Untitled

Abstract

With increased technology development, governments and specifically urban managers require the most recent technology to make their smart cities safer. Intelligent transportation systems (ITS) have been highly incorporated within this area to improve effective smart city applications. A digital twin (i.e., a copy of a real-world construct) is required to manage a building structure and resources during smart city construction to optimize the overall perf ormance. This research objective can be achieved by incorporating big data techniques that maximize available resources with the help of a graphical processing unit (GPU) that can manipulate complex data successfully. The GPU integrates fuzzy rough set theory (FRST) and a multilayer perceptron neural model (MNM) to minimize the multi-classification problem during analysis. This neural model uses multiple layers that can understand the data structure and classify resources with minimal computational errors. It uses the AdaBoost learning parameter to minimize complexity while making decisions in smart city applications. This framework uses data from the Internet of Things (IoT) to manage resources and energy in a smart city. The FRST-enabled MNM approach explores the gathered data with appropriate forecasting accuracy while minimizing the running time and computation complexity and increasing the processing speed.