军事知识服务平台
基于DT-RRT混合算法的山区公路多维约束智能选线
Intelligent Route Selection for Mountainous Highways based on DT-RRT Hybrid Algorithm with Multi-Dimensional Constraints
摘要 随着我国公路建设向纵深发展,大批新建项目处于艰险山区,选线设计面临多维耦合约束等挑战,既有智能选线算法常难以产出优化解.对此,首先建立了复杂约束条件下的山区公路智能选线优化模型,此后基于距离变换(DT)算法框架,设计了针对 DT 搜索"开始前、过程中、结束后"3 阶段的多维耦合约束动态检测与处治方法;此外,针对现有 DT 算法"强于局部搜索、弱于全局感知"的问题,引入快速探索随机树(RRT)算法,采用 DT 算法扫描"已知"区域、借助 RRT 算法探索"未知"区域,同时采用多准则锦标赛决策方法融合 DT 与 RRT 搜索结果,实现了基于 DT-RRT 混合算法的智能选线启发式搜索与优化.最终,通过某复杂山区公路选线案例,对 DT-RRT 混合算法与传统人工设计、既有 DT 算法进行了详细对比分析与有效性验证.
Abstract:
With the expansion of highway construction in China,a large number of new projects are located in rugged mountainous areas,where route selection and design face challenges such as multi-dimensional coupling constraints.Existing intelligent route selection algorithms often struggle to produce optimized solutions.To address this,an intelligent route selection optimization model for mountainous highways under complex constraint conditions is first established.Based on the distance transform(DT)algorithm framework,a dynamic detection and treatment method for multi-dimensional coupling constraints is designed for the three stages of DT search:before initiation,during execution,and after completion.Furthermore,to overcome the limitation of existing DT algorithms in being"strong in local search but weak in global perception,"the rapidly exploring random tree(RRT)algorithm is introduced.The DT algorithm is employed to scan"known"areas,while the RRT algorithm is utilized to explore"unknown"areas.A multi-criteria tournament decision-making method is adopted to integrate the search results of DT and RRT,enabling heuristic search and optimization for intelligent route selection based on the DT-RRT hybrid algorithm.Finally,through application to a case study of route selection for a complex mountainous highway,a detailed comparative analysis and validation of the effectiveness of the proposed DT-RRT hybrid algorithm are conducted against traditional manual design and existing DT algorithms.
刊 名 公路 ISTICPKU
年,卷(期): 2026, 71(9)
分类号: U412.32
机标分类号: TP311.13 E911 TU984
在线出版日期: 2026年10月14日
基金项目: 2025YB035:江西省交通运输厅科技项目,