Research On Target Infrared Radiation Characteristics Measurement Based On Multi-Channel Temperature Control

During the flight of a fighter aircraft,the fuselage skin will have an uneven temperature rise due to the strong aerodynamic heating effect,which greatly increases the infrared radiation energy of the fuselage skin.With the development of infrared imaging detection technology,the detection and track mode is being changed from short wave of 3～5μm band to long band of 8～12μm.Which result more and more serious threat for the combat aircrafts.Therefore,the infrared stealth ability will become even important in the future,and it is of great application value and significance to carry out research on the infrared radiation characteristics of aircraft skin.In order to study the infrared radiation characteristics of the fuselage skin,uses a part of fuselage skin with a stealth coating as the research target because of the confidentiality of stealth aircraft.First of all,for the ground test of the skin,this article has been equipped with a ground temperature control test system and has designed a temperature controller based on STM32F407 processor to control the fuselage skin’s temperature.This controller measures temperature through Pt100 thermal resistance and measuring bridge and it’s equipped with a segmented PID control algorithm,which generates control signals in the form of PWM.Control signal is output from the controller’s I/O port to the solid state relay,in order to drive silicone heaters to work at different equivalent powers.This controller has achieved no over-temperature control from 100°C to 200°C,and the stabilization time is 60% shorter than traditional PID controllers.In order to achieve higher accuracy of the experiment,the experiments of blackbody calibration were carried out based on MR170 Fourier Transform Infrared Spectrometer produced by ABB Group Company and to complete the calibration of the test instrument.Secondly,infrared radiance models based on Support Vector Machine and Wavelet neural network have been established,in order to eliminate interference from external factors.Using standard blackbody’s data to train the network because of the unknowing of the emissivity of the fuselage skin.Using the network training results to calculate the emissivity of the measured blackbody,the error is within the expected range by comparing result with the theoretical blackbody.Which proved that the accuracy of the two algorithms designed in this article.At last,testing the infrared radiation characteristics of the fuselage skin and training data of the fuselage skin based on two established infrared radiance models.The training results show that both two methods can effectively eliminate interference from external factors and make the calculated emissivity of the fuselage skin more accurate.Compared the difference of the two algorithms,proved that the modeling method of Wavelet neural network is slightly better than the Support Vector Machine neural network.