# Urban connected vehicle lane planning based on improved Frank Wolfe algorithm

## TL;DR

This paper proposes an improved lane planning model for urban connected vehicles using enhanced optimization algorithms to improve traffic safety and efficiency.

## Contribution

The novelty lies in combining improved whale optimization and Frank-Wolfe algorithms for urban connected vehicle lane planning.

## Key findings

- The improved whale optimization algorithm achieved 95.27% accuracy and 92.65% recall on the Iris dataset.
- The improved Frank Wolfe algorithm converges in about 30 iterations with a convergence limit of 10-4.
- The model reduces road system travel cost as the fairness index threshold increases.

## Abstract

As the new generation of information technology matures and improves, the functions of intelligent connected vehicles become more and more perfect, and the number of urban connected vehicles is also increasing. To provide an effective optimization scheme to the mixed traffic flow road network in the networked environment, the study investigates the lane planning for urban connected vehicles method. First, a lane planning for urban connected vehicles bi-level programming model is constructed. Then, the upper-level model is solved using improved whale optimization, and the lower-level model is solved using improved Frank-Wolfe algorithm. The results showed that the accuracy and recall of the proposed improved whale optimization algorithm on the Iris dataset were 95.27% and 92.65%, respectively, which were superior to traditional whale optimization algorithm, moth flame optimization algorithm, moth flame optimization algorithm combined with chaos strategy, and adaptive firefly optimization algorithm. The proposed improved Frank Wolfe algorithm can converge at around 30 iterations, with a convergence limit of around 10-4, which is superior to the traditional Frank Wolfe algorithm. The minimum total travel cost of the road system gradually decreases with the increase of the fairness index threshold. The experimental results demonstrate the effectiveness of the proposed urban connected vehicle lane planning model and solving algorithm. The research results contribute to improving the operational safety and efficiency of the road network TS, thereby improving the current traffic situation of the urban TS.