Double-swirl combustion instability control method and system

Abstract

The invention discloses a double-swirl combustion instability control method and system. The method comprises the following steps: firstly, generating a first-generation excitation function signal by a machine learning algorithm; and then building a test system composed of a double-swirl combustor, a combustion chamber, a gas supply system, an electromagnetic valve, a pressure sensor, a data acquisition card, a controller and a computer. The controller controls the electromagnetic valve to adjust the inner rotational flow fuel flow according to the excitation function signal. The pressure sensor collects combustion chamber wall surface pressure and transmits the pressure to the computer. Finally, a cost function is defined according to pressure oscillation for updating, so that machine learning enters the second generation. Circulating like this for several generations, when there is no longer a smaller cost function in machine learning, it means that the algorithm has converged. At the moment, the excitation function signal corresponding to the minimum cost function is the optimal excitation function signal. According to the method of the invention, the advantage of global optimization of the machine learning algorithm is brought into full play, combustion instability is restrained to the maximum extent, and the service life of an aero-engine and a gas turbine is prolonged.

Description

Technical field [0001] The present invention relates to a dual cycosity combustion unstable control technology, and in particular, to a two-cyclone combustion unstable control method and system. Background technique [0002] In recent years, engine pollutants, essentially of nitrogen oxides (NOX) are increasingly strict. Poverty Premix Premix Combustion Technology (LPP) is currently most likely to achieve extremely low NOx emissions, and there is a multi-use center-grade dual-fluid combustion tissue mode. Because the lean premix flame is extremely disturbed, the combustion is unstable, and the application of the LPP combustion chamber has brought a serious challenge. Therefore, inhibition or prevent the occurrence of unstable combustion is critical in LPP combustion chamber research. [0003] Currently, there are two combustion unstable control methods: passive control and active control. [0004] Passive control mainly changes the acoustic and flow field structure of the combust...

AI Technical Summary

Problems solved by technology

However, the current method of adjusting the fuel flow rate by applying an oscillating signal to control combustion instability is only artificially setting the modulation frequency and amplitude, and it is difficult to achieve the global optimum.

Images