Smart Traffic Flow Control System for City Governance Driven by Neural Network Methods

Abstract

Recent advancements in intelligent transportation systems (ITS) like Internet of Things (IoT), machine learning, and deep learning methods are employed for effective traffic flow prediction governing smart traffic control. In this paper, a deep-learning-based multi-neural network method framework is proposed to short-term traffic forecasting which called as Smart Traffic Flow Control System for City Governance driven by Neural Network Methods(STFCS−NNM). With respect to the key difference it makes, the method employs recurrent neural networks (RNNs), namely long short-term memory (LSTM) and gated recurrent unit (GRU) architectures to recognize temporal dependencies and nonlinear correlations within traffic data. The models are trained and checked on real-world traffic data collected from the Caltrans Performance Measurement System (PeMS) aggregated in 5-minute intervals at IoT-enabled sensors. Compared with traditional statistical models such as ARIMA and SARIMA, experimental results show that the proposed models have better prediction accuracy, which validates their effectiveness in the field of traffic flow dynamics modeling and intelligent traffic management.