This paper presents a proactive model and tool for traffic analysis and management that integrates deep learning for traffic parameter prediction with microscopic traffic simulation providing traffic analysts with the ability to visualise the traffic network state ahead of time, generate traffic control measures, and visualise the consequences of the applied measure(s). The model adopts an integrated assess-forecast- simulate approach in which traffic flow characteristics are applied to deep Convolutional Neural Network and Long Short- Term Memory (CNN-LSTM) stacked autoencoders in order to forecast traffic flow and speed, which is subsequently passed on to a traffic microsimulation tool – Simulation of Urban Mobility (SUMO) – where the predicted parameters are used to generate a traffic future state simulation. We test our model using sensor- collected traffic and weather data from the geographical area of Greater Manchester, United Kingdom. The empirical results show the benefits of the model for urban traffic analysis.
|Name||2019 IEEE International Conference on Systems, Man and Cybernetics (SMC)|