Innovative Intelligent Algorithms Assisted by Statistical Models for Urban Traffic Flow Control

This paper focuses on the problem of urban traffic flow control and proposes an innovative solution based on an intelligent algorithm assisted by a statistical model to address the increasingly complex traffic conditions. This paper integrates multi-source data from road sensors, GPS devices, and traffic management systems, extracts temporal features, spatial features, and traffic features from the preprocessed data, uses a linear regression model, solves the model parameters through the least squares method, and uses the cross-validation method to evaluate the model. In addition, Q-learning is introduced to adjust the traffic signal control strategy. The effectiveness of the strategy in this paper in practical application is verified by comparing it with the traditional fixed signal light control strategy. The congestion index is reduced and the maximum number of vehicles passing is increased to 1517 vehicles/minute. The solution proposed in this paper not only realizes the accurate prediction of urban traffic flow but also alleviates traffic congestion and improves road traffic efficiency by dynamically adjusting the signal control strategy.