A hybrid strategy of OTPA and machine learning for efficient root-cause analysis of NVH in multi-source systems

Aug 5, 2026

Vehicle comfort goals and competitive pressure drive systematic reduction of Noise, Vibration, and Harshness (NVH), making reliable source/path ranking essential. Transfer Path Analysis (TPA) supports identification and elimination of dominant contributors, and Operational Transfer Path Analysis (OTPA) avoids measuring interface forces and transfer matrices. However, OTPA may suffer from data loss due to pseudo-inversion of the indicator matrix. A hybrid random-forest regression estimates the transmissibility matrix, replacing truncated SVD while preserving key information. Verification shows a 2.28% SPL improvement at the driver’s ear by updating the critical path.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Introduction

Transfer Path Analysis background

Classical TPA challenges

Component-based TPA (CB-TPA)

Proposed hybrid strategy (OTPA + machine learning)

Transmissibility estimation via random forest

Efficiency validation and performance improvement

Applied Acoustics 242 (2026) 111060 Contents lists available at ScienceDirect Applied Acoustics journal [homepage:](homepage: www.elsevier.com/locate/apacoust)[ www.elsevier.com/locate/apacoust](homepage: www.elsevier.com/locate/apacoust) | A hybrid strategy of OTPA and machine learning for efficient root-cause analysis of NVH in multi-source systems
Sharif Khakshournia a, Shaygan Shahed Haghighia, Marzie Majidia, Farhad Najafnia a, Hamed Haddad Khodaparastb,*
a Noise and Vibration Department, Automotive Industries Research & Innovation Center of SAIPA (AIRIC), Tehran, Iran
b Faculty of Science and Engineering, Swansea University, Bay Campus, Fabian Way, Swansea, Crymlyn Burrows SA1 8EN, United Kingdom | | | | --- | --- | --- | | A R T I C L E I N F O | | A B S T R A C T | | Keywords:
Operational transfer path analysis Random forest
NVH diagnosis Sensitivity analysis Machine learning Contribution analysis | | The growing awareness of health benefits, along with the competitive emphasis on vehicle comfort, has led automakers to place greater attention on reducing Noise, Vibration, and Harshness (NVH). One of the most beneficial techniques for NVH engineers to identify, rank, and eliminate dominant noise and vibration sources and paths is Transfer Path Analysis (TPA). Unlike traditional TPA, Operational Transfer Path Analysis (OTPA) requires neither the preliminary acquisition of the transfer matrix between excitation and response points nor the measurement of forces transferred through the active and passive side connection points. Although the OTPA method offers significant advantages over classical TPA methods, it still faces challenges such as data loss caused by the pseudo-inversion of the indicator matrix. In this paper, we estimate the transmissibility matrix using a machine learning-based regression algorithm (random forest). We demonstrated that machine learning is an effective alternative to the truncated Singular Value Decomposition (SVD) method for estimating the transmissibility matrix, as it is a swift solution that preserves essential information in the indicator matrix. The efficiency of the method has been verified by a 2.28 % improvement in the Sound Pressure Level (SPL) of the driver’s ear noise of a sedan-type vehicle through the modification of the most critical path found by this approach. | 1. Introduction As automotive technologies evolve with electric powertrains and new structural design concepts, accurate identification and ranking of noise sources become increasingly crucial to address emergent or previously insignificant sources. Pinpointing and ranking the primary noise and vibration sources and paths in vehicles are pivotal for effective mitigation strategies. The development of advanced analytical techniques enables manufacturers to precisely identify the specific components or mechanisms that contribute most significantly to interior noise and vibration across relevant frequency bands. This targeted approach allows focused efforts on addressing root causes rather than attempting broad-based symptomatic solutions. These analytical methods, such as Transfer Path Analysis (TPA) and contribution analysis, empower manufacturers to stay ahead of evolving customer expectations for refinement and comfort while ensuring compliance with stringent noise regulations [1–3]. TPA-based methods analyze propagation of noise and vibration, from the excitation source in the active subsystem to the target point in passive subsystems, through paths in a complex system. Classical TPA methods depict the passive subsystem response and the influences of source excitation on the target point via the interface forces between the active and passive subsystems [4]. Hence, the method requires the interface forces acting during operational conditions and then Frequency Response Functions (FRFs) of the passive subsystem. Despite its apparent simplicity, the technique faces some several significan

FAQs

What is the main goal of the proposed OTPA and machine learning strategy?

To efficiently estimate the transmissibility matrix for NVH root-cause analysis in multi-source systems and improve critical-path identification for noise mitigation.

How does OTPA differ from classical TPA?

OTPA does not require preliminary transfer-matrix acquisition between excitation and response points, nor measurements of forces transferred through active/passive connection points under operational conditions.

Why does the paper use random forest instead of truncated SVD?

Random forest regression estimates the transmissibility matrix quickly while preserving essential information in the indicator matrix, addressing limitations linked to pseudo-inversion in OTPA.

What validation result demonstrates the method’s effectiveness?

The approach achieved a 2.28% improvement in Sound Pressure Level (SPL) at the driver’s ear noise of a sedan-type vehicle after modifying the most critical path found by the method.

Aug 5, 2026

Upload by Evangeline