Bilevel Programming Model of Discrete Traffic Network Based on Machine Learning Optimization Hybrid Algorithm

Abstract

With the rise of living standards, more and more people choose to travel by car, which increases the pressure on the urban traffic network (TN). The TN needs to be re planned and designed. The main goal of the design, organization and management of the TN is to improve the service level and access to the TN. During the planning process, a reasonable and scientific TN model should be developed according to the characteristics of traffic operation in different regions. Therefore, this paper studies and designs the bilevel programming model of discrete TN based on the hybrid algorithm of machine learning and optimization. This paper first describes the construction of the bilevel programming model and the network structure, then designs the hybrid algorithm and sets the weight coefficient, and finally analyzes the sensitivity of the comparison of the average load degree of the road section and the time budget.