Early Detection of Combustion Instabilities using Deep Convolutional Selective Autoencoders on Hi-speed Flame Video

Adedotun Akintayo,Kin Gwn Lore,Soumalya Sarkar,Soumik Sarkar

DOI: https://doi.org/10.1145/1235

2016-03-25

Abstract:This paper proposes an end-to-end convolutional selective autoencoder approach for early detection of combustion instabilities using rapidly arriving flame image frames. The instabilities arising in combustion processes cause significant deterioration and safety issues in various human-engineered systems such as land and air based gas turbine engines. These properties are described as self-sustaining, large amplitude pressure oscillations and show varying spatial scales periodic coherent vortex structure shedding. However, such instability is extremely difficult to detect before a combustion process becomes completely unstable due to its sudden (bifurcation-type) nature. In this context, an autoencoder is trained to selectively mask stable flame and allow unstable flame image frames. In that process, the model learns to identify and extract rich descriptive and explanatory flame shape features. With such a training scheme, the selective autoencoder is shown to be able to detect subtle instability features as a combustion process makes transition from stable to unstable region. As a consequence, the deep learning tool-chain can perform as an early detection framework for combustion instabilities that will have a transformative impact on the safety and performance of modern engines.

Computer Vision and Pattern Recognition,Machine Learning,Neural and Evolutionary Computing

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the early detection of combustion instability during the combustion process. Specifically, the instabilities in the combustion process can lead to significant performance degradation and safety hazards, especially in land - based and air - based gas turbine engines. These instabilities are usually manifested as self - sustaining large - amplitude pressure fluctuations and exhibit periodic coherent vortex structure shedding at different spatial scales. However, due to the sudden nature (i.e., bifurcation - type) of such instabilities, it is extremely difficult to detect them before the combustion process becomes completely unstable. To solve this problem, the author proposes a method based on a deep convolutional selective auto - encoder, using high - speed flame video frames to achieve early detection of combustion instability. This method selectively masks stable flame image frames by training the auto - encoder, allowing unstable flame image frames to pass through. During this process, the model learns to recognize and extract rich descriptive and explanatory flame shape features. Through such a training scheme, the selective auto - encoder can detect the subtle instability features when the combustion process transitions from the stable region to the unstable region. Therefore, this deep - learning tool chain can serve as an early - detection framework for combustion instability, having a transformative impact on the safety and performance of modern engines.

Early Detection of Combustion Instabilities using Deep Convolutional Selective Autoencoders on Hi-speed Flame Video

3D Convolutional Selective Autoencoder For Instability Detection in Combustion Systems

Combustion Instability Detection Based on Multi-Signal Information Fusion for an Industrial Gas Turbine Combustor

Vision Based Flame Detection Using Compressed Domain Motion Prediction and Multi-Feature Fusion.

A Robust Autoregressive Long-Term Spatiotemporal Forecasting Framework for Surrogate-Based Turbulent Combustion Modeling Via Deep Learning

Flame lift-off detector based on deep learning neural networks

A Comprehensive Investigation of LSTM-CNN Deep Learning Model for Fast Detection of Combustion Instability

Cross-Modal Virtual Sensing for Combustion Instability Monitoring

A deep learning-based dynamic deformable adaptive framework for locating the root region of the dynamic flames

Multi-Scale Video Flame Detection for Early Fire Warning Based on Deep Learning

Flame Detection in Surveillance

Flame Detection Using Deep Learning

Deep Convolutional Generative Adversarial Networks for Flame Detection in Video

Gas-Fired Flame Stability Based on Optical Flow Method and Deep Learning

Deep Convolutional Generative Adversarial Networks Based Flame Detection in Video

Data-driven Analysis of Turbulent Flame Images

Dynamical Mode Recognition of Coupled Flame Oscillators by Supervised and Unsupervised Learning Approaches

Autoadaptive Flame Detection and Classification Using Deep Learning of FastFlameNet CNN

Dimensionality Reduction and Dynamical Mode Recognition of Circular Arrays of Flame Oscillators Using Deep Neural Network

Real-time Combustion State Identification Via Image Processing: A Dynamic Data-Driven Approach

Physics-aware neural network flame closure for combustion instability modeling in a single-injector engine