基于全连接神经网络的分层旋流火焰燃烧振荡预报

In order to guide the active control system to suppress the combustion instabilities, relevant researches on combustion oscillation prediction based on different methods should be carried out to verify their availabilities. The time-averaged images of premixed internally-staged-swirling stratified flame of CH₄ are adopted here as the basis of our research. In order to simplify the information contented in flame images, two pre-processing methods on images are used here, which are resolution degradation and extracting properties of flame structure variables. The neural networks with fully connected layers are implemented to predict combustion oscillation with pre-processed data. It is found that both methods can obtain good prediction accuracies (better than 90%). Besides, the network using degraded resolution as inputs might still behave well with an accuracy over 90% at an extremely low resolution (3×3). A positive correlation of prediction accuracy and resolutions is found. In the terms of flame structure features, the transition of flame stability dynamics is captured under a limited range of parameter variations. The combustion oscillation prediction time is less than 2ms using the proposed data-driven methods, which provide support for real-time prediction of combustion oscillation.