Sensing in IoT for Smart City Systems
Volodymyr Kochan, Oleksandr Matsiuk, Nataliia E. Kunanets, Volodymyr V. Pasichnyk, Oleksiy Roshchupkin, Anatoliy Sachenko, Iryna Turchenko, Олексій Дуда, Віта Семанюк, Svitlana Romaniv · 2019
The Internet of Things (IoT) includes a large set of sensors of various physical quantities, operating principles and parameters. In this case, sensor errors are traditionally dominant in measuring channels. In this paper general methods of increasing the accuracy of sensors using neural networks are considered. Due to the generalization of properties, neural networks can significantly improve the accuracy of sensors with reduced complexity of the transition to their individual transformation functions.