Noise reduction of the automobile multi-mode muffler using differential gap control and neural network control

Abstract

In this study, a controllable multi-mode exhaust system was investigated by analysing the acoustic structure of a low-noise low-back-pressure air exhaust system and forming its controller. After selecting the exhaust system structure that considered the frequency characteristics of the vehicle’s exhaust noise with a large acoustic transmission loss, the basic muffler of the controllable multi-mode exhaust system was designed. The basic structure of the controllable multi-mode muffler with a curved side inlet and a curved side outlet was determined. The theoretical and experimental values of the acoustic transmission loss for straight pipes and for curved pipes were compared with no significant difference. To control the curved multi-mode exhaust system, a differential gap controller and a neural network controller were designed and simulated for control tests. The study findings reveal that the use of the proposed model for noise control achieves a noise reduction of 40–50% over systems with no noise control, proving that this exhaust system is effective in noise control. The feasibility of the noise-controllable multi-mode exhaust system was investigated using both a differential gap controller and a neural network controller.