Assessing the level of confidence for expressing extended uncertainty through control errors on the example of a model of a means of measuring ion activity ACTA IMEKO ISSN: 2221-870X June 2021, Volume 10, Number 2, 199 - 203 ACTA IMEKO | www.imeko.org June 2021 | Volume 10 | Number 2 | 199 Assessing the level of confidence for expressing extended uncertainty: a model based on control errors in the measurement of ion activity Oleksandr Vasilevskyi1 1 Vinnytsia National Technical University, 95 Khmelnitsky Shose str., 21021, Vinnytsia, Ukraine Section: RESEARCH PAPER Keywords: measurements; uncertainty; level of confidence; control errors; activity of ions; expanded uncertainty Citation: Oleksandr Vasilevskyi, Assessing the level of confidence for expressing extended uncertainty through control errors on the example of a model of a means of measuring ion activity, Acta IMEKO, vol. 10, no. 2, article 27, June 2021, identifier: IMEKO-ACTA-10 (2021)-02-27 Section Editor: Maik Rosenberger, Ilmenau University of Technology, Germany Received March 31, 2020; In final form March 14, 2021; Published June 2021 Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source are credited. Corresponding author: Oleksandr Vasilevskyi, e-mail: o.vasilevskyi@gmail.com 1. INTRODUCTION There is a need to develop methods for estimating the uncertainty of dynamic measurements that meet international requirements for evaluating and expressing the quality of measurements, which is a topical scientific task in the field of metrology. International standards support the use of combined uncertainty uc(y) as a parameter for quantifying the uncertainty of the measurement result [1]-[4]. Although uc(y) is widely used to express uncertainty, in some cases, such as in trade, industry, regulatory acts or issues of health and safety, it is advisable to further specify a measure of uncertainty that determines the interval for the measurement result. This additional measure of uncertainty, which corresponds to interval estimation, is called expanded uncertainty. It is therefore relevant to determine a particular measurement technique to evaluate the trust level for calculating expanded measurement uncertainty. The purpose of the article is to develop a mathematical apparatus for estimating the confidence level in calculating the expanded uncertainty of measurements of ion activity, taking into account the manufacturer's (developer's) and consumerβs control errors, which will make it possible to establish the interval around the measurement result within which the majority of the distribution of values can be attributed to the measured value. 2. MAIN MATERIALS OF THE RESEARCH In the literature [1]-[9], only partial consideration is given to methods of establishing a trust level for calculating the expanded uncertainty of measurement. A mathematical apparatus that would allow a reasonable confidence level in the uncertainty measurement is not described. It is therefore advisable to propose and describe a methodology for estimating trust level based on the control errors of the manufacturer and the consumer, which will make it possible to establish the value of the coefficient of coverage k for the calculation of the expanded uncertainty of measurement. As an example, the developed means was used to measure the activity of ions. The confidence level for calculating the expanded uncertainty is proposed based on the control errors of the manufacturer and the consumer using the formula π· = 1 β ππ = 1 β (πΌ + π½), (1) Where πΌ represents the control errors of the manufacturer, π½ represents the control errors of the consumer and ππ represents the total value of control errors. The procedure for determining the confidence level is described on the basis of the combined uncertainty of the results ABSTRACT A method for estimating the level of confidence when determining the coverage factor based on control errors is proposed, using the example of measurements of ion activity. Using information on tolerances and uncertainty, it is possible to establish a reasonable interval around the measurement result, within which most of the values that can be justified are assigned to the measured value. mailto:o.vasilevskyi@gmail.com ACTA IMEKO | www.imeko.org June 2021 | Volume 10 | Number 2 | 200 of reusable measurements of ion activity using a measuring instrument built on the principle of converting voltage to frequency [10], [11], which is described by the following transformation equation: πU/F = πpow π0 π (π0 β πΌt(273.15 + π‘) ππ πππ ) π2 (2) where ΡΠ₯Ρ is the activity of the ions; πΌt is the temperature coefficient of steepness, equal to 198.4 Γ 10β3 Β°Cβ1; t is the temperature of the medium under investigation, k2 is the gain scaling factor (DA1 and DA2 in Figure 1), Upow is the value of the reference voltage of the voltage converter at a frequency of 10 V; Ο = R C is the time constant of the voltage converter at a frequency that is used to set the full scale output frequency range (R = 1 kΞ©, Π‘ = 47 Β΅F), f0 is the frequency of the quartz resonator microcontroller (20 ΠHz), U0 is the standard potential of the reference electrode at the initial isopotential point and na is the charge of the ith ion [11]. The electrical functional circuit of the ion activity measuring device based on the conversion of voltage to frequency is shown in Figure 1. The general law of uncertainty when measuring ion activity depends on many factors, such as the activity of interfering ions due to the limited properties of ion-selective electrodes or the presence of measurement error due to temperature, zero drift, instability of the power supply, etc., and it can be difficult to identify which factor is dominant [10]. This allows us to adopt the law of distribution for the centred value of the error when measuring the normal activity of ions, which we describe with the expression π(ππ) = 1 π’π (ππ)β2π e β (ππβποΏ½ΜοΏ½)2 2 π’π 2(ππ) (3) where ποΏ½Μ οΏ½ is the estimated value of ion activity and uc(pX) is the combined uncertainty of the measurement of ion activity. The value of the combined uncertainty of ion activity measurement when using a measuring device built on the basis of the conversion of voltage to frequency is calculated by the formula π’π 2(pX) = π’π΄ 2(ππ) + π’ππ΅ 2 , (4) π’ππ΅ 2 = β ππ 2π’π 2(ππ) π π=1 + +2 β β ππ ππ π(ππ, π‘)π’π (ππ)π’π (π‘) π π=2 πβ1 π=1 , (5) where uA(pX) is the evaluation of type A standard uncertainty, ui(pX) represents components of type B uncertainty for measurements of ion activity with the respective sensitivity coefficients Ρi, uj(t) represents components of type B uncertainty for measurements of temperature with the corresponding sensitivity coefficients cj and r(pX, t) is the correlation coefficient between ion activity (pX) and temperature (t) [10]-[14]. Taking the transformation equation (2) into account, the type B combined uncertainty is determined by the formula π’ππ΅1 2 = ( ππU/F ππππ ) 2 [π’π΅πΈ 2 + π’π΅π/πΉ 2 ] + (6) + ( ππU/F ππ0 ) 2 π’π΅π©π0 2 + ( ππU/F ππ‘ ) 2 π’π΅π 2 + + ( ππU/F πππππ€ ) 2 π’π΅πpow 2 , π’ππ΅ 2 = π’ππ΅1 2 + 2 ππU/F ππππ π’ππ΅1 πππ‘U/F ππ‘ π’ππ΅π‘ Γ Γ β (π‘ππ β π‘π )(ππππ β πππ ) πππ π=1 ββ (π‘ππ β π‘π ) 2πππ π=1 β (ππππ β πππ ) 2πππ π=1 (7) where π π‘π/πΉ = 4 πpow π0 π π1 πΈπ πΌπ‘ π‘ is the equation of transformation for measurements of temperature (Figure 1) [10], Es = Π R0 is the sensor supply voltage (R0 is the sensor resistance at a temperature of 0 Β°Π‘) and k1 is the gain scaling factor (DA3 in Figure 1), ππU/F ππππ = πΌπ‘ (273.15 + π‘) πpow π0 π [π0 β πΌπ‘ (273.15 + π‘) ππ πππ ] 2 π2ππ = 12.98 ππ β1 is the sensitivity factor for ion activity at a temperature of 25 Β°Π‘, ππU/F ππ0 = β πpow π0 π [π0 β πΌπ‘ (273.15 + π‘) ππ πππ ] 2 π2 = β0.22 Vβ1 is the coefficient of the voltage sensitivity of the standard potential of the reference electrode, ππU/F ππ‘ = πΌπ‘ πππ πpow π0 π [π0 β πΌπ‘ (273.15 + π‘) ππ πππ ] 2 π2ππ = 0.3 Β°Cβ1 is the coefficient of sensitivity at an additional measured temperature, ππU/F ππpow = π0π (π0 β πΌπ‘ (273.15 + π‘) ππ πππ ) π2 = β9.08 Vβ1 is the coefficient of the voltage sensitivity of the stable power supply source of the voltage-to-frequency converter (VFC), nij is the number of pairwise measurements of temperature and ion activity, π’π΅πΈ = πΎ ππmax 100%β3 β3 pX is the type B uncertainty caused by the presence of a primary electrode with a low accuracy class (ο§ = 0.7 %), π’π΅π©π0 = π©π0 β3 = 0.95 mV is the type B uncertainty caused by the instability (π©π0 = 1.65 mV) of the standard potential of the reference electrode, π’Ππ = π₯t Ubiasmax πPoutβ3 β 2.89 β 10β3 Β°Π‘ is the type B uncertainty caused by the existence of a bias voltage (Ubiasmax πPoutβ = 0.2Β·10 β3) of the operational amplifier when the temperature deviates (π₯π‘ = 5 Β°Π‘) from the nominal value, π’π΅πPOW = π©ππ β3 β 1.16 mV is the type B uncertainty caused by instability in the reference voltage power supply source (π©ππ = 2 mV) and π’π΅VFC = πΏnonlππmax 100%β3 β6 pX is the type B uncertainty caused by the nonlinearity (πΏnonl) of the voltage converter (AD650) [10]-[14]. The combined uncertainty values are calculated in [10] and [14], and the maximum uncertainty value for measuring the ACTA IMEKO | www.imeko.org June 2021 | Volume 10 | Number 2 | 201 activity of phosphate ions is 17.21 Γ 10β3 pX in the range of measurements from 6 to 0.3 pX. A compatible two-dimensional confidence level density when measuring the activity of ions, taking into account the allowable deviation of the measurement error Ξ΅, which is established by the consumer, is described by the expression π(ππ, ) = 1 2 Ο β π’π (ππ) β π’ e β (ππβποΏ½ΜοΏ½)2 2 π’π 2(ππ) β (Ξ )2 2 π’π 2 . (8) In most practical cases, as noted in [9] and [15], the admissible deviation Ξ΅ of the parameter controlled by the user, depending on the combined uncertainty of the measurement results uc(y), is determined by the formula [%] = 6 π’π (π¦) π 100 = 6 π’π (π¦) πmax β πmin 100 , (9) where d is the width of the tolerance field determined by the values of the upper Xmax and lower Xmin range of measurement for the physical quantity. By substituting the values of the maximum combined uncertainty when measuring ion activity uc(pX) = 17.21Β·10β3 pX and the values of the upper (Xmax = 6 pX) and the lower (Xmin = 0.3 pX) ranges of the measurement limit in equation (9), we obtain the value of the permissible error set by the consumer, which is Ξ΅ = 1.81 %. The uncertainty uΞ΅ of the tolerance field of the controlled parameter at which the result can be considered reliable in practice is recommended to be equal to 6uc(y) [9], [13], [15]-[18]. Thus, uΞ΅ = 6 Γ uc(y) = 6 Γ 17.21Β·10β3 = 103.26Β·10β3 pX. Taking into account expression (8), the manufacturerβs control errors Ξ± are estimated by the formula πΌ = β« [β« e β (Ξππ)2 2 π’π 2(ππ) β (Ξ )2 2 π’π 2 2 Ο β π’π (ππ) β π’ ππππ ββ + ππππ₯ ππππ + β« e β (Ξππ)2 2 π’π 2(ππ) β (Ξ )2 2 π’π 2 2 Ο β π’π (ππ) β π’ ππ₯ β ππππ₯ ] (10) and the consumerβs control errors π½ are estimated by the formula π½ = β« β« e β (Ξππ)2 2 π’π 2(ππ) β (Ξ )2 2 π’2 2 Ο β π’π (ππ) β π’ πΞπππΞ + πmaxββππ πmin+βππ πmin ββ + β« β« e β (Ξππ)2 2 π’π 2(ππ) β (Ξ )2 2 π’2 2 Ο β π’π (ππ) β π’ πΞπππΞ πmaxββππ πmin+βππ β πmax . (11) The admission field is the value of the measured value ΞpX. In this case, this is the activity of the pX ions, which is defined by the formula RS 232 PC RD TD MCU MSP430F123 CA1 CA0 19 20 Vcc Vss 2 4 XOUT XIN 5 6 TA0 TA1 TA2 22 23 24 UTxD URxD 15 16 RST 7 + 5 V P2.0 P2.1 8 9 20 ΠΠΡ Vss Vdd + 5 V 2 9 5 H CLK DI H T 1 6 1 1 4 3 P3.0 P3.1 11 12 P3.3 P3.1 P3.2 P3.3 13 14C6 C7 DD4 1 DD5 pX U 3 2 4 + 5 V 7 6 100M LMC 6001 - 5 V 3 2 4 + 5 V 7 AD820 6 - 5 V 18k 6k 0 1 2 3 Π0 Π1 ΠGND 100k 100n 3 4 7 6 U/F AD650 2 1 8 5 + V in R Ct Ct F out LC Vcc Vss 561ΠΠ1 + 5 V Y DA2 DA4 DD2 R1 R8 R9 R3 R6 R12 R 10 C1 C3 T UT 3 2 4 + 5 V 7 AD820 6 R2 3 4 7 6 U/F AD650 2 1 8 5 + V in R Ct Ct F out LC Vcc Vss + 5 V DD1 R7 R5 R11 C2 C4 R4 R5 IN GND 1171 CΠ42 2 1 3 DD2 R13 C8 R6 DD3 DA1 DA3 C1 C5 Figure 1. The functional electrical circuit of the device for measuring ion activity. ACTA IMEKO | www.imeko.org June 2021 | Volume 10 | Number 2 | 202 Ξππ = ποΏ½ΜοΏ½ 100 πΏmax , (12) where ποΏ½Μ οΏ½ is the estimation of the ion activity in the upper measurement range, πΏmax is the maximum relative measurement error (for measurements of ion activity, this value is 0.7 % in the range of measurements from 6 to 0.3 pX). Consequently, when calculating the control errors of the manufacturer and the consumer, the field of admission βpX can be calculated by (12), which, with the corresponding numerical values, is equal to 2.1Β·10β3 pX. By substituting calculated tolerances in the formula for estimating the control errors of the manufacturer (10) and the consumer (11) and solving them using the Maple 12 package, we obtain the following numerical values: πΌ = 0.24Β·10β67 (πΌ β 0), and π½ = 0.00196. The total value of the control errors is Pn = 0.00196, and the confidence level for calculating the expanded uncertainty of the measurement, according to formula (1) will be equal to D = 1 β Pn = 1 β 0.00196 = 0.998. Figure 2 illustrates the change in the manufacturerβs and consumerβs control errors depending on the parameter ΞΌ, which is equal to uΞ΅/uc(pX). This parameter establishes the relationship between the permissible uncertainty of uΞ΅, which is specified by the consumer (via normative documents), and the combined uncertainty of uc(pX), which is set by the manufacturer. The figure shows uΞ΅ < uc(pX), uΞ΅ β uc(pX) and uΞ΅ > uc(pX). Thus, based on the constructed characteristic of the change in reliability (Figure 2), which is obtained by calculating the consumer and manufacturerβs control errors, a range can be calculated within which the majority of the distribution of values that can be attributed to the measured value is likely to be located, depending on the value of the accepted tolerance for the monitored parameter. As can be seen from Figure 2, with a tolerance of six combined uncertainties, the confidence level is 99.8 %. Given the level of confidence based on control errors, the expanded uncertainty can be calculated from the formula π = ππ β π’π (π¦) = π99.8 β 17.21 β 10 β3 , (13) where kp is the coverage ratio for the established confidence level (p = D), which is taken from the Student's table. Thus, with 30 degrees of freedom and a confidence probability of 99.8 %, the coverage factor kp is 3.385, and with 10 degrees of freedom and a probability of 99.8 %, it is 4.14. Accordingly, if these coverage factor values are entered into (13), the expanded uncertainty will be U99.8 = 71.25Β·10-3 pX for 10 degrees of freedom and U99.8 = 58.26Β·10β3 pX for 30 degrees of freedom. 3. CONCLUSIONS The work describes the characteristics of the change in the manufacturerβs and consumerβs control errors, on the basis of which a general characteristic of the change in the control errors that occurs during the measurement of ion activity is obtained. Using this characterisation, it is possible to determine graphically the confidence level within which most of the distribution of values obtained by measuring the activity of ions is probably located. The obtained nomograph makes it possible to determine with a high probability the confidence level at a given tolerance for the controlled parameter to determine the coverage factor in calculating the extended uncertainty of ion activity measurement. The described approach for determining confidence probability on the basis of control errors can be applied to any type of measurements, provided a separate calculation is performed to characterise the control errors of the manufacturer of the measuring instrument and the consumer (i.e. errors of the first and second kind). Figure 2. Characteristics of the change of control errors depending on the parameter ΞΌ. ACTA IMEKO | www.imeko.org June 2021 | Volume 10 | Number 2 | 203 REFERENCES [1] BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML, Evaluation of measurement data β guide to the expression of uncertainty in measurement, Joint Committee for Guides in Metrology, Bureau International des Poids et Mesures JCGM 100:2008 (2008). [2] BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML, Evaluation of measurement data β supplement 2 to the 'Guide to the expression of uncertainty in measurement' β extension to any number of output quantities, Joint Committee for Guides in Metrology, Bureau International des Poids et Mesures JCGM 102:2011 (2011). [3] BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML, Evaluation of measurement data β an introduction to the βGuide to the expression of uncertainty in measurementβ and related documents, Joint Committee for Guides in Metrology, Bureau International des Poids et Mesures JCGM 104:2009 (2009). [4] ISO/IEC, Uncertainty of measurement β part 1: introduction to the expression of uncertainty in measurement, ISO/IEC Guide 98-1:2009 (2009). [5] O. M. Vasilevskyi, Calibration method to assess the accuracy of measurement devices using the theory of uncertainty, Int. J. Metrol. Qual. Eng. 5(4) (2014) 403, 9 pp. DOI: 10.1051/ijmqe/2014017 [6] IEC, Application of uncertainty of measurement to conformity assessment activities in the electrotechnical sector, IEC GUIDE 115-2007 (2007). [7] ISO/IEC, General requirements for the competence of testing and calibration laboratories, ISO/IEC 17025:2005 (2005). [8] O. M. Vasilevskyi, Methods of determining the recalibration interval measurement tools based on the concept of uncertainty, Technical Electrodynamics 6 (2014) pp. 81-88. [9] S. F. Beckert, W. S. Paim, Critical analysis of the acceptance criteria used in measurement systems evaluation, Int. J. Metrol. Qual. Eng. 8 (2017) art. 23. DOI: 10.1051/ijmqe/2017016 [10] O. M. Vasilevskyi, V. M. Didych, Elements of the Theory of Constructing Potentiometric Means of Measuring Control of Ion Activity with an Increased Probability [Monograph], Vinnytsia: VNTU, 2013. Online [Accessed 2 June 2021] http://vasilevskiy.vk.vntu.edu.ua/file/d23a8f23c4fdc081c30ebb0 9bc7182e6.pdf [11] V. M. Didych, O. M. Vasilevskyi, V. O. Podzharenko, Potentiometer facilities of ions activity measurement of humus elements in soil, Visnyk of Vinnytsia Politechnical Institute 5 (2008) pp. 5-10. Online [Accessed 2 June 2021] https://visnyk.vntu.edu.ua/index.php/visnyk/article/view/639/ 638 [12] O. M. Vasilevskyi, Metrological characteristics of the torque measurement of electric motors, Int. J. Metrol. Qual. Eng. 8 (2017) 7, 9 pp. DOI: 10.1051/ijmqe/2017005 [13] O. M. Vasilevskyi, V. M. Didych, A. Kravchenko, M. Yakovlev, I. Andrikevych, D. Kompanets, Y. Danylyuk, W. WΓ³jcik, A. Nurmakhambetov, Method of evaluating the level of confidence based on metrological risks for determining the coverage factor in the concept of uncertainty, Proceedings Volume 10808, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments, WUT Wilga Resort, Poland, 2 β 10 June 2018. DOI: 10.1117/12.2501576 [14] O. M. Vasilevskyi, V. Y. Kucheruk, V. V. Bogachuk, K. Gromaszek, W. WΓ³jcik, S. Smailova, N. Askarova, The method of translation additive and multiplicative error in the instrumental component of the measurement uncertainty, Proc. SPIE 10031, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments, WUT Wilga Resort, Poland, 29 May β 6 June 2016. DOI: 10.1117/12.2249195 [15] D. J. Wheeler, An honest gauge R&R study, 2006 ASQ/ASA Fall Technical Conference, Columbus, US, 12 β 13 October 2006. Online [Accessed 2 June 2021] http://www.spcpress.com/pdf/DJW189.pdf [16] O. M. Vasilevskyi, P. I. Kulakov, I. A. Dudatiev, V. M. Didych, A. Kotyra, B. Suleimenov, A. Assembay, A. Kozbekova, Vibration diagnostic system for evaluation of state interconnected electrical motors mechanical parameters, Proc. SPIE 10445, Photonics Applications in Astronomy, Communications, Industry, and High Energy Physics Experiments, WUT Wilga Resort, Poland, 29 May β 6 June 2016. DOI: 10.1117/12.2280993 [17] O. M. Vasilevskyi, V. Y. Kucheruk, V. V. Bogachuk, K. Gromaszek, W. WΓ³jcik, S. Smailova, N. Askarova, The method of translation additive and multiplicative error in the instrumental component of the measurement uncertainty, Proc. SPIE 10031, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments, WUT Wilga Resort, Poland, 29 May β 6 June 2016. DOI: 10.1117/12.2249195 [18] V. O. Podzharenko, O. M. Vasilevskyi, Diagnostics of technical condition of electromechanical systems for the logarithmic decrement, Proceedings of Donetsk National Technical University (2005). https://doi.org/10.1051/ijmqe/2014017 https://doi.org/10.1051/ijmqe/2017016 http://vasilevskiy.vk.vntu.edu.ua/file/d23a8f23c4fdc081c30ebb09bc7182e6.pdf http://vasilevskiy.vk.vntu.edu.ua/file/d23a8f23c4fdc081c30ebb09bc7182e6.pdf https://visnyk.vntu.edu.ua/index.php/visnyk/article/view/639/638 https://visnyk.vntu.edu.ua/index.php/visnyk/article/view/639/638 https://doi.org/10.1051/ijmqe/2017005 https://doi.org/10.1117/12.2501576 http://vasilevskiy.vk.vntu.edu.ua/file/ref/bf7f9857715c0be9f72e906112b07474.pdf http://vasilevskiy.vk.vntu.edu.ua/file/ref/bf7f9857715c0be9f72e906112b07474.pdf https://doi.org/10.1117/12.2249195 http://www.spcpress.com/pdf/DJW189.pdf https://doi.org/10.1117/12.2280993 http://vasilevskiy.vk.vntu.edu.ua/file/ref/bf7f9857715c0be9f72e906112b07474.pdf http://vasilevskiy.vk.vntu.edu.ua/file/ref/bf7f9857715c0be9f72e906112b07474.pdf https://doi.org/10.1117/12.2249195