Robust Engine Order Cancellation with Reconstructed-Component Decision Logic and Dynamic Order Selection

Pakhomov M. Robust Engine Order Cancellation with Reconstructed-Component Decision Logic and Dynamic Order Selection. In: 2026 AES 6th International Automotive Audio Conference. Audio Engineering Society; 2026. Paper 10. Available from: https://aes.org/publications/elibrary-page/?id=23404

@inproceedings{Pakhomov2026_23404,
  author = {Pakhomov, Mikhail},
  title = {{Robust Engine Order Cancellation with Reconstructed-Component Decision Logic and Dynamic Order Selection}},
  booktitle = {2026 AES 6th International Automotive Audio Conference},
  note = {Paper 10},
  year = {2026},
  month = jul,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23404}
}

TY  - CPAPER
TI  - Robust Engine Order Cancellation with Reconstructed-Component Decision Logic and Dynamic Order Selection
AU  - Pakhomov, Mikhail
T2  - 2026 AES 6th International Automotive Audio Conference
M1  - Paper 10
PY  - 2026
DA  - 2026/07/29
UR  - https://aes.org/publications/elibrary-page/?id=23404
PB  - Audio Engineering Society
LA  - en
AB  - Engine orders remain a noticeable source of tonal noise in vehicle cabins. In production EOC systems, attenuation is only part of the story; stable behavior across operating conditions is equally important. After successful cancellation, the target order can become masked by broadband background noise (low SNR). In addition, in-band intrusions may appear in the same frequency region. Under these conditions, observation-based criteria that rely on the current visibility of the tone in the microphone signal can cause false deactivation and mode chattering. We present an engine order cancellation method for an architecture with four front-row error microphones. Orders are modeled in real time using a quadrature sincos basis synchronized with rotational-speed information. The main contribution is a control-informed, order-level state logic that uses a reconstructed estimate of the current control contribution through secondary-path models (loudspeaker-to-microphone). This enables a practical distinction between absent/irrelevant and suppressed & masked cases and stabilizes operation in low-SNR-after-cancel and intrusion conditions. In addition, the proposed method includes a dynamic order-management layer (candidate _ top-K) with immediate activation when a dominant, non-cancelled order emerges and optional fade-in/out transitions during switching. The study considers three operating scenarios: masking after successful cancellation, in-band intrusions, and shifts of dominant orders.
ER  -