Diagonalization Filter Matrix for Active Noise Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional active noise cancellation systems face challenges in achieving optimal convergence due to complex acoustic environments and limited adaptation time, leading to unsatisfactory performance, especially in vehicle cabins where multiple sound zones require precise noise cancellation.
Innovation Solution
The implementation of a diagonalization filter matrix that separates noise cancellation into offline acoustic tuning and real-time adaptation, simplifying the system by decoupling the ANC effort into offline diagonalization and real-time adaptation, reducing computational complexity and improving convergence rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional active noise cancellation systems use full adaptive filter systems for real-time noise cancellation, then noise cancellation performance can be achieved, but convergence is slow due to complex acoustic environments and limited adaptation time
Solution Approach 1:
The patent segments the noise cancellation system into two independent parts: an offline diagonalization filter matrix that handles acoustic environment modeling, and an online adaptive filter that handles real-time noise cancellation. This segmentation allows each component to be optimized independently, with the offline part performing heavy computational tasks during system setup and the online part performing lightweight real-time processing, thereby significantly improving convergence rate while maintaining noise cancellation performance.
Solution Approach 2:
The patent applies preliminary action by pre-computing the diagonalization filter matrix during an offline calibration phase before actual noise cancellation operations begin. This preliminary computation captures the acoustic characteristics of the environment and stores them in a compressed form, allowing the real-time system to skip complex acoustic modeling and directly use the pre-computed information, thus dramatically accelerating convergence during operation.
2Adaptability or versatility
If the system processes multiple sound zones simultaneously with full adaptive filtering, then comprehensive noise cancellation coverage is achieved, but computational complexity increases significantly
Solution Approach 1:
The patent segments the computational workload by separating the heavy offline diagonalization computation from the lightweight online adaptive filtering. Each sound zone can be independently processed using the pre-computed diagonalization matrix, allowing multiple zones to be handled simultaneously without proportionally increasing real-time computational complexity. The offline part handles the complex matrix operations once during calibration, while the online part performs simple vector-matrix multiplications for each zone during operation.
Solution Approach 2:
The patent performs preliminary computation of the diagonalization filter matrix for each sound zone during offline calibration, capturing and compressing the acoustic characteristics of each zone. This pre-computed information is then reused for all real-time processing across multiple zones, eliminating the need to re-compute complex acoustic models for each zone during operation and thus reducing overall computational complexity while maintaining comprehensive coverage.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach results in faster convergence rates and improved noise cancellation performance by decoupling the ANC system, allowing for independent sound zone management and minimizing interference between zones, thus enhancing robustness and maintaining performance across multiple zones.
Implementation Method 1
Active noise cancellation (ANC) may be used to generate sound waves or anti-noise that destructively interferes with undesired sound waves
Implementation Method 2
Sensors placed near the listeners' ear positions provide the error signals for the feedback structure
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Estimated output signals of the reference signals are generated using an estimated filter path transfer function that provides an estimated effect on sound waves traversing a physical path, the estimated filter path transfer function performing processing according to a diagonalization matrix and reference signals. Anti-noise signals are generated from the reference signals using an adaptive filter driven by learning unit signals received from a learning algorithm unit, the learning unit signals based in part on error output signals generated from the estimated output signals, the anti-noise signals including signals per sound zone and per reference signal, each sound zone including a microphone and one or more loudspeakers. A sum across references is performed on the anti-noise signals to generate a set of output signals per sound zone. The set of output signals are processed by the diagonalization matrix to generate a set of output signals per loudspeaker.