Real-time physico-neural solutions for MOCVD
- Univ. of Colorado, Boulder, CO (United States)
This paper presents an integrated physical neural network approach for the modeling and optimization of a vertical MOCVD reactor. A first-principles physical model for the reactor was solved numerically using the Fluid Dynamics Analysis Package (FIDAP). This transient model included property variation and thermodiffusion effects. Artificial Neural Network (ANN) models were then trained to predict the growth rate profiles within the reactor. The data used to train the network was obtained from FIDAP simulations for combinations of process parameters determined by statistical Design of Experiments (DOE) methodology. It is shown that the trained ANN predicts the behavior of the reactor accurately. Optimum process conditions to obtain a uniform thickness of the deposited film were determined and tested using the ANN model. The results demonstrate the power and robustness of ANNs for obtaining fast on-line responses to changing input conditions. This capability of ANNs is particularly important for implementing run-to-run and on-line control of the MOCVD process.
- OSTI ID:
- 462602
- Report Number(s):
- CONF-950828-; ISBN 0-7918-1705-9; TRN: IM9719%%72
- Resource Relation:
- Conference: 1995 National heat transfer conference, Portland, OR (United States), 5-9 Aug 1995; Other Information: PBD: 1995; Related Information: Is Part Of 1995 national heat transfer conference: Proceedings. Volume 4: Transport phenomena in manufacturing and materials processing; Transport phenomena in materials joining processes; Transport phenomena in net shape manufacturing; HTD-Volume 306; Mahajan, R.L. [ed.] [Univ. of Colorado, Boulder, CO (United States)]; PB: 261 p.
- Country of Publication:
- United States
- Language:
- English
Similar Records
Removal of Congo Red from Aqueous Solution by Hydroxyapatite Nanoparticles Loaded on Zein as an Efficient and Green Adsorbent: Response Surface Methodology and Artificial Neural Network-Genetic Algorithm
Prediction of Weld Penetration in FCAW of HSLA steel using Artificial Neural Networks