Analysis of Sensor Data for Hydrogen Production in a Biofilm Photoreactor Using Multilayer Perceptron Network

Neural Networks are one of the most appreciate techniques in the field of the analysis of the data set. In this paper the usage of multilayer perceptron networks (MLP) for the prediction of the hydrogen production from the sensor data is presented. The results with R2 value of over 0.95 show clearly, that it is possible to build an effective system for the prediction of the hydrogen production and concentration rates based only on the data covering biofilm thickness.