Application of back propagation neuron network on data linkage transmission of semiconductor hydrogen detection device

Abstract

In this study we have fabricated a hydrogen detection device based on semiconductor type with the advantages of smaller in size, higher sensitivity and capable to operate under room temperature. We have also demonstrated an improvement by applying back propagation neuron network to the system in order to minimize the amount of data transmission without influencing data quality. In the practical situation that monitoring the designate sites requires multiple sensing and wireless data transmission. Such function requires massive data transmission volume and would normally result higher power consumption and data lost in transmission. This study has developed a self-made hydrogen detection device and applied neuron network to assist the real time detection system deducing transmission volume. The empirical result has successfully reduced 97.4% of the transmitted data without sacrificing the original data quality. Which suggested an effective methodology to reduce data transmission volume requirement and conserve detection device power usage.