Sistema de multilateração acústica para localização passiva de animais vocalizantes e outras fontes sonoras

Abstract

Brazil is the country with the greatest biodiversity on the planet. It is estimated that more than 10% of the organisms known to science inhabit Brazilian forests. In biology and environmental sciences, it is often necessary to locate animals that cannot be monitored using traditional methods, either because of their small size, difficulty of access, or to avoid disturbing them in their habitat. Bioengineering, a subfield of Biomedical Engineering, focuses on solving problems related to biological systems. This paper presents a method capable of locating sound sources such as vocalizing animals without handling them and without interfering with their habitat. Methods: First, a pilot study was conducted using two-dimensional trilateration based on triangular geometry, employing three microphones to determine the position of a sound source on a plane. For spatial localization, a novel approach for three-dimensional passive acoustic multilateration was developed, based on the time of arrival (TOA) of sound at four microphones arranged in a Cartesian xyz geometry with known positions. Connected to the four microphones are four synchronized, low-cost microcontroller systems coupled to the acoustic sensors. When a sound reaches the microphones, its arrival time at each sensor is recorded. An algorithm is proposed to process the TOA datasets and determine the spatial location of sound sources by comparing the volumes of two tetrahedra calculated through different procedures. Results: The practical results demonstrated a maximum average relative error of 3.03% and an average relative error of 1.97%. The total cost of the system was approximately six hundred dollars (R$ 3,500.00). Conclusion: This article introduces a passive method for locating sound-emitting objects with centimeter-level precision in environments spanning tens of meters. The method and system were tested in the field and successfully determined both the position of sounds produced in a controlled environment and the position of vocalizing songbirds. The method is based on simple mathematical principles, and its cost, limitations, and future developments are discussed in detail. In addition to locating sound sources and vocalizing animals, this method can be used to locate any sound-emitting source. Secondary findings from the research are also presented, including the study of the vocalizations of the frog species *Scinax tymbamirim* in its natural environment.