A Short-Term Traffic Flow Forecasting Model Based on OOA-BP Neural Network

Short-term traffic flow forecasting is the key to alleviating traffic congestion. Accurately forecasting the traffic flow of a certain road section within a short time can greatly improve the traffic efficiency, so this paper constructs a traffic flow forecasting model based on the Osprey Optimization Algorithm (OOA) optimized BP neural network (BPNN). To prevent the disadvantages of the BPNN, such as the randomness of weights and thresholds, which leads to the consumption of a large amount of training resources and reduced prediction accuracy during the prediction process, this paper adopts the OOA to optimize the BPNN. This method has strong global search and local search capabilities and fast convergence speed. Since it can accelerate the training speed of the BPNN and effectively solve the problem of the prediction model being trapped in a local optimum, the optimal solution can be found within a relatively short time. The experimental results show that under the same data set, OOA-BPNN model outperforms BPNN model, particle swarm optimization (PSO), genetic algorithm (GA) and other optimization models in terms of mean absolute error (MAE) and mean absolute percentage error (MAPE).