A multispectral radiation thermometry platform

based on wireless data transmission

KAIHUA ZHANG,1 KUN YU,1 AND YUFANG LIU 1, 2, 3
1Henan Key Laboratory of Infrared Spectrum Measures and Applications, School of Physics, Henan Normal University, Xinxiang, Henan 453007,
PR China
2 Institute of Physics, Henan Academy of Sciences, Zhengzhou, Henan 450046, PR China
[email protected]

Received XX Month XXXX; revised XX Month, XXXX; accepted XX Month XXXX; posted XX Month XXXX (Doc. ID XXXXX); published XX Month XXXX

In metal smelting, precise temperature control is of vital address the uncertainty caused by emissivity variations [5].
importance for reaction rates, efficiency, and product Brightness methods are greatly influenced by background thermal
quality. Traditional methods such as thermocouples have radiation and medium absorption, while colorimetric methods are
inherent limitations, but multispectral radiation stringent in the selection of two temperature measurement bands.
thermometry (MRT) offers high resolution and reliability. Multispectral radiation thermometry technology, achieving
This paper proposes a multispectral radiation temperature inversion independent of target emissivity, has
thermometry platform featuring wireless data experienced rapid advancement [6-10]. Researchers have
transmission, which enables remote data transfer and proposed algorithms that assume emissivity models as a means to
precise temperature measurements. The platform was overcome the challenge of unknown emissivity [11-16]. These
meticulously calibrated and six common emissivity models typically include wavelength emissivity models [17] or
models were inverted with high accuracy. The results of temperature emissivity models [18]. While assuming emissivity
temperature measurements conducted at a copper models resolves the issue of temperature determination in the
smelting site demonstrated an excellent degree of presence of unknown emissivity, their applicability is often limited
agreement with those obtained using disposable to specific materials within specific environments. To address the
thermocouples. The platform has the potential to be limitations of assuming emissivity models, researchers have
applied in harsh environments, offering a novel approach explored methods for processing multispectral radiation
to temperature measurement in metal smelting thermometry data without assuming specific emissivity models.
processes. These methods include the use of multi-wavelength emissivity
constraint algorithms [19], generalized inverse matrix with an
external penalty function [20], and direct processing algorithms
In metal smelting, temperature serves as a pivotal parameter for [21]. While these methods are effective to some extent, they are
controlling reaction rates, enhancing production efficiency, and sensitive to deviations between assumed and actual emissivity
ensuring product quality [1]. Accurate temperature measurement models. Other methods based on similarity between measured and
within the smelting furnace enables precise process control, crucial blackbody spectral radiance intensity [22] or moving narrowband
for improving production efficiency, safeguarding product quality, spectral windows [23] have limitations in terms of proximity to the
and ensuring production safety [2]. object's true temperature range and real-time temperature
Currently, widely employed temperature measurement measurement. In recent years, neural network technology has
methods in metal smelting processes include thermocouple provided a new direction for processing multispectral radiation
thermometry and radiation thermometry [3, 4]. The advantage of thermometry data. However, these methods often require a large
thermocouple thermometry lies in its simple structure and number of data samples and long training times [24]. Research on
measurement of the true temperature of the object. However, it temperature measurement algorithms has enhanced the
suffers from poor dynamic response, limited temperature applicability and accuracy of multispectral radiation thermometry.
measurement range, high consumption, and inability to measure However, devices used for spectral measurement and temperature
temperature variations in real-time during the smelting process. inversion typically have large volumes, posing significant challenges
Radiation thermometry is characterized by high resolution, for practical application in metal smelting sites.
sensitivity, reliability, short response time, unlimited temperature This paper presents a multispectral radiation thermometry
measurement, non-contact, and non-destructive measurement platform based on wireless data transmission, which separates
advantages, holding a significant position in the field of temperature spectral measurement from data processing. This enhances the
measurement in metal smelting. However, currently prevalent applicability of the platform in metal smelting sites. The
brightness and colorimetric pyrometers fail to fundamentally temperature measurement platform was meticulously calibrated,
and temperature validation experiments were conducted in the earlier publication [25]. The computer computes the average
laboratory. Finally, the device was applied for temperature temperature inversion and ultimately outputs the inferred
measurement at a copper smelting site to verify its applicability and temperature of the target.
measurement accuracy. As the output of the fiber optic spectrometer is voltage-related
The schematic diagram of the multispectral radiation electrical signal data, calibration of the spectrometer is necessary
temperature measurement platform is shown in Fig. 1. The prior to temperature inversion of the data obtained. This calibration
platform comprises four main components: an optical signal process involves converting the output electrical signal data to a
acquisition module, a multispectral signal transmission module, an data type related to the radiant energy of the object under
ATK-ESP8266 WiFi transmission module, and the TLink cloud measurement. The calibration of the spectrometer necessitates the
platform. The optical signal acquisition module is responsible for utilization of a multi-temperature calibration method proposed in
capturing the thermal radiation signal of the measured object and our previous study [26].
converting it into electrical signals. The multispectral signal In order to employ the multi-temperature calibration method, it
transmission module is responsible for exporting the data collected is necessary to measure the spectral signals of the blackbody
by the fiber optic spectrometer. The ATK-ESP8266 WiFi furnace at multiple temperature points. As a case in point, we may
transmission module uploads the collected multispectral data to the cite the selection of the British ISOTECH high-temperature
TLink cloud platform, where it is imported into the data processing blackbody furnace, model R970. The blackbody furnace has an
equipment. effective emissivity exceeding 0.995 over a temperature range of
423 K to 1473 K. The temperature of the blackbody furnace is set by
a controller, while an independent indicator displays the actual
radiation temperature. The sensor of this independent indicator is
installed inside the blackbody chamber, which has a diameter of 20
mm and a depth of 65 mm. At temperature intervals of 50 K, ranging
from 873 K to 1173 K, the blackbody radiation signals were
measured, and subsequently, the spectral response curve and
background radiation curve of the MRT platform were calculated
using the multi-temperature calibration theory, as depicted in
Figure 2. Prior to the occurrence of temperature inversion, it is of
the utmost importance to convert the electrical signal data
outputted by the fiber optic spectrometer into radiation energy
Fig. 1. The schematic diagram of the MRT platform. signals associated with the target under measurement, based on the
aforementioned calibration results.
The sample is subjected to heating via a silicon carbide heating
element. This is encapsulated in a cylindrical cavity constructed
from glass fiber insulation material. The furnace surface is covered
with a high-temperature and high thermal conductivity alumina
ceramic plate, while the outer layer is covered with ceramic fiber
insulation material. The sample is placed on the alumina ceramic
plate, and heating is applied through heat transfer. The heating
system employs a high-precision proportional-integral-derivative
(PID) controller for temperature control, and the sample
temperature is monitored using an S-type thermocouple. The
maximum temperature of the furnace can reach 1473 K.
The optical signal acquisition module comprises an optical
system and a fiber optic spectrometer. The radiation emitted by the
target is initially focused by the concave mirror onto the right-angle
mirror, and then converged by the right-angle mirror into the fiber
optic port. Finally, the radiation is transmitted through the optical
fiber to the fiber optic spectrometer for measurement. The fiber
optic spectrometer employed in this setup is the NIR25S from Idea
Optics. The system incorporates a high-speed linear array InGaAs
detector as the optoelectronic sensor and integrates a step filter to Fig. 2. The response function and background radiation of the MRT
platform..
eliminate higher-order diffraction, enabling the acquisition of
radiation signals in the spectral range of 0.9-2.5 μm. The spectral In order to assess the time response characteristics and
radiation data of the target, measured by the fiber optic temperature inversion performance of the MRT platform,
spectrometer, is uploaded to the TLink cloud platform through the simulations were initially conducted employing six emissivity
data transmission and data transmitting modules. A remote models. Utilizing the Planck formula, spectral radiance values were
computer downloads the data in real-time and feeds it back to the computed for various temperatures based on these emissivity
data processing system for temperature inversion. The functions, as depicted in Figure 3. Subsequently, the calculated
temperature inversion method utilized in the temperature spectral data was transmitted to the Tlink cloud platform as a signal,
measurement platform adheres to the approach detailed in our where a remote computer downloaded the data and executed
temperature inversion for the six model materials utilizing a data wavelengths, considering factors such as measurement accuracy
processing method reliant on Euclidean distance optimization. This requirements and operational efficiency. Furthermore, while a
comprehensive analysis enabled an evaluation of the platform's greater number of wavelengths may enhance accuracy, it may also
capabilities across different emissivity scenarios, thereby introduce greater complexity in equipment setup and data
contributing to a deeper understanding of its performance under processing. Therefore, a comprehensive evaluation of various
various conditions. factors is necessary in practical temperature measurement
applications. This nuanced understanding aids in optimizing
temperature measurement processes to meet specific industrial
requirements effectively.
To validate the performance of a multispectral temperature
measurement platform under real-world conditions, a copper
smelting facility in China was selected as the testing ground. During
the experiments, the platform's front-end equipment was precisely
aligned with the liquid metal surface within an industrial copper
smelting crucible to capture real-time spectral data of the molten
copper. Continuous flushing of protective gas over the
measurement area was employed to mitigate oxidation of the
molten copper, ensuring data fidelity. Subsequently, temperature
spectral data were retrieved and processed at the backend with
minimal latency, facilitating near real-time temperature assessment.
Additionally, periodic measurements using industrial-grade
disposable thermocouples were performed to corroborate the
accuracy of the collected data. A comparative analysis between the
temperature platform's readings and those obtained from the
thermocouples, as depicted in Figure 5, was conducted. A
meticulous experimental design and thorough data analysis were
employed to conduct a comprehensive evaluation of the
multispectral temperature measurement platform's efficacy in
industrial copper smelting environments. This endeavor not only
underscores the platform's potential in industrial applications but
also highlights avenues for further enhancement and optimization.

Fig. 3. The calculated spectral data and emissivity of six models.

Fig. 5. Comparison of real-time temperature measurement data from the

MRT platform with disposable thermocouple measurements.
Fig. 4. The relative error and response time of inversion results for different
numbers of wavelengths. According to the data presented in Figure 5, the inverse data from
the temperature measurement platform shows good consistency
Figure 4 presents the relative error and response time of with the measurements from the disposable thermocouple,
inversion results for different numbers of wavelengths. It can be indicating that the temperature platform data meets the
observed that an increase in the number of wavelengths requirements for industrial temperature measurement. However,
corresponds to a gradual decrease in relative error, although this is there are certain deviations between the two, primarily due to
accompanied by an increase in response time. This trade-off several factors. Firstly, industrial sites often contain significant
emphasizes the critical importance of striking a balance between amounts of smoke and water vapor, and the temperature
accuracy and efficiency in temperature inversion processes. It measurement platform has not been adequately calibrated for
highlights the necessity of thoughtfully selecting the number of these real-world conditions. Future work should involve the
introduction of online calibration or elimination algorithms to 9. Y. Zhang, Z. Zou, and F. Yan, "A data processing algorithm for multispectral
enhance temperature accuracy. Secondly, the surface of molten radiation thermometry based on multi-segment linear model and secondary
copper is prone to oxidation. Although oxidized substances are inversion," Measurement 201, 111753 (2022).
swept away from the measurement area by the protective gas, the 10. L. Chen, X. Zhang, S. Gao, Y. Cui, C. Yang, X. Wei, J. Jiang, Y. Niu, and C. Wang,
short integration time of spectral acquisition may still allow the "Multi-spectral radiation thermometry based on an Alpha spectrum-LM
algorithm under the background of high temperature and intense reflection,"
radiation spectrum of oxidized substances to enter the optical path,
Optics express 30, 36603-36621 (2022).
thereby affecting the accuracy of the target radiation spectrum.
11. H. Liu, and S. Zheng, "Measurement of temperature and wavelength-
The development of a multispectral radiation thermometry dependent emissivity distributions using multi-wavelength radiation
platform based on wireless data transmission represents a thermometry," Optics Communications 472, 125895 (2020).
significant advancement in temperature measurement technology 12. L. Mengqi, Y. Mingqing, N. Chunhui, and L. Datong, "Development of
for metal smelting applications. By separating spectral portable high-speed multi-spectral radiation thermometer," Laser Technology
measurement from data processing, the platform offers enhanced 47 (2023).
accuracy and efficiency, overcoming limitations associated with 13. A. Araújo, and R. Silva, "Surface temperature estimation in determined
traditional temperature measurement methods. Calibration multi-wavelength pyrometry systems," Review of Scientific Instruments 91
experiments confirmed the platform's reliability, while validation (2020).
tests conducted at a copper smelting facility demonstrated its 14. S. Zheng, W. Cai, C. Zhao, D. Li, J. Zhang, and Q. Lu, "On the measurement
effectiveness in real-world operating conditions. Despite certain of flame temperature and emissivity based on multispectral imaging
deviations observed between the platform's readings and those technique," Measurement 196, 111272 (2022).
15. K. Yu, H. Guo, K. Zhang, Y. Liu, and Y. Liu, "Multi-wavelength radiometric
obtained from thermocouples, the platform's performance met the
thermometry data processing algorithm based on the BFGS algorithm,"
requirements for industrial temperature measurement. Future
Applied optics 60, 1916-1923 (2021).
research directions include refining calibration algorithms and 16. S. Gao, C. Zhao, L. Chen, J. Jiang, P. Yu, Z. Zhang, and C. Wang, "Error analysis
optimizing spectral measurement techniques to further enhance and reflection correction for radiation temperature measurements at high
the platform's accuracy and applicability in metal smelting background temperatures," Measurement science and technology 32,
environments. 055003 (2021).
Funding. National Natural Science Foundation of China (U23A20377), 17. C. D. Wen, "Investigation of steel emissivity behaviors: Examination of
Henan Provincial Joint Fund for Science and Technology Research and Multispectral Radiation Thermometry (MRT) emissivity models," International
Development Program (225200810077), Program for Innovative Research Journal of Heat & Mass Transfer 53, 2035-2043 (2010).
Team (in Science and Technology) in University of Henan Province 18. S. Zhao, X. Li, X. Zhou, K. Cheng, and X. Huai, "Investigation of the effects of
(23IRTSTHN013), and Natural Science Foundation of Henan province Ni-based alloys DD6 on the normal spectral emissivity during oxidation," in
(222300420011, 222300420209). Asian Joint Workshop on Thermophysics and Fluid Science(2016).
19. J. Luo, B. Sun, X. Sun, and K. Xue, "Emissivity constraints for calibration
Disclosures. The authors declare no conflicts of interest. constant models of multi-wavelength pyrometers," Infrared Physics &
Data availability. Data underlying the results presented in this paper are Technology 127, 104408 (2022).
not publicly available at this time but may be obtained from the authors upon 20. J. Xing, Z. Liu, J. Luo, and B. Han, "Generalized inverse matrix normalization
reasonable request. algorithm to extract high-temperature data from multiwavelength
pyrometry," Review of Scientific Instruments 91 (2020).
21. W. Wójcik, V. Firago, A. Smolarz, I. Shedreyeva, and B. Yeraliyeva,
References "Multispectral High Temperature Thermography," Sensors 22, 742 (2022).
1. B. Li, J. Du, Y. Sun, S. Zhang, and Q. Zhang, "On the importance of heat source 22. X. Liu, X. Hao, B. Xue, B. Tai, and H. Zhou, "Two-dimensional flame
model determination for numerical modeling of selective laser melting of temperature and emissivity distribution measurement based on element
IN625," Optics & Laser Technology 158, 108806 (2023). doping and energy spectrum analysis," Ieee Access 8, 200863-200874 (2020).
2. W. Shi, X. Wu, J. Bai, L. Kong, X. Li, J. Cao, and W. Li, "Determination of 23. Y. Zhao, J. Lv, K. Zheng, J. Tao, Y. Qin, W. Wang, C. Wang, and J. Liang,
running temperature in entrained flow gasifier in regard to coal ash melting "Multispectral radiometric temperature measurement algorithm for turbine
thermal kinetic effects," Fuel 339, 127458 (2023). blades based on moving narrow-band spectral windows," Optics express 29,
3. S. J. Torres, F. Rivas, F. Perez, R. Parra, S. Torres, and S. E. Godoy, 4405-4421 (2021).
"Radiometric methods for high temperature flame estimation: robustness 24. H. Huang, Z. Zhang, Y. Niu, C. Zhang, and J. Jiang, "Review of multispectral
analysis and application to an industrial flash smelting furnace," IEEE radiation thermometry data processing algorithms," Infrared Physics &
Transactions on Instrumentation and Measurement (2023). Technology 129, 104573 (2023).
4. M. Can, C. Gözönünde, N. Arifoviç, F. Yıldız, and H. Nasibov, "Large-area Fe–C 25. K. Zhang, Y. He, K. Yu, and Y. Liu, "Fast data processing method for
eutectic fixed-points for radiation and contact thermometry," Measurement multispectral radiation thermometry based on Euclidean distance
science and technology 34, 124001 (2023). optimization," Optics express 32, 1342-1356 (2024).
5. J. Ji, M. Ding, B. Wang, and C. Sun, "Error Analysis and Correction of Radiation 26. K. Zhang, and Y. Liu, "Modified two-temperature calibration method for
Pyrometer in Measuring Surface Temperature of Turbine Blades with the emissivity measurements at high temperatures," Applied Thermal Engineering
Influence of Reflection," IEEE Transactions on Instrumentation and 168, 114854 (2020).
Measurement (2023).
6. M. Mecklenburg, W. A. Hubbard, E. R. White, R. Dhall, S. B. Cronin, S. Aloni,
and B. C. Regan, "Nanoscale temperature mapping in operating
microelectronic devices," Science 347, 629 (2015).
7. Z. Wang, J. Dai, and S. Yang, "The development of a multispectral pyrometer
for achievable true temperature field measurements of the explosion flame,"
Measurement science and technology 34, 065501 (2023).
8. Y. Luan, D. Mei, and S. Shi, "Light-field multi-spectral radiation
thermometry," Optics letters 46, 9-12 (2021).