Integrated optimization of dynamic lane allocation and signal timing at urban intersections

Allocating spatiotemporal resources dynamically is deemed to be an effective technique for accommodating traffic arrivals at urban intersections. This paper develops a two-layer optimization framework for integrated lane allocation and signal timing at isolated intersections. The lane allocation at the upper-layer is formulated as an integer linear programming to minimize the sum of the critical flow factor. Whereas the lower-layer signal control problem is formulated as an integer quadratic programming to minimize the average vehicle delay. An iterative solution procedure is then introduced. Considering that frequent lane switching may affect the control performance, this paper proposes a dynamic implementation scheme to regulate the lane switching frequency. Numerical tests show that the proposed integrated optimization method outperforms the baseline method in reducing delays by 19.08%. Moreover, for the integrated optimization, further applying the proposed dynamic implementation scheme reduces the lane switching frequency, which effectively balances switching frequency and overall control performance.