Traffic Flow Prediction using Recurrent Neural Network and LSTM
Recently, the problem of traffic has gotten worse due of industrialization, especially in major urban areas. People are increasingly choosing to use their own automobiles over public transit due to the greater level of life in cities. As a result, there is a significant rise in traffic, which causes a number of other related problems. One must spend longer time travelling than they would have otherwise, which also increases the amount of gasoline consumed, in order to go from one point to another. Numerous detectors are used to collect traffic data, and this data is very dynamic and non-stationary. For instance, it is impossible to declare with better certainty how many cars are turning at a junction. As a result, creating a mathematical model to calculate green time is difficult. However, the development of Intelligent Transportation Systems (ITS) in contemporary times has made it possible to monitor traffic incidents, communicate, process information, and take user action. The capacity to correctly forecast the traffic stream's pattern is one of the most important requirements for this system to function well. In this study, a method using artificial neural networks and real-time traffic data was suggested to forecast future traffic flow. By studying the past, neural networks are able to predict the future. This study specifically forecasts traffic volume at two levels: short-term for intervals of 15 minutes, and mid-term for intervals of one hour throughout a particular day. Applications for dynamic traffic management that use these predictions include signal control, congestion management, and trip time forecasting, among others.
Keywords – Traffic Prediction, Neural Networks, RNN model, Long-Short term memory network