An improved least square support vector regression algorithm for traffic flow forecasting

Due to the nonlinear, non-stationery, complex and stochastic characteristics of short-term traffic flow time series, traditional prediction methods do not work well. This paper presents a short-term traffic flow forecasting model based on the least square support vector regression (LSSVR) algorithm, which is optimized by a glowworm swarm optimization (GSO) algorithm. The GSO algorithm is used to determine two core parameters in the learning process, which significantly influence the predicting performance in the model. An actual example of traffic flow data on one section of highway in Chengdu, China is used to evaluate the performance of the proposed LSSVR-GSO model. The experimental results show that the proposed LSSVR-GSO model has more accurate predicting results than the LSSVR model optimized by the genetic algorithm and the back-propagation neural network model.