.png){kind=icon}
返回
A modified motion planning algorithm for horizontal highway alignment development
.png){kind=icon}
PRE
AI
DOI：10.1111/mice.12534.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
A horizontal alignment can be represented by three key factors: number of horizontal points of intersection (HPIs), their locations, and corresponding horizontal curve radii. Deciding all the three factors simultaneously requires extensive effort, which is not practically feasible in the manual alignment development process. Most available computer-aided methods prioritize some or all the three factors in the automated alignment development processes. However, approximation in HPI location or pre-selection of HPI number and curve radius are the few limitations of these methods. This study presents a modified motion-planning based algorithm for developing new horizontal alignments with optimized costs and impacts. It simultaneously uses a low-discrepancy sampling technique to develop increasingly dense potential HPIs, rapidly exploring random trees to find a suitable number of intermediate HPIs at appropriate locations and sequential quadratic algorithm to select optimally fitted curve radii. The proposed algorithm is integrated with the GIS database for realistic location-dependent cost and environmental impact assessment. Two real-world study areas were selected to compare the results with the one reported in the literature and to evaluate backtracking capability. Results indicated the proficiency of the proposed algorithm in developing new alignments. The sensitivity analyses revealed the effect of design speed and right-of-way width on the alignment generation. The proposed algorithm can automate the new horizontal highway alignment development process and support highway engineers in planning and development.
Keyword:
VERTICAL ALIGNMENT
DISTANCE TRANSFORM
GEOMETRIC DESIGN
OPTIMIZATION
MODEL
SEARCH
COST
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
C
IF:
9.1
论文数:
2.0K
被引数:
10.0K
机构
引用论文
.png){kind=icon}
分享收藏分享收藏
查看更多内容