GA框架下融合多目标优化的禁行区域规划算法设计

Design of multi⁃objective optimization algorithm for restricted area planning under the GA framework

在城市环境中,建设工程不仅会存在安全风险,还可能导致交通拥堵、事故和延误等问题。为保证公共交通畅通和施工现场的安全,文中在GA框架下提出了一种融合多目标优化的禁行区域规划算法。采用区域生长算法,利用马尔可夫链分析交通节点的状态转移概率,明确节点之间的关联度并构建谱聚类的相似图,提取施工禁行区域规划单元的路网。再应用遗传算法进行迭代寻优计算,求出最优施工禁行区域规划方案。实验结果表明,所提算法给出的施工禁行区域规划方案实施后,整个禁行期间交通冲突次数集中在0~50次之间,冲突次数频率的数值约为10,更好地保证了交通运行安全。

In urban environments,construction projects not only pose safety risks,but may also lead to problems such as traffic congestion,accidents,and delays.In order to ensure the smooth flow of public transportation and the safety of construction sites,the design of a multi-objective optimization algorithm for restricted area planning under the GA framework is proposed in this article.Observing the construction affected area and using the region growth algorithm,the construction affected area is divided into multiple prohibited traffic planning units.Using Markov chain analysis to analyze the state transition probability of traffic nodes,clarify the correlation between nodes,and construct a similarity graph of spectral clustering to extract the road network of construction restricted area planning units.Establish a planning model for construction restricted areas with the goal of minimizing differences in traffic complexity and regulatory loads.Apply Genetic Algorithm for iterative optimization calculation to find the optimal construction restricted area planning scheme.The experimental results show that after the implementation of the construction prohibited area planning scheme provided by the studied algorithm,the number of traffic conflicts during the entire prohibited period is concentrated between 0 and 50,the frequency of conflicts is around 10,which better ensures traffic operation safety.