Design Of High-speed Temperature Measurement System Based On Multi-spectral Radiation

Transient temperature precision measurement has a wide range of needs in engineering.A variety of backgrounds,including rocket engine combustion,aviation engine combustion,ammunition explosion,laser processing,etc.,require the measurement of transient temperature changes in order to analyze the material characteristics of the explosive materials and the processing through temperature changes.Conventional contact temperature measurement methods are difficult to meet the needs of transient temperature measurement due to the thermal inertia of the temperature sensor.Especially in high temperature environments,the sensor material is hard to meet the requirements of the measurement environment,and under extreme conditions such as explosions.It is even more difficult to obtain information about the entire temperature change process in the transient explosion process through the contact method.Therefore,non-contact measurement based on spectrum is an effective method to solve transient temperature measurement.The multi-spectral temperature measurement is a method that can reflect temperature through light intensity and wavelength.It is a temperature measurement method developed based on the idea of colorimetric temperature measurement.Compared with colorimetric temperature measurement,multispectral temperature measurement has a wider range of applications and is not limited to the gray body radiation objects.However,the use of multi-spectral temperature measurement technology needs to solve the problems of spectral separation,the nonlinear sensitivity characteristics and the consistency calibration of the spectrum of the spectral detector,and the information processing.In this paper,related researches are carried out for the needs of transient temperature measurement and the related theories,technologies and system implementation of multispectral temperature measurement.This paper first studied the multi-spectral separation technology based on theoretical analysis and computer simulation,proposed an optimized design method for a uniform spectrum separation system about combination of prism and grating;Then,a spectral separation system with a more uniform spectral distribution and a spectral resolution of better than 2 nm in the 400-800 nm spectral range was designed;Based on the engineering problem of transient temperature measurement,we designed a temperature measurement circuit system with a collection speed of 20 μs based on the high-speed photodiode array and STM32 main control chip;Aiming at the non-linear sensitivity characteristics and consistency calibration of the spectrum detector,the system calibration and data preprocessing methods are studied.The corresponding signal processing,calibration,and calibration software are designed to complete the calibration of the system,besides,the temperature measurement method based on neural network is studied.Through neural network training,the model between multi-spectral radiation intensity and temperature is established.Static blackbody furnace calibration experiments show that the system achieves a better result within the temperature range of 1550～2000 degrees Celsius.Using the static combustion experiment of tungsten sheet,the reliability of the temperature measurement accuracy of the equipment was verified.In addition,the gunpowder combustion process experiment verified that the system can obtain all the temperature changes of the gunpowder combustion process at high speed.