Posted 30 Dec 2010 11:51
Measurement uncertainty and error is the metrology of the basic proposition is frequently used measurement testers important concept. It has a direct bearing on the reliability of measurement results and the exact same value transfer. However, the concept is unclear because many people, it is easy to confuse the two or misused, this paper study "Evaluation and Expression of Uncertainty in Measurement," the experience, focus on a difference between the two. Is the first clear error of measurement uncertainty and the conceptual differences between the two.It is the confidence probability is given by a true value may fall into the interval. It can be in multiples of standard deviation, or confidence level that the half width of the interval. It is not really a specific error, it only is a form of quantitative parameters that can not fix that part of the range of error. It comes from the accidental effect and the effect of the imperfect correction system is used to characterize the reasonable value given the dispersion of the measured parameters. Uncertainty according to their access method is divided into A, B two types of assessment components. A class component is assessed through statistical analysis of observations made out of Uncertainty, B class assessment component is based on experience or other information to estimate and assumes the existence of approximate characterization of the standard deviation of the uncertainty components. In most cases the error is the measurement error, which is the traditional definition of measurement results and the difference between the true value being measured. Can generally be divided into two categories: systematic error and random error. Error is the objective, it should be a definite value, but because in most cases, the true value is not known, so the true errors can not know exactly. We only under certain conditions to find the best approximation of the true value and true value is called agreement. Through the understanding of the concept, we can see that measurement uncertainty and the measurement error of the main differences are the following: an. Objective assessment of the difference: Measurement uncertainty is that the dispersion of the measured values; measurement error measurement results is that the degree of deviation from the true value. II. The difference between evaluation results: measurement uncertainty is unsigned parameters, with standard deviation or standard deviation or confidence interval in multiples of half-width, said by the people under the experimental, data, experience and other information to assess, through A, identified two types of quantitative evaluation method B; measurement error are plus or minus value, the value of measurement minus the true value being measured, the true value is unknown, they can not get accurate, true value when used instead of the agreed true value, only the estimated value can be obtained. III. The difference between factors: measurement uncertainty through the analysis and assessment by the people are, and therefore people are measured, and measurement of the amount of knowledge about the process; measurement errors is an objective reality, free from outside factors, not people change the level of awareness; Therefore, when uncertainty analysis should take into account all factors, and to verify the assessment of uncertainty. Otherwise, the analysis underestimated the measurement results may be very close to the true value (ie, very small errors) in the case of uncertainty was assessed by large measurement errors may in fact larger in the case, not given chosen to determine the degree of small. IV. Distinguished by the nature of the difference: measurement uncertainty when uncertainty was assessed by their nature generally do not distinguish, if the need to distinguish be expressed as: random effects introduced by the uncertainty components and the uncertainty introduced by the systematic effect degree of weight; measurement errors can be divided according to the nature of random error and systematic errors into two categories, by definition, random error and systematic errors are infinite times the ideal case, the concept of measurement. Five. The difference between the measurement results of correction: the word itself implies uncertainty can be estimated as a value, it is not that specific, the exact error, though can be estimated, but not used to correct values, only measurement results can be fixed to consider amending the uncertainty in the uncertainty introduced by imperfect; and systematic error estimate of the measurement if the results are known can be corrected, and the measurement results has been fixed. A measure, as amended, may be more close to the true value, but not only did not reduce the uncertainty, but sometimes even more. This is mainly because we can not know exactly how much real value, only the measurement results toward or away from the true value to estimate the extent of it. Although the measurement uncertainty and error with all these different, but they remain in close contact. The concept of uncertainty in the application and development of error theory, and error analysis of measurement uncertainty is still the theoretical basis of the assessment, in the estimation of B class components, the error analysis is even more indispensable. For example, the characteristics of measuring instruments can be the maximum permissible error, error of indication and other terms described. In the technical specifications and the measuring instruments specified tolerance limits, known as the maximum permissible error or allowable error limit. It is the manufacturer of the equipment required by a model error of indication of the allowed range, not the actual existence of an instrument error. Maximum permissible error of measurement instruments can be found in the instrument manual, that when the numerical sign, usually absolute error, relative error, refer to the error or their combination, said. Such as earth 0.1PV, soil 1% and so on. Maximum permissible error of measuring instruments is not uncertainty, but uncertainty can be used as a basis for measurement. Measurements introduced by the uncertainty of measuring instruments according to the maximum permissible error of the instrument evaluation method of assessment by B class. Another example is the value of measuring instruments indicated agreement with the corresponding input difference between the true value, the indication for the error of measuring instruments. For the physical gauge, indication is its nominal value. Usually high level of measurement of a standard or complex is provided by the value, true value as agreed upon (often called calibration value or standard value.) In the test work, when the measure given the expanded uncertainty of the standard value were seized equipment to the maximum permissible error of 1 / 3 to 1 / 10, and was seized indication error of the instrument specified within the maximum permissible error , can be judged as qualified.