基于粒子群算法的改建道路纵断面自动设计研究

董华珍，聂涔

摘要:

为改变现有改建道路纵断面设计软件拉坡设计人工耗时长、随意性大、主观性强、设计方案参差不齐的状况，本文首次提出了基于粒子群算法的改建道路纵断面自动设计，以提高设计效率、减少工程投资。文章在满足相关规范、路面改造方案前提下，以改造纵断面填挖方高度总和最小为目标函数，以纵坡长度、纵向坡度、竖曲线半径为约束条件，以变坡点的里程、高程为变量，利用粒子群算法原理，构建了基于粒子群算法的改建道路纵断面自动设计模型。研究结果表明：该设计模型达到了改建道路纵断面自动适应贴合现状道路高程、路面改造工程量最小的目的，当粒子群算法迭代次数达到400多次时，粒子适应度函数趋于收敛，当迭代次数设为500次时，可以获得满意的成果。实例验证表明：与传统的人工设计相比，算法自动设计的效率更高，自动设计过程仅需2~3s，拟合原地面线的效果更好，能节省近一半的路面加铺工程量，优化设计效果显著。基于粒子群算法的改建道路纵断面自动设计，可为道路改造设计及设计咨询提供了理论基础与技术支持。

Abstract:

The existing longitudinal section design software of reconstructed road pull-up design is time-consuming, arbitrary, subjective and the design scheme is uneven. In order to improve the design efficiency and reduce the project investment, for the first time，the automatic design of reconstructed road profile based on particle swarm optimization algorithm was proposed.Under the premise of satisfying relevant codes and the pavement reconstruction plan, the objective function was created to take the minimize value of the total fill and excavation height of the reconstructed longitudinal section. The function was constrained by the length of the longitudinal slope, the longitudinal declivity, and the radius of the vertical curve.The variables of the function were the position and elevation of the variable slope point.The objective function used the principle of particle swarm optimization algorithm, and constructed the automatic design model of the reconstructed road profile based on particle swarm optimization algorithm.The research results show that the design model achieves the purpose of minimizing the existing road elevation and road reconstruction project. When the number of iterations of particle swarm optimization algorithm reaches 400 times, the particle fitness function tends to converge. When the number of iterations is set to 500 times, satisfactory results can be obtained. The example shows that compared with the traditional manual design, the efficiency of the automatic design algorithm is higher, the automatic design process only takes 2~3 seconds, the effect of fitting the original ground line is better, the investment of road engineering reconstruction can be saved nearly half, and the optimization design effect is remarkable. The automatic longitudinal section design of reconstructed road based on particle swarm optimization, provides theoretical basis and technical support for the design and consultation of road reconstruction.