Digital Smoothing Of The Langmuir Probe I-V Characteristic

Electrostatic probes are among the most important diagnostic devices in plasma, providing an experimentally simple method for measuring various plasma parameters.A Langmuir probe in its most basic form is simply a thin wire which is inserted into a plasma.The probe is biased at a variable voltage, positive and negative to the plasma and the current collected provides information on various plasma properties.The measured current as a function of the applied voltage is known as the probeⅠ-Ⅴcharacteristic.The use of a Langmuir probe relies on efficient analysis of thisⅠ-Ⅴcharacteristic.From the characteristic one can determine the electron density, the electron temperature, the plasma potential and the electron energy distribution function(EEDF).To determine the plasma parameters it is important to determine the first and second derivative of theⅠ-Ⅴcharacteristic accurately.Any noise present in the measurement is greatly amplified by the differentiation rendering the first and second derivative unless some method of noise suppression is applied.In the filtering design respect, we use arc tangent function to simulate theⅠ-Ⅴcurve.The range of voltage is-10V～15V, the lengths of date are 1032 and 126. The noise added is random Gauss noise, we use several methods to smooth theⅠ-Ⅴcurve:polynomial fitting, Savitzky-Golay filter, a Blackman window and swave filter. On the use of Savitzky-Golay filter, we receive the first and second derivative of theⅠ-Ⅴcharacteristic when it is fitting.The thesis make compare between two wavelet bases:db wavelet and sym wavelet. It also compare three threshold selecting strategy:penalty strategy,Birge-Massart strategy and stabilized strategy.In the filtering result estimate respect, we compare the mean square error and correlation coefficient between the fitted and origined data.We also plot curve to compare them..In program designing respect, we design filter programs of all methods, compute the comparison parameter and plot the curve using Matlab.In experimental respect, we use the optimized method to smooth the datum, receiving optimistical result.