Use of MLP Neural Networks for the Conversion of Hot-Film Sensor Signals to Air Velocity Measurement

• Grant number: 25/08357-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: August 01, 2025
• End date: July 31, 2026
• Field of knowledge: Physical Sciences and Mathematics - Geosciences - Meteorology
• Principal Investigator: Livia Souza Freire Grion
• Grantee: Miguel Rodrigues Tomazini
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

This work proposes the use of an MLP neural network for converting air velocity data obtained from a hot-film sensor, transforming temporal voltage readings (Volts) into time series of velocity (m/s), with an emphasis on application to real-world data. This study continues previous research that implemented an MLP network trained exclusively on synthetic data. The current goal is to validate and extend this approach through the use of field-collected data under different environmental conditions, comparing the results with the analytical method to assess the feasibility and performance of the proposed methodology. To this end, adaptations and improvements will be made to the code to ensure its suitability for the new dataset. Through this process, the aim is to achieve a more conclusive analysis of the feasibility and effectiveness of using an MLP neural network for anemometric conversion, as well as to enhance the method's accuracy, contributing to studies in atmospheric turbulence and boundary layer modeling.