Ocena dokładności pomiaru temperatury radiometrem mikrofalowym

W artykule przedstawiono analizę czułości i dokładności pomiaru mocy promieniowania termicznego przy pomocy radiometru mikrofalowego, oraz zaprezentowano metodę cyfrowej detekcji synchronicznej, umożliwiającą poprawę parametrów użytkowych radiometru. Na zakończenie przedstawiono analizę statystyczną wyników pomiarów, potwierdzającą wnioski z analizy czułości i dokładności radiometru mikrofalowego z cyfrowym procesorem sygnałowym, pracującego w paśmie 1,5 GHz.

---

In the paper the analysis of sensitivity and measurement accuracy of thermal radiation power using microwave radiometer is presented. The metod of digital synchronous detection improving operational parameters of the radiometer is also described. A microwave radiometer measures power of thermal radiation emitted by bodies with temperatures larger then 0 K. Since power levels of this radiation are extremely low in the microwave range, the precise determination of fundamental radiometer parameters, like its sensitivity accuracy, are especially important. In the literature, minimum change of input noise power of the receiver system detected at the radiometer output is marked as D[Delta]Tmin and zero reading was analysed. It was equipped with digital signal processing unit operating simultaneously as a synchronous detector and a low-pass filter. In any radiometer of this type the influence of gain changes on the system noise temperature in eliminated. Temperature measurement in the radiometer consists in the measurement of a voltage mean value at the output of post-detection filter. This value, when assuming rectangular detector characteristics, is proportional to the mean-square value of a narrow-band stationary and ergodic random process which is, in fact, a signal at the radiometer input. So, the temperature measurement is in effect a measurement of a mean square value of radiometer input signal and is not, as any other, an error-free measurement. Talking this into account, standard uncertainty of temperature measurement using microwave radiometer can be expressed as an error mean square of mean square value of the measurement. Normalisation of the process at the post-detection filter output enables us to express fundamental parameters of any radiometer in terms of probability density function of anormal distribution. This is particularly useful when values measured at the low-pass filter output are instant values like for digital processing of signals in a low frequency channel of the radiometer. Standard uncertainty of temperature measurement equals radiometer sensitivity if we assume that its value depends only on random error of mean square value at the low-pass filter output. At the end of the paper statistical analysis of measurement results presented. It confirms conclusions resulting from analysis of sensitivity and resolution of the microwave radiometer with digital signal processing unit working at 1.5 GHz band.