A Method Based On Computational Fluid Dynamics For Simulation And Optimization Of Sounding Temperature Sensor

The research on the atmospheric temperature in troposphere and lower stratosphere is of important significance for improving the accuracy of weather forecasting and global climate change prediction. The temperature rise of sounding temperature sensor caused by solar radiation is an important factor which affects temperature measurement accuracy during the observation of the upper atmosphere.At present, wind tunnel experiment and empirical estimation are the methods used to study the solar radiation temperature rise. However, limited by the technical difficulty and the formula applicability, the methods are unable to offer precise numerical temperature measurements. To tackle the issue, a computational fluid dynamics method is implemented to study the senor .Bead thermistor sensor with different lead length, lead angle and lead numbers are built to study the law of radiation temperature rise. The simulation results show that the sensor with longer lead obtain lower temperature rise. The temperature rise of sounding temperature senor with 180° lead angle and four leads are less affected by the solar angle. A temperature rise equation is obtained by fitting the CFD results using a genetic algorithm method, where the solar elevation and azimuth angles act as two major variables.The models of the bead thermistor with the stand are built to do numerical simulation, because the stand will affect the temperature rise of the thermistor sensor.The simulation results show that the radiation temperature rise is positively correlated with thermal conductivity, altitude and solar radiation intensity and is negatively correlated with solar elevation angle. The bead thermistor with stand of 30 mm lead length acquire the smaller temperature rise; the sensor with smaller size and the structure without beam is beneficial to reduce the radiation rise of the sensor.A BP neural network algorithm method is implemented to obtain the temperature rise equation. The verification of the formula shows that the calculation and simulation value consists with each other in a high level. The maximum error is within 0.1 K,which reflect the reliability of the formula.In order to apply the results to practice more easily, the radiation error correction software of the sounding temperature sensor was designed on Qt. It is used for real-time correction to the radiation temperature rise errors of the sounding temperature sensor. Users can also search history data and simulation results of the temperature rise error.