Long-Term Traffic Volume Prediction Based on Type-2 Fuzzy Sets with Confidence Interval Method

The paper uses Gaussian Interval Type-2 fuzzy sets theory to deal with the 24-h historical traffic volume data with uncertainty and randomness to get the 24-h prediction of traffic volume with higher precision. The central limit theorem is adopted to convert point data of mass traffic flow in sampling time period into interval data (also called confidence interval) which is being used to get the state variables of traffic volume. The historical traffic volume data are divided into a couple of intervals according to the similarity degree of the same sampling period for each day. Each interval represents a traffic state. Take these traffic states as data foundation, Markov chains model is built. Viterbi algorithm is used to find the maximum possible state sequence by decoding from Markov chain model by which the fuzzy inference system is constructed. The confidence interval data retain the uncertainty and randomness of traffic flow, meanwhile reduce the influence of noise from the detection data through the transition from traffic volume data to traffic state data. Fuzzy inference system based on Markov chains and Viterbi algorithm discusses the inherent relevance of traffic volume data. The combined Markov Gaussian Interval Type-2 fuzzy sets leads to interval forecasting traffic volume output with the ability of describing the possible range of the traffic volume as well as the traffic volume prediction data with high accuracy. The simulation results based on actual data show the validity and rationality of the model.