|
1, the error generation
Error into random error and systematic error.
Error can be expressed as: error = measured - true value = random error + systematic error.
Therefore, any error can be decomposed into systematic errors and random errors: As the measurement tool (or measuring instruments) the inherent error, measurement principle or measurement method itself, the theory of defects, experimental and laboratory staff in the psychological constraints brought about by physiological conditions measurement error is called systematic errors.
Systematic error is characterized by the same measurement conditions, the repeated measurements from the measurement results are always too large or too small, and the error value, or according to certain rules must change. Reduce the system error method usually can change measurement tools or measurement methods, measurement results can also be considered correct value.
Random error: random error also known as random error, even if complete elimination of systematic errors in this ideal case, the repeated measurements the same measurement object, still due to various accidental, unpredictable uncertainties arising from interference with measurement error, known as random error.
Random error is characterized by the same measurement object repeated measurements, the error of measurement results obtained show irregular fluctuations, both may be positive (the measurement results is too large), may also be negative (small measurements), and the absolute error value fluctuations without rules. However, the distribution of errors subject to statistical laws, to demonstrate the following three features: a single peak, that error is small than big error; symmetry, that is equal to the probability of error and negative error; bounded, that is a great error probability is almost zero.
We can see from the distribution of random errors, increase the number of measurements, according to statistical theory dealing with the measurement results can reduce the random error.
2, precision, precision and accuracy
With the same measurement tools and methods multiple measurements under the same conditions, if the measured value random error is small, the results of each measurement, fluctuations in small, indicating good measurement repeatability, measurement accuracy is good, also known as stability is good, therefore, accidental error of measurement reflects the size of the measurement precision.
According to error theory shows that when measuring the number of infinite number of circumstances, can make the random error tends to zero, while the measurements obtained with the deviation from the true value - the measurement accuracy, will fundamentally depend on the size of systematic errors, which reflects the size of systematic errors may reach the accuracy of measurement.
Accuracy is a measure of accuracy and precision of the general, in the actual measurement, may affect the accuracy of the main systematic error, random error may be largely, of course, may affect both the measurement accuracy must not be ignored. In some measuring instruments, used the concept of precision, in fact, the system includes both error and random error, such as commonly used in precision instruments often divided on the instrument rating.
Accuracy referred to as precision instrumentation, also known as accuracy. Accuracy and error can be said to twins, because there are errors there, only the concept of accuracy. Meter accuracy in short, measured close to the instrument accuracy of the true value, usually a relative percentage error (also known as Xiangdui equivalent error) said.
|
