Machine Learning for Turbulent Combustion Modelling

燃烧建模是一个非常复杂的领域，结合了计算流体动力学，传热传质和化学动力学。化学动力学的求解占据了模拟中的大部分时间。为了缓解这种情况，可以训练人工神经网络来模拟化学物种的进化。人工神经网络能够模拟高度非线性的函数，例如化学物种随时间的演变。因此，本计画的目的是应用人工神经网路来模拟化学物种的时间演化，并将其应用于工业上。

Combustion modelling is an incredibly complex field, combining computational fluid dynamics, heat and mass transfer and chemical kinetics. The solution of the chemical kinetics takes up the majority of the time in a simulation. To try to alleviate this, artificial neural networks can be trained to emulate the evolution of the chemical species in time. Artificial neural networks are capable of modelling highly non-linear functions, such as the evolution of chemical species in time. Therefore, this project aims to apply artificial neural networks to model the temporal evolution of chemical species, and then apply it to an industrial situation.