Ultrasonic Time of Flight Estimation using Artificial Neural Networks

The measurement of wind speed is a key topic for the optimization of energy generation in wind farms. Currently, there are numerous ways to perform its estimation, however, the usual mechanical anemometers are rudimentary and present a considerable margin of error due to the friction between parts and the inertia of the system. Therefore, there is a greater tendency for the use of ultrasonic anemometers, which are based on the calculation of the time of flight of the ultrasonic wave, that is, the time interval required for a wave emitted from a transmitter ultrasonic transducer to reach a receiving transducer. Among the ways of estimating the time of flight of the ultrasonic wave, there are methods based on time intervals and others on digital signal processing, but there are, on the other hand, those based on artificial intelligence. In this article, an artificial intelligence model based on Artificial Neural Networks capable of estimating the time of flight of the ultrasonic wave is presented. Furthermore, the construction process is described, as well as its experimental results commented.