Study On The Thermal Control Technique For Earth-moon Imaging Spectrometer Load System For High-altitude Balloon

CLARREO and TRUTHS projects have been launched in the United States and Europe respectively,these two projects were planned to launch the laboratory radiation benchmark into space and establish a ‘traceable SI space radiation calibration standard laboratory' to improve the global on orbit radiation calibration level.The Earth-Moon imaging spectrometer is the key instrument to establish the ‘traceable SI space radiation calibration standard laboratory' in China,which is loaded in the cube pod and carried on the High-altitude balloon platform and working in the stratosphere near the altitude of 20 km.In order to ensure the detection accuracy of the spectrometer,the temperature level of the spectrometer and the temperature gradient of the load system should be kept,so the thermal control technology of the load system should be studied and thermal design should be carried out.In this paper,the thermal environment of the load system is firstly analyzed,and based on this,the heat balance equation of load system is established.After that,according to the common working conditions of winter solstice and summer solstice,the values of three kinds of external heat flow,namely solar flux from the Sun,Sunlight reflected off the Earth(albedo),and infrared(IR)energy emitted by the Earth are calculated,and the influence of aerodynamic heat is analyzed.Afterwards,the flight state of the load system is divided into two phase: rising phase and level flight.The surface convective heat transfer coefficient of the load system is calculated for both phases respectively,and the instantaneous subsection fitting method is used in the coefficient calculation of rising phase.The above analysis and calculation results are used to guide the thermal design and thermal simulation of the load system.Then,according to the low temperature thermal design target of the spectrometer frame temperature,the sensitivity analysis of the thermal design parameters is carried out to determine the main thermal design parameters that affect the temperature of the spectrometer frame.Firstly,according to the thermal balance equation and thermal control principle of the load system,12 thermal design parameters are selected for sensitivity analysis.Secondly,Morris screening method and Spearman rank correlation coefficient formula based on finite element simulation are used to analyze the local and global sensitivity of parameters respectively.In order to analyze the influence of multiparameter coupling on the frame temperature of spectrometer,the Sobol' sensitivity analysis method based on BP neural network agent model is used to get the first and total sensitivity indices of the parameters,and the BP-Garson method is used to verify the results.The results indicate that the most sensitive parameters in sensitivity ranking are the external surface convective heat transfer coefficient and the surface radiation characteristics of pod,and the thermal conductivity between the spectrometer frame and other components.In the thermal design,these parameters need to be focused on extremely.Based on the results of thermal environment analysis and sensitivity analysis of thermal design parameters of load system,the detailed thermal design and thermal analysis of load system are carried out.According to the thermal design index of the load system,the passive thermal control and active thermal control measures of the load system are designed respectively.In particular,the selection of heat conduction filler for the CCD detector is tested and verified.Then,the winter solstice and summer solstice conditions of the load system are defined,and the temperature distribution results of the load system under the two conditions are obtained by the finite element simulation of the load system established previously.The temperature results indicate that the temperature of the spectrometer frame can be reduced to-5±2? within one hour and maintained for more than three hours under the two conditions,and the temperature of the optical window is always higher than-8?,which meets the requirements of proposed thermal design index.Finally,the optimization design of the sensitivity analysis method of thermal design parameters based on agent model is carried out.In order to improve the speed and accuracy of generating parameter data set for sensitivity analysis,the method based on I-DEAS embedded macro development technology and MATLAB text replacement is adopted to realize the automatic input and modification of thermal design parameters in TMG.In order to improve the training speed of agent model and the calculation speed of parameter sensitivity,RBF neural network and Kriging interpolation proxy model are introduced.The calculation speed of two agent models is about 1.1 times and 60 times of BP neural network.At the same time,the convergence of parameter sensitivity based on BP neural network is also verified.The speed-up optimization of sensitivity analysis method is of great significance to the sensitivity analysis and model modification for a large number of parameter models.