Short-Term Traffic Flow Prediction Based on the Optimization Study of Initial Weights of the Attention Mechanism

Traffic-flow prediction plays an important role in the construction of intelligent transportation systems (ITS). So, order to improve accuracy short-term traffic flow prediction, a model (GWO-attention-LSTM) based on combination optimized attention mechanism and long memory (LSTM) is proposed. The LSTM uses assign individual weight feature information extracted via LSTM. This can increase model’s focus information. initial parameters are also using grey wolf optimizer (GWO). By simulating hunting process wolves, GWO algorithm calculates position maps it mechanism. In this way, short-time constructed. data trunk roads center Qingdao (China) used as research object. Multiple sets comparison models set up for analysis. results show that GWO-attention-LSTM has obvious advantages over other models. error MAE values decreased by 7.32% 14.35% average compared with attention-LSTM model. It concluded better performance provide effective help management control theory research.