Research And Prediction Of AMS Temperature Dynamic Response Characteristics During Special Operation Of The International Space Station

Alpha Magnetic Spectrometer(AMS)is a large-scale scientific experiment carried out on the International Space Station.Its scientific purpose is to explore antimatter and dark matter in the universe.During the on-orbit operation of AMS,each of its components has strict operating temperature requirements,and the thermal control system is the basis for the stable and normal operation of each component.While the thermal control system is working for a long time,a large amount of AMS thermal data is recorded.Therefore,based on these historical thermal data,using machine learning as a tool to analyze and discuss the thermal environment of AMS is of great significance to the future stable operation of AMS and even to space thermal science.In the past ten years,the AMS thermal control system has recorded The huge thermal data below gives the possibility of this research method.Based on the thermal balance of the AMS in orbit,this article first analyzes the thermal environment of the AMS in orbit,and derives various influencing factors that affect the temperature of the AMS based on the orbital parameters of the International Space Station(ISS),and the International Space Station The influence of the change of its own flight attitude and component status on the external heat flow of the AMS.Next.using the actual thermal data recorded by the AMS,the dynamic response characteristics of the AMS temperature with the β angle change and the special operation of the ISS are analyzed from six directions.Finally,after selecting the appropriate measurement points and parameters,the AMS historical thermal data set was sorted out,and the AMS thermal data was processed by means of mathematical statistics and machine learning.First,the relationship between AMS temperature and influencing factors is analyzed by factor analysis,and the previous inference is verified from the perspective of mathematical statistics through factor loading.Then use the Python platform to establish an artificial neural network,take the influencing factors as the input,and the measured point temperature as the output.After three working conditions are put into the data set for training,the predicted value of the measured point temperature is obtained,and various predictions are achieved.The effect of AMS temperature under working conditions.The actual temperature data was used for comparison and verification,and the results were good.The research results show that the surface temperature in the six directions of AMS has obvious distribution rules on the β angle,which is in line with the conclusion that the β angle is the most important factor affecting the external surface temperature of AMS.Among them,the measurement points IN-5 and IN-6 on the top surface of the AMS may have a low temperature alarm,and the PDS surface measurement points of the AMS may have a high temperature alarm.At the same time,the special operation of the International Space Station will also affect the temperature of the AMS:the working conditions for locking SARJ are divided into locking when the β angle is positive and locking when the β angle is negative.When the β angle is around 0°Time-locking solar panels has little effect on the external heat flow of AMS.When changing the angle lock angle:when the β angle is positive,the absolute value of the adjustment angle increases,which will reduce the external heat flow on the WAKE side;when the absolute value of the adjustment angle decreases,the external heat flow on the WAKE side will increase.When the β angle is negative,the result is the opposite.After processing the AMS data set by factor analysis,the factor loading verifies that the WAKE side temperature is mainly determined by three influencing factors,of which the β angle is the dominant factor,and the angle and the locking of the solar panel are the secondary factors.Through the processing of the AMS data set by artificial neural network training,the predicted value of the temperature data is obtained,which is in good agreement with the actual value,and provides a new idea for the prediction and analysis of AMS temperature under multiple working conditions.