Predictive mathematical modelling devised at PolyU supports the development of real-time, intelligent traffic management systems.
Many urban areas in the Asia-Pacific region, such as Bangkok in Thailand, suffer from traffic congestion. Therefore, it is important to have better management of transportation arteries. To help highway managers predict traffic conditions and prevent congestion, Dr Agachai Sumalee at PolyU’s Department of Civil and Environment Engineering has developed the Stochastic Cell Transmission Model (SCTM).
In the SCTM, each of five operational modes corresponds to a specific level of congestion on freeway segments that is formulated as a discrete time bilinear stochastic system. A set of conditions then characterizes the probability of each mode’s occurrence. Derived from the theory of finite mixture distribution, the joint traffic density estimates the overall effect of the five modes. Using predictive mathematics, the SCTM allows the analysis of minor events that lead to major delays and thus modelling traffic flow density on freeway segments with stochastic demand and supply.
In terms of surveillance, the SCTM can be used to provide short-term predictions using historical and online data of travel demand and traffic state. In addition, with its use of stochastic delay the model is effective for long-term traffic prediction.
The SCTM has been used by the Thai Expressway Authority to improve road traffic and around Bangkok, successfully streamlining the daily commute in an area serving more than 14 million people.
As Dr Sumalee put it, “we are the first research team to successfully develop a mathematic model that works for creating real-time, intelligent traffic management systems”. With his research excellence in transportation systems, transport modelling and planning, Dr Agachai Sumalee earlier won the 2014 Asia-Pacific Economic Cooperation (APEC) Science Prize for Innovation, Research and Education. ♦