Gas flow temperature is an important parameter of the process, determines the quantitative and qualitative indicators of the original product, the presence of defects, the state of technological equipment, as well as the safety of the process. Therefore, its measurement must be carried out continuously, with high accuracy, low inertia and high reliability, since the information signal about the value of temperature is used in information-measuring systems and automatic control and regulation systems.
At present, the problem of energy efficiency remains extremely important. Modern building technologies allow you to create houses with the minimum power consumption, using energy efficient external protections, including plastic windows. This leads to a reduction in the heat loss of the room, but there is a danger of reducing the required air exchange.
Investigations of methodical errors of two-color compensative and classical energy and spectral ratio pyrometry are performed under conditions of changeable radiative characteristics of metal alloys. To quantitatively estimate the radiative characteristics we proposed such parameters as an average level of emissivity and selectivity factor. As average (adjustment) values of these parameters we chose the values, which correspond to tungsten in vacuum with temperature 1600 K (for wavelengths 0.7 and 0.9 µm).
The successful implementation of the systems of automation in metallurgical industry is impossible without the presence of reliable primary information about the parameters of technological process. A temperature is one of basic parameters that determine the quantitative and quality indicators of the final products. Therefore the technological processes of metallurgical industry need the presence of various primary transformers of temperature with high accuracy, sensitivity, stability and resistance to interference.
Study of the dependence of the temperature determination error on the emissivity factor of materials is conducted in the paper. The mathematical models, which describe the ratio of thermodynamic temperature and measured imaginary temperatures, taking into account the emissivity factor, are analyzed. The constructions of the full radiation, brightness radiation, and spectral ratio radiation pyrometers are underpinned by the considered models.
The current article describes the results of the study of the neural networks temperature prediction error dependence on measurement errors, which are random, nonlinear and multiplicative errors. It is noted applicability of the architecture of neural network for temperature prediction. The formula of temperature step response for ideal sensor is given.
The choice of ways of obtaining information about the temperature of the surface layers of a moving object is the crucial in determining the type of temperature transducers. This choice depends on the row of specific factors: the speed of movement of the measured object relative the object, the state of the surface, the presence of related agents in near-surface layers, the presence of disturbing factors - noise, vibration, etc. The major factor seems to be the value of the heat carried out from the diagnosed object with the help of intermediate substance.
Two main methods of contact measurement of steel temperature are applied in industry: with help of disposable and multiple used thermotransducers. Reusable ones are fixed to the bottom of the graphite tip fixed directly on the reinforcement. Then they are able to measure the temperature of the graphite surface. Since it needs to even the graphite temperature and the temperature of the melt metal, the measuring takes some extra time. Therefore the thermotransducer is mounted in the metal for a long time.
Current article considers the results of the study of air and water flow temperature prediction error on the number of inputs in neural network. Authors guide the architecture of neural network for temperature prediction. The formula of temperature step response for real sensor is given. Also, the method for calculating the time constants for the temperature step response formula using real measurement data is considered.
While studying the physical foundations of the temperature standard, we obtained a quantum unit of temperature as the value of the temperature jump when one electron-phonon scattering per unit time. We expressed it in terms of the ratio of fundamental physical constants h/kB; it is equal to 3.199 493 42 ∙ 10-11 K with a relative standard uncertainty of 59.2 ∙ 10-8. The investigated quantum standard is recommended for use as an "intrinsic standard", which does not require continuously repeated measurements (to check its accuracy) in relation to the current unit of temperature.