Acoustic Source Localization with Aerial Drones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aerial drones face challenges in acoustic source localization due to strong, wideband, non-stationary ego-noise from their propeller units, which interferes with the detection of sound sources in physical spaces.
Innovation Solution
The system employs a sparse sensor array design combined with mobility-induced beam forming and intra-band and inter-measurement fusion to enhance the signal-to-noise ratio (SNR) by splitting wideband acoustic signals into narrow sub-bands, measuring power in each cell, and performing intra-band and inter measurement fusion to identify the geo-location of acoustic sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wideband acoustic signal processing is used for acoustic source localization, then the detection capability is improved, but the ego-noise from propeller units interferes with sound source detection
Solution Approach 1:
The wideband acoustic signal is divided into multiple narrow sub-bands using frequency binning. This segmentation allows the system to process each sub-band separately, improving the signal-to-noise ratio for individual frequency components while mitigating the impact of wideband propeller noise through selective filtering and fusion of sub-band measurements.
2Measurement precision
If conventional acoustic localization methods are used, then the system complexity is low, but the location accuracy is insufficient in the presence of propeller noise
Solution Approach 1:
The system incorporates temporal dimension by performing measurements at multiple positions along the flight path and fusing these measurements over time. This multi-dimensional approach (combining spatial positions with temporal sequencing) enables accurate source localization despite the complexity introduced by noise mitigation requirements.
Solution Approach 2:
The patent introduces an intermediary processing stage that fuses measurements from multiple sub-bands and multiple positions. This intermediary fusion process acts as a mediator between the raw noisy measurements and the final location estimate, improving accuracy while managing system complexity through structured signal processing.
3Measurement precision
If measurements are taken at multiple positions for fusion, then the location accuracy improves, but the measurement and processing time increases
Solution Approach 1:
The system performs preliminary frequency binning and sub-band separation before the fusion process. By preparing the signal structure in advance through frequency decomposition, the subsequent fusion of multi-position measurements becomes more efficient, reducing overall processing time while maintaining improved location accuracy.
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 improves the SNR by 15 to 18 dB, achieving location accuracy of approximately 40 cm within a 6 m×3 m scan region, overcoming the limitations of conventional methods.
Implementation Method 1
measuring, by the one or more hardware processors, power in each of the cells by forming a beam at each of the cells
Data Source
AI summary
Detecting sound sources in a physical space-of-interest is challenging due to strong ego-noise from micro aerial vehicles (MAVs)' propeller units, which is both wideband and non-stationary. The present subject matter discloses a system and method for acoustic source localization with aerial drones. In an embodiment, a wideband acoustic signal is received from an aerial drone. Further, the wideband acoustic signal is splitted into multiple narrow sub-bands having cells. Moreover, from a measurement position corresponding to each of the multiple narrow sub-bands, power in each of the cells is measured by forming a beam to each of the cells. In addition, intra-band and inter measurement fusion of the measured power at each of the cells is performed. Also, geo-location of an acoustic source corresponding to the wideband acoustic signal is identified upon performing intra-band and inter measurement fusion of the measured power.