Abstract
This thesis is a further study on the thesis in which we searched for a suitable system to take wave measurements on the pond at the Antwerp Maritime Academy in order to predict the waves, potentially with a neural network. In addition, we need a sensor that can measure the waves on a smaller scale, in a test basin e.g. The research method used is a literature study to describe the functioning of the systems, complemented by practical implementation of developing the sensor, comparing it with the WaveGauge and using the measured data in a neural network. We succeeded in developing an accurate sensor using an Arduino microcontroller for use in our test tank. This was compared with Obscape's wave gauge, and this showed that the sensors are comparable. Fortunately enough, we noticed a little flaw here that is in the self-designed sensor, so we were able to correct it. Finally, we managed to use a neural network to predict the waves less than 1 cm accurately in 70% of cases with wind data as input.