Neural fixed-parameter active noise controller for variable frequency tonal noise

Sound emitted by audible warning devices for emergency vehicles usually exceeds maximum admissible rates of noise in the aspect of occupational health and safety of the crew inside. Therefore, the idea of active noise control systems for emergency vehicles is an important issue. Most of active noise control systems for emergency vehicles developed till now are based on the concept of adaptive control. A major disadvantage of adaptive noise control methods is the lack of robustness for external disturbances. In this paper the concept of robust, fixed-parameter controller for variable frequency tonal noises is presented. The concept is based on a neural network performing a role of intelligent data base for notch filter coefficients.