A Long Short-Term Memory Network for Plasma Diagnosis from Langmuir Probe Data

Abstract

Electrostatic probe diagnosis is the main method of plasma diagnosis. However, the traditional diagnosis theory is affected by many factors, and it is difficult to obtain accurate diagnosis results. In this study, a long short-term memory (LSTM) approach is used for plasma probe diagnosis to derive electron density (Ne) and temperature (Te) more accurately and quickly. The LSTM network uses the data collected by Langmuir probes as input to eliminate the influence of the discharge device on the diagnosis that can be applied to a variety of discharge environments and even space ionospheric diagnosis. In the high-vacuum gas discharge environment, the Langmuir probe is used to obtain current–voltage (I–V) characteristic curves under different Neand Te. A part of the data input network is selected for training, the other part of the data is used as the test set to test the network, and the parameters are adjusted to make the network obtain better prediction results. Two indexes, namely, mean squared error (MSE) and mean absolute percentage error (MAPE), are evaluated to calculate the prediction accuracy. The results show that using LSTM to diagnose plasma can reduce the impact of probe surface contamination on the traditional diagnosis methods and can accurately diagnose the underdense plasma. In addition, compared with Te, the Ne diagnosis result output by LSTM is more accurate.

A bidirectional long short-term memory network for electron density diagnostic with double probe

“…The chamber [26] is a cylindrical stainless steel vessel with a diameter of 0.8 m and a length of 1 m (see figure 6). The background vacuum can reach 10 −4 Pa. During the experiment, the gas discharge pressure was controlled at 0.1 Pa. After argon is filled into the chamber, the filament current is controlled by the computer.…”

“…In a previous experiment [26], Wang et al successfully applied machine learning methods to single probe diagnostics of plasma without knowing the parameters of the discharge device, and performed data collection using contaminated single probes. The predicted results have a certain compensation effect, but still cannot completely eliminate the impact of contamination.…”

Development of the Measurement of Lateral Electron Density (MOLE) Probe Applicable to Low-Pressure Plasma Diagnostics

“…Recently, to improve throughput and productivity, process monitoring technology has become significant because the price of a patterned wafer has steadily grown and loss by unstable process is no longer negligible [ 6 , 7 , 8 , 9 ]. Particularly, advanced process control (APC), which is a process-tuning technology based on real-time signals from monitoring devices, has attracted strong interest from industry [ 10 , 11 ].…”