Detección, conteo y caracterización de vehículos en vías alternas mediante señales registradas con sismómetros para tiempos determinados

The study seeks to detect, count and characterize road traffic events by means of a vehicle measurementimplemented on 3 alternate roads in the Buenos Aires neighborhood of the city of Medell ́ın, using twotypes of seismometer arrangements; linear and two-dimensional, in order to capture its vibration signals,using the Linux operating system, 2 seismic stations and 4 geophones that capture these disturbances.Once the signals are obtained, they are analyzed using two programming languages; Python and Matlab,used to carry out two types of event recognition machine learning from two reference signals that arechosen for the training of the two learning machines. Then, from the measurements in the configuration ofthe two-dimensional and linear reference seismometer arrangements, the results are studied to find whatpercentage of effectiveness they present and the utility in recognizing vehicle capacity; thus determiningwhich of the seismic array configurations is most efficient for capturing road traffic signals.

• Autores: Gallego Guzmán, Sebastián; Flórez López, Alejandro
• Fecha de publicación: 2020
• Institución: Universidad de San Buenaventura
• Repositorio: Repositorio USB
• Idioma: spa
• OAI Identifier: oai:bibliotecadigital.usb.edu.co:10819/8034
• Acceso en línea: http://hdl.handle.net/10819/8034
• Palabra clave: Medición; Identificación; Evento; Dato; Traza; Señal de vibración; Geófono; Aprendizaje automático; Diagrama de caja; Arreglo bidimensional; Arreglo lineal; Transformada wavelet; Transformadade fourier; Descriptores y/o predictores; Intervalo de tiempo; Measurement; Identification; Event; Data; Trace; Vibration signal; Geophone; Machine lear-ning; Boxplot; Two-dimensional array; Linear array; Wavelet transform; Fourier transform; Descriptors orpredictors; Time interval; Vehículos; Señales
• License: Atribución-NoComercial-SinDerivadas 2.5 Colombia