Scientists have shown that the process of formation of traffic jams in the cities can be described using the model of a SIR model that predicts the spread of infectious diseases. Simulation of traffic in Melbourne and five other cities using the new model showed good agreement with real data, write the scientists in Nature Communications.
Although traffic congestion affect the lives of almost every resident of the city, the scientists there is no universal model that would predict the formation and disappearance of road congestion. Many modern approaches to the description of urban traffic based on microscopic models, which require high computing power and complex calibration. In addition, the poor in cities due to insufficient number of registered devices, it is impossible to obtain enough data for setting initial conditions of the model. Other (macroscopic) models, usually implemented using neural networks, and are often unable to give detailed quantitative information on traffic congestion in a particular place.
A team of scientists under the leadership of Miad Saberi (Meead Saberi) from the University of New South Wales had drawn inspiration to create a new model of traffic in a rather unexpected place — the attention of researchers was attracted to the epidemiological SIR-model (from the English words Susceptible — Infected — Recovered), which is now constantly used for modeling influenza epidemics and other diseases in large cities. Its popularity SIR-model was for ease of construction and use. It consists of three differential or difference equations that describe the dynamics of changes in the number of cases, recovered and susceptible to illness (those who can be infected). The number of cases increases, reducing the number susceptible to the disease. Sick and then recover — the number of cases decreases, increasing the number recovered.
If you replace the number of cases of people on the number of cars standing in a traffic jam, and healthy and cured on coming to the jam and leaving him, and then to modify slightly the system of equations — obtain a simple model describing the congestion of the city. Scientists tested their model in seven cities to prove their assumption: the increase and decrease in congestion of roads of the cities is “infectious” nature. To do this, they took Google maps about traffic six major cities (Paris, London, Melbourne, Sydney, Chicago and Montreal) and compared the predictions of their model on the share of loaded transport knots in the total number of road network with real data. A more detailed comparison of the predictions of their model, the scientists have produced, with advanced data on traffic congestion in Melbourne. Despite the differences in all the studied cities, the scientists saw a similar “infectious” dynamics of the formation and dissolution of traffic jams.