A novel fractional order partial grey model for forecasting short-term traffic flow

Abstract Short-term traffic flow forecasting is an integral part of intelligent transportation systems, and a key foundation for traffic departments and travelers to obtain reliable traffic information. This paper constructs a fractional order partial grey model with a sine function control matrix by combining a partial differential equation. The model is capable of predicting short-term traffic flow with spatiotemporal and oscillatory characteristics. In the new model, the introduction of a fractional order accumulation operator may help reduce random fluctuations in data and enhance the flexibility of the model. Moreover, the model dynamically adjusts its parameters using a sine function control matrix, allowing it to better adapt to data with different oscillation types, thereby improving the versatility of the model. Additionally, the characteristic curve method is used to provide the time response function of the model, and the particle swarm optimization algorithm is employed to optimize the model parameters. To demonstrate the validation of the new model, two different road segments from two different highways were selected as research objects. Compared with two other partial grey models and two other algorithm models, the new model has the highest accuracy in simulation and prediction. Furthermore, this new model is applied to the simulation and prediction of the short-term traffic flow, and the results show that the simulated and predicted data have great similarity with the real data trend, which exhibits good reliability and applicability in solving the short-time traffic flow prediction problem.