Pasar al contenido principal

ALGORITMOS DE PROCESADO DEL SONIDO CON HERRAMIENTAS DE APRENDIZAJE AUTOMATICO BASADAS EN LA FISICA PARA SINTESIS DE CAMPO SONORO

PID2024-161353NB-C22

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para la Investigación y el Desarrollo Experimental
Subprograma Subprograma Estatal de Generación de Conocimiento Científico-Técnico y Desarrollo Experimental
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2024
Plan Estatal 2024-2027 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 2
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Low-Complexity Machine Learning Models for Active Noise Control in Nonlinear Systems

  • Ferrer Contreras, Miguel|||0000-0002-8743-1887
  • de Diego Antón, María|||0000-0001-9948-3396
  • Gonzalez, Alberto|||0000-0002-6984-3212
[EN] Active Noise Control (ANC) systems are typically based on adaptive filters. However, electroacoustic transducers and their associated electronic components often exhibit nonlinear behaviors that linear controllers cannot accurately model, resulting in suboptimal performance. While neural networks and other advanced models have been proposed to address these limitations, their high computational demands and inherent latency frequently restrict real-time deployment. This work investigates the use of lightweight, computationally efficient machine learning (ML) models that operate on a sample-by-sample basis using simple digital operators. The proposed models are applied to ANC under nonlinear conditions, including distortions in the primary path, the secondary path, and the reference signal. The approach enhances noise attenuation while preserving low computational complexity, thereby enabling real-time implementation on embedded systems. Simulation results confirm the effectiveness of the method across a variety of nonlinear scenarios, demonstrating superior noise reduction and control accuracy compared to conventional linear ANC schemes, and achieving this at a significantly lower cost than alternative nonlinear approaches., This work was supported in part by State Research Agency (AEI), MICIU/AEI/10.13039/501100011033 under Grant PID2021-125736OB-I00 and Grant PID2024-161353NB-C22, in part by Generalitat Valenciana, Conselleria de Innovation, Universities, Science and Digital Society, through PROMETEO Programme under Project CIPROM/2022/20, and in part by the EU through IN-NOVA, active reduction of noise transmitted into and from enclosures through encapsulated structures under Grant 101073037.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

An Error Innovation-based Adaptive Kalman Filter for multi-channel Active Noise Control

  • Islam, Tahmida
  • Ferrer Contreras, Miguel|||0000-0002-8743-1887
  • de Diego Antón, María|||0000-0001-9948-3396
  • Gonzalez, Alberto|||0000-0002-6984-3212
[EN] This paper presents an Error Innovation-based Adaptive Estimation Kalman Filter (eIAEKF) tailored for multi-channel Active Noise Control (ANC) systems using MIMO filtered-x structures. Conventional Kalman-based ANC methods are computationally demanding and typically rely on fixed or manually tuned noise covariance matrices, limiting their robustness under changing acoustic conditions. The proposed eIAEKF introduces an error-driven mechanism that adaptively updates the measurement-noise covariance using the measured error signal, directly aligning the estimation process with the ANC objective. Building on the principles of innovation-based adaptive estimation, the algorithm reduces computational complexity by avoiding the need for an explicit transition matrix and by simplifying the computation of the Kalman gain, while also improving robustness to poor initialization. Simulation results under both stationary and dynamic scenarios, including changes in the primary noise spectrum, source position, and microphone location, show that eIAEKF achieves faster convergence and higher attenuation than an adaptive Kalman-like baseline, while maintaining numerical stability. Real-time experiments on a hardware platform further demonstrate the practical viability of the proposed approach., This work was supported by in part by MICIN under Grants PID2021-125736OB-I00 and PID2024-161353NB-C22 (MCIN/AEI/10.13039/501100011033/FEDER UE, "ERDF A way of making Europe") , in part by GVA under Grant CIPROM/2022/20, and in part by EU through Marie Sklodowska-Curie Action, HORIZON-MSCA-2021-DN, IN-NOVA (Active reduction of noise transmitted into and from enclosures through encapsulated structures) under Grant 101073037.