, wherein the sent light is converted into received light as a function of the measured variable by absorption by the mixed sample;
producing a receiver signal by means of the receiver from the received light; and
determining the measured value based on the receiver signal and a calibration function, wherein the calibration function includes a term based on a predetermined model of the aging of the reagents such that the measured value is determined based at least partially on the age of the reagents.
2. The method as claimed in claim 1 , wherein the calibration function is generated by means of a kernel method.
3. The method as claimed in claim 2 , wherein the kernel method is a support vector machine or a kernel Fisher discriminant.
4. The method as claimed in claim 3 , wherein the measured value is calculated by
c ( x )= a 0 +Σ i-1 N a i ·k ( x,x i ),
with a kernel function
k ( x,x i )=exp(−μ∥ x i −x∥ 2 ),
wherein x is a data vector comprising the receiver signal of the at least two wavelengths, x i are support vectors, a o , a i are coefficients, μ is a kernel parameter, and N is a number of support vectors in the form of a natural number.
5. The method as claimed in claim 4 , wherein the data vector x further includes at least one of the parameters, ambient temperature of a analyzer, temperature in the analyzer, sampling point in time, process duration, reagents, reagent age, and/or mixing ratios of reagents.
6. The method as claimed in claim 4 , wherein the support vectors x i , coefficients a o , a i and kernel parameter μ are determined before the measured value is determined.
7. The method as claimed in claim 1 , wherein the reagents are not replaced for at least 10 weeks.
8. The method as claimed in claim 1 , wherein the method is executed in an analyzer for analyzing for at least one ion concentration.
9. The method as claimed in claim 1 , wherein coefficients and parameters of the calibration function are at least partially based on the aging of the reagents.
10. An analyzer for determining a measured value of a measured variable of a sample, the analyzer comprising:
a mixing cell structured to enable mixing of a sample with one or more reagents to produce a mixed sample;
a measuring device including an emitter structured to emit sent light of at least two wavelengths and a receiver associated with the emitter for receiving received light, wherein the sent light is converted into received light as a function of a measured variable by absorption by the mixed sample, and wherein the receiver generates a receiver signal from the received light; and
a controller configured to determine a measured value of the mixed sample based on the receiver signal and a calibration function, wherein the calibration function includes a term based on a predetermined model of the aging of the reagents such that the measured value is determined based at least partially on the age of the reagents.
11. The analyzer as claimed in claim 10 , wherein the calibration function is generated by means of a kernel method.
12. The analyzer as claimed in claim 11 , wherein the measured value is calculated by
c ( x )= a 0 +Σ i-1 N a i ·k ( x,x i ),
with a kernel function
k ( x,x i )=exp(−μ∥ x i −x∥ 2 ),
wherein x is a data vector comprising the receiver signal of the at least two wavelengths, x i are support vectors, a 0 , a i are coefficients, μ is a kernel parameter, and N is a number of support vectors in the form of a natural number.
13. The analyzer as claimed in claim 12 , wherein the coefficients and parameters of the calibration function are at least partially based on the aging of the reagents.
14. The analyzer as claimed in claim 10 , wherein the controller is a transmitter.