Artificial Neural Network-Based Active Noise Cancellation of Ambulance

Abstract

This study investigates the use of artificial neural networks (ANNs) in Active Noise Cancellation (ANC) systems, demonstrating their effectiveness in reducing acoustic noise. By training the system with real-world audio data, such as ambulance sounds, the ANN-based approach successfully minimized noise while maintaining performance balance. The neural network design employed the tanh activation function across hidden and output layers, structured with three hidden layers (2, 3, and 5 nodes). Optimal performance was achieved through careful parameter tuning, including a learning rate of 0.001 and a momentum value of 0.9. The research underscores the capability of multi-layer perceptron (MLP) networks in accurate signal prediction and highlights the significance of parameter optimization. This adaptable MLP-based method holds promise for a wide range of signal prediction and noise cancellation applications.