Active Noise Control Filter Classification for Hearing Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Active noise control (ANC) systems in hearing devices face challenges due to varying wearing situations and individual ear anatomy, leading to suboptimal performance, especially in non-adaptive systems, where sampling rate mismatches and inaccurate feedback microphone placement cause inefficiencies.
Innovation Solution
A method that identifies the wearing situation and noise incidence direction, using a classifying algorithm to select the best-suited filter set from a pre-determined set stored in memory, allowing for independent sampling rates and adaptive filter application, with optional gain adjustment and disruption handling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If adaptive filters with LMS algorithm are used to optimize ANC performance for different wearing situations, then ANC performance is improved, but device complexity increases due to sampling rate mismatch requirements
Solution Approach 1:
The system segments the filter selection process into discrete filter sets corresponding to different wearing situations (e.g., earbud vs. over-ear, different fit conditions). Each filter set is pre-configured for specific conditions, eliminating the need for continuous adaptive adjustment and resolving the sampling rate mismatch problem while maintaining ANC performance.
2Productivity
If the feedback microphone is used to minimize error signal, then filter optimization is achieved, but measurement accuracy deteriorates due to mismatch between microphone location and eardrum position
Solution Approach 1:
The system creates a virtual model of the eardrum position by processing the feedback microphone signal through transfer function estimation. This virtual copy allows the system to optimize filters based on eardrum-level noise conditions rather than microphone-level conditions, improving measurement accuracy while maintaining optimization efficiency.
3Adaptability or versatility
If multiple filter sets are stored for different wearing situations, then adaptability is improved, but memory requirements and device complexity increase
Solution Approach 1:
The system performs preliminary classification of the current wearing situation using machine learning models that analyze sensor data (accelerometer, gyroscope, microphone). Based on this classification, the appropriate pre-configured filter set is selected in advance, enabling rapid adaptation without real-time complex computations and reducing memory access overhead.
Solution Approach 2:
The system organizes filter sets by key parameters such as wearing type (earbud/over-ear), fit quality, and noise environment. This parameter-based organization allows efficient retrieval and switching between filter sets based on detected conditions, managing complexity through systematic categorization rather than arbitrary filter storage.
4Measurement precision
If complex two-stage filter selection process is used to estimate transfer function, then measurement accuracy is improved, but processing time and energy consumption increase
Solution Approach 1:
The system performs transfer function estimation and filter optimization in advance during manufacturing or initial setup, storing the results in pre-configured filter sets. During actual use, the system only needs to select the appropriate pre-computed filter set based on wearing situation classification, dramatically reducing processing time and energy consumption while maintaining measurement accuracy.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
The invention relates to a method of classifying and using acoustic filters for active noise control in hearing systems, the filters being determined beforehand and being stored in a memory in the earphones. It is therefore possible to select and use quickly and efficiently a certain filter when wearing the earphones in order to improve filter performance and filter stability.