Loading...
Publication
Deteção Acústica através de Fibras Óticas
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 8.16 MB | Adobe PDF |
Authors
Abstract(s)
A monitorizacao do ambiente maritimo enfrenta desafios crescentes, exigindo solucoes passivas (de modo a nao serem detetadas), escalaveis e com baixo custo de operacao. A tecnologia Distributed Acoustic Sensing (DAS), baseada em fibras oticas, surge como alternativa promissora aos hidrofones classicos para detecao e caracterizacao acustica de tráfego maritimo, aproveitando infraestruturas existentes de cabos. Esta dissertacao avalia a viabilidade do DAS em cenarios reais (REPMUS 2024: Sesimbra, Troia e Carcavelos) e em ensaios controlados de tanque, comparando diferentes cabos e condicoes de acoplamento.
Para confirmar a origem dos sinais captados no DAS, foram usados registos Automatic Identification System (AIS). Foram selecionadas janelas de dados em torno da hora e posicao da passagem de navios sobre o cabo (reportados por AIS). Esses dados foram pre-processados usando um filtro Butterworth passa-alto de 4a ordem, removendo a componente DC, e usando uma janela de Hamming antes de fazer a analise espectral. Sobre estes dados foram feitas analises de potência integrada por banda (Frequency Band Energy - FBE), transformadas de Fourier (FFT) quer de grandes segmentos com grande resolucao em frequencia, quer na forma de espetrogramas para observar variacoes no tempo, e foram calculadas relações sinal-ruido (SNR) sempre que possivel. No mar, observou-se concentração de energia relevante em baixas frequencias, orientando a filtragem para o intervalo util. Em tanque, a comparação entre cabos e acoplamentos evidenciou diferencas marcadas por frequencia, com melhor SNR em configuracoes na coluna de água e pior desempenho quando enterrados, bem como efeitos dependentes do contexto no uso de ressonadores. Nos testes preliminares feitos no ambito do projeto Fibermars, observou-se que os cabos enterrados conseguiam ter melhores relacoes sinal-ruido do que os cabos pousados no fundo ou na coluna de água. Os resultados aparentemente contraditorios obtidos em tanque indicam que esses maus resultados se deviam ao ruido de fluxo (flow-noise) da água sobre os cabos.
Os testes efetuados em tanque foram extremamente importantes para o projeto Fibermars, por serem os primeiros deste genero. No entanto, foram limitados pelas limitacoes da fonte emissora, que tinha uma banda util limitada e que não abrangia as frequencias muito baixas, e pela ausencia de uma calibracao absoluta, que levou a que todas as medicoes sejam relativas a condicao base. Ainda assim, os resultados suportam o potencial do DAS como complemento aos meios tradicionais de vigilancia e para reforco do conhecimento situacional maritimo, recomendando campanhas adicionais com varrimento espectral alargado, calibracao e controlo de variáveis ambientais.
Maritime environment monitoring faces increasing challenges, requiring passive solutions (so as not to be detected), scalable, and with low operating cost. Distributed Acoustic Sensing (DAS) technology, based on optical fibers, emerges as a promising alternative to classical hydrophones for the detection and characterization of maritime traffic, leveraging existing cable infrastructures. This dissertation evaluates the feasibility of DAS in real scenarios (REPMUS 2024: Sesimbra, Troia, and Carcavelos) and in controlled tank tests, comparing different cables and coupling conditions. To confirm the origin of the signals captured by DAS, Automatic Identification System (AIS) records were used. Data windows were selected around the time and position of vessel passages over the cable (as reported by AIS). These data were pre-processed using a 4th-order high-pass Butterworth filter to remove the DC component and applying a Hamming window prior to spectral analysis. On these data, band-integrated power analyses (Frequency Band Energy FBE) were performed, as well as Fourier transforms (FFT) on both long segments for high frequency resolution and in the form of spectrograms to observe temporal variations. Signal-to-noise ratios (SNR) were also calculated whenever possible. At sea, relevant energy concentration was observed at low frequencies, guiding filtering towards the useful band. In the tank, the comparison between cables and coupling configurations revealed marked frequency-dependent differences, with better SNR in water-column configurations and worse performance when buried, as well as context-dependent effects when using resonators. In preliminary tests conducted within the Fibermars project, it was observed that buried cables could achieve better signal-to-noise ratios than cables laid on the seabed or in the water column. The apparently contradictory results obtained in the tank suggest that these poor results were due to flow noise generated by water over the cables. The tank tests were extremely important for the Fibermars project, as they were the first of their kind. However, they were limited by the characteristics of the source, which had a limited useful bandwidth and did not cover very low frequencies, and by the absence of absolute calibration, which meant that all measurements are relative to the baseline condition. Nevertheless, the results support the potential of DAS as a complement to traditional surveillance means and for enhancing maritime situational awareness, recommending additional campaigns with broader spectral coverage, calibration, and control of environmental variables.
Maritime environment monitoring faces increasing challenges, requiring passive solutions (so as not to be detected), scalable, and with low operating cost. Distributed Acoustic Sensing (DAS) technology, based on optical fibers, emerges as a promising alternative to classical hydrophones for the detection and characterization of maritime traffic, leveraging existing cable infrastructures. This dissertation evaluates the feasibility of DAS in real scenarios (REPMUS 2024: Sesimbra, Troia, and Carcavelos) and in controlled tank tests, comparing different cables and coupling conditions. To confirm the origin of the signals captured by DAS, Automatic Identification System (AIS) records were used. Data windows were selected around the time and position of vessel passages over the cable (as reported by AIS). These data were pre-processed using a 4th-order high-pass Butterworth filter to remove the DC component and applying a Hamming window prior to spectral analysis. On these data, band-integrated power analyses (Frequency Band Energy FBE) were performed, as well as Fourier transforms (FFT) on both long segments for high frequency resolution and in the form of spectrograms to observe temporal variations. Signal-to-noise ratios (SNR) were also calculated whenever possible. At sea, relevant energy concentration was observed at low frequencies, guiding filtering towards the useful band. In the tank, the comparison between cables and coupling configurations revealed marked frequency-dependent differences, with better SNR in water-column configurations and worse performance when buried, as well as context-dependent effects when using resonators. In preliminary tests conducted within the Fibermars project, it was observed that buried cables could achieve better signal-to-noise ratios than cables laid on the seabed or in the water column. The apparently contradictory results obtained in the tank suggest that these poor results were due to flow noise generated by water over the cables. The tank tests were extremely important for the Fibermars project, as they were the first of their kind. However, they were limited by the characteristics of the source, which had a limited useful bandwidth and did not cover very low frequencies, and by the absence of absolute calibration, which meant that all measurements are relative to the baseline condition. Nevertheless, the results support the potential of DAS as a complement to traditional surveillance means and for enhancing maritime situational awareness, recommending additional campaigns with broader spectral coverage, calibration, and control of environmental variables.
Description
Keywords
Detecao Acustica Distribuida Fibra Otica Acústica Submarina Consciencia Situacional Maritima Cabos de Telecomunicações Submarinos Distributed Acoustic Sensing Optical Fiber Underwater Acoustics Maritime Situational Awareness Submarine Telecommunication Cables