标题： 基于高维扩展的模拟分岔用于现实网络中的交通信号优化 显示英文标题

标题： Simulated Bifurcation with High-dimensional Expansion for Traffic Signal Optimization on Real-world Networks

摘要： 随着城市化进程的加快和交通拥堵的加剧，优化交通信号系统以提高道路通行能力和缓解拥堵已成为一个关键问题。 本研究提出了一种基于现实道路拓扑结构和典型四向八相交通信号控制方案的短期交通预测模型。 该模型考虑了不同方向的交通流量差异和信号相位变化频率，将这些因素整合到全局交通优化的优化目标中。 该目标函数的结构类似于统计物理中的自旋玻璃系统。 引入了模拟分叉优化算法，并以传统的模拟退火作为基准。 结果表明，模拟分叉在效率和效果方面均优于模拟退火。 利用北京的实际交通流量和道路网络数据，我们初始化了模型并进行了数值优化实验。 结果表明，模拟分叉在计算效率方面显著优于模拟退火，能够有效解决具有多重自旋相互作用的组合优化问题，并将时间复杂度降低到$O(N^{1.35})$。 该解决方案解决了全局交通信号优化的 NP 难问题。 重要的是，模拟分叉生成的信号相位模式符合实际交通信号系统的运行要求，展示了其在优化大型复杂城市交通网络信号控制方面的潜力。 这项工作为未来的城市交通管理和智能交通系统提供了坚实的理论和实践基础。 显示英文摘要

摘要： With accelerating urbanization and worsening traffic congestion, optimizing traffic signal systems to improve road throughput and alleviate congestion has become a critical issue. This study proposes a short-term traffic prediction model based on real-world road topologies and a typical four-way, eight-phase traffic signal control scheme. The model accounts for traffic flow disparities across directions and signal phase change frequencies, integrating these factors into an optimization objective for global traffic optimization. The structure of this objective function is similar to spin-glass systems in statistical physics. A Simulated Bifurcation optimization algorithm is introduced, with traditional simulated annealing as a benchmark. The results show that Simulated Bifurcation outperforms simulated annealing in both efficiency and effectiveness. Using real traffic flow and road network data from Beijing, we initialized the model and conducted numerical optimization experiments. The results indicate that Simulated Bifurcation significantly outperforms simulated annealing in computational efficiency, effectively solving combinatorial optimization problems with multiple spin interactions, and reducing the time complexity to $O(N^{1.35})$. This solution addresses the NP-hard problem of global traffic signal optimization. Importantly, the signal phase patterns generated by Simulated Bifurcation align with the operational requirements of real traffic signal systems, showcasing its potential in optimizing signal control for large, complex urban traffic networks. This work provides solid theoretical and practical foundations for future urban traffic management and intelligent transportation systems.