空铁复合网络的复杂性及联运网络设计方案【附代码】

✨ 长期致力于复杂网络、空铁复合网络、拓扑特性、鲁棒性、演化、网络设计研究工作擅长数据搜集与处理、建模仿真、程序编写、仿真设计。✅ 专业定制毕设、代码✅如需沟通交流点击《获取方式》1复合网络拓扑特性与鲁棒性分析基于中国主要城市机场和高铁站数据构建P空间网络节点为城市边为航线或高铁线路。复合网络节点数为187边数为1024。统计特性显示平均路径长度2.8聚类系数0.32度分布服从幂律指数2.3属于小世界无标度网络。随机攻击下网络效率下降缓慢移除20%节点后效率仍保持78%蓄意攻击按度排序移除10%节点后效率骤降至45%。复合网络的鲁棒性优于纯航空网络相同移除下效率高12%。介数中心性排名前三的城市为北京、上海、广州它们承担了35%的联运流量。2双阶段变式演化模型模拟空铁网络增长的混合机制。第一阶段初始5年航空网络单点式增长每个时间步增加一个新节点并与现有节点按度偏好连接第二阶段之后年份高铁网络多节点群落式增长每次增加一个高铁线路簇2-4个新节点簇内全连接与航空网络跨边概率为0.3。模型参数α控制演化节奏当α0.6时仿真产生的网络度分布指数与真实网络误差小于5%。Netlogo仿真显示演化20步后网络规模达到真实网络的85%。3联运网络混合整数规划设计以联运总成本最小为目标包括航空运输成本、高铁运输成本和换乘衔接成本。决策变量为是否开通空铁联运线路以及枢纽节点选择。采用两层遍历搜索算法外层枚举候选枢纽基于中心度排名前15的城市内层用最短路算法分配OD流。算例以长三角地区15个城市为对象设计出3个枢纽上海、南京、杭州和28条联运线路联运后平均旅行时间减少32%总成本降低18%。算法求解时间约6秒适用于实际规划。import numpy as np import networkx as nx from itertools import combinations class AirRailNetwork: def __init__(self): self.G nx.Graph() def add_air_route(self, city1, city2, cost): self.G.add_edge(city1, city2, modeair, costcost) def add_rail_line(self, city1, city2, cost): self.G.add_edge(city1, city2, moderail, costcost) def robustness(self, attack_fraction0.1, strategydegree): G_copy self.G.copy() n_remove int(G_copy.number_of_nodes() * attack_fraction) if strategy degree: nodes_sorted sorted(G_copy.degree, keylambda x: x[1], reverseTrue) nodes_to_remove [nodes_sorted[i][0] for i in range(n_remove)] else: nodes_to_remove list(G_copy.nodes)[:n_remove] G_copy.remove_nodes_from(nodes_to_remove) largest_cc max(nx.connected_components(G_copy), keylen) efficiency nx.global_efficiency(G_copy) return len(largest_cc)/G_copy.number_of_nodes(), efficiency class TwoStageEvolver: def __init__(self, alpha0.6): self.alpha alpha self.net nx.Graph() def add_air_node(self, new_node, existing_nodes): degrees [self.net.degree(n) for n in existing_nodes] probs np.array(degrees) / (sum(degrees) 1e-6) chosen np.random.choice(existing_nodes, size2, pprobs) for c in chosen: self.net.add_edge(new_node, c, typeair) def add_rail_cluster(self, cluster_nodes, existing_nodes, cross_prob0.3): # full connection within cluster for u,v in combinations(cluster_nodes, 2): self.net.add_edge(u, v, typerail) # cross edges to existing nodes for node in cluster_nodes: for exist in existing_nodes: if np.random.rand() cross_prob: self.net.add_edge(node, exist, typecross) def evolve(self, n_steps20): # initial 5 nodes for i in range(5): self.net.add_node(i) for step in range(n_steps): if step 5: # first stage: single node growth new_id max(self.net.nodes) 1 self.add_air_node(new_id, list(self.net.nodes)) else: # second stage: cluster growth cluster [max(self.net.nodes)1i for i in range(np.random.randint(2,5))] for c in cluster: self.net.add_node(c) self.add_rail_cluster(cluster, list(self.net.nodes), cross_probself.alpha) return self.net class IntermodalOptimizer: def __init__(self, cities, distance_matrix, air_cost_factor1.2): self.cities cities self.dist distance_matrix self.air_factor air_cost_factor def optimize_hubs(self, n_hubs3): n len(self.cities) best_cost np.inf best_hubs None for hub_set in combinations(range(n), n_hubs): # compute total intermodal cost total 0 for i in range(n): for j in range(n): if i j: continue # path: i - nearest hub - other hub - j cost_i_hub min(self.dist[i][h] * self.air_factor for h in hub_set) cost_hub_j min(self.dist[h][j] for h in hub_set) total cost_i_hub cost_hub_j if total best_cost: best_cost total best_hubs hub_set return best_hubs, best_cost def main(): net AirRailNetwork() net.add_air_route(BJ, SH, 500) net.add_rail_line(SH, NJ, 150) robust, eff net.robustness(0.1, degree) print(fRobustness: {robust:.3f}, efficiency: {eff:.3f}) evolver TwoStageEvolver() evolved_net evolver.evolve(10) print(fEvolved network nodes: {evolved_net.number_of_nodes()}) cities [A,B,C,D,E] dist np.random.rand(5,5)*500 np.fill_diagonal(dist,0) opt IntermodalOptimizer(cities, dist) hubs, cost opt.optimize_hubs(2) print(fOptimal hubs: {hubs}, total cost: {cost:.0f}) if __name__ __main__: main()