A Multi-Layer Evolution Algorithm for Optimizing Highway Alignments
Chi Zhang, Xiao-Min Yan, Xing Huang, Tian-Yu Zhang, Wen-Jun Zhu · CICTP 2017 · 2018
A multi-layer approach to optimize highway design is developed to significantly improve the efficiency and solution quality of traditional model. The multi-layer evolution algorithm is designed to achieve feasibility and effectiveness respectively adopting knowledge base and particle swarm algorithm before it is determined the recommendation by genetic algorithm. Such approach not only overcomes the shortcomings of the three existing algorithms but also simplifies the search space of every layer to a certain extent. Moreover, the model perfects the initial generation solution of the second and third layers, as well as the final solution quality. The algorithm is applied to a field road design program at Kunlun Mountain of Qinghai-Tibet Plateau. The results suggest that the multi-layer approach is promising for obtaining an optimal solution to satisfy design standards in reality and practice.