The Cooperation between Combustion Theory and Data Science Paves the Way to Advanced Combustion Diagnosis
Abstract: This paper discusses how the development of data science influences that of combustion diagnosis technologies. In combustion, the energy is released from the chemical reaction between fuels and air. Flame is the glowing gaseous part of a fire. The image of the flame provides rich information on the combustion conditions such as fuel rich and fuel lean. Many features can be extracted from the flame images and the time-series analysis of the features can be directly employed to monitor the combustion conditions. The commonly used features include mean, standard deviation, third moment, Shannon entropy. The concept of attractorin nonlinear time-series analysis provides an effective framework to quantify the structure of the data embedded in high dimensional spaces. GARCH (Generalized Autoregressive Conditional Heteroskedasticity ) is widely used to quantify the processes in which time-varying variances appear. The research on combustion and data mining mutually benefit and provide a basis for advanced combustion diagnosis.
Keywords: combustion diagnosis; flame image; data mining; time-series; attractor; GARCH
How to Cite?: Tao Jia, Zizhen Jia, "The Cooperation between Combustion Theory and Data Science Paves the Way to Advanced Combustion Diagnosis", Volume 12 Issue 7, July 2023, International Journal of Science and Research (IJSR), Pages: 804-806, https://www.ijsr.net/getabstract.php?paperid=SR23710160954, DOI: https://dx.doi.org/10.21275/SR23710160954