数学模型作业(3)

  1. 数学模型作业(2)中计算欧式距离有误,已更改。
  2. 内容(代码)接数学模型作业(2)
  3. 初学python和数学模型,不足之处请大佬指出。
  4. 给个赞(上次已经骗过关注了~~~)
  1. 导入相关参数
shape = 613
alpha1 = 25
alpha2 = 15
beta1 = 20
beta2 = 25
theta = 30
delta = 0.001
  1. 建立 List 格式的垂直校正点集合 V 和水平校正点集合 H,保存在一个 txt文件中。
# 建立垂直校正点集合V和水平校正点集合H
V = []
H = []
VH = []
for i in range(0, shape):
    if prop[i] == 1:
        V.append(i)
    if prop[i] == 0:
        H.append(i)
VH = V + H
file = open('校正点集合.txt', 'w')
for i in range(len(VH)):
    s = str(VH[i]) + '\n'
    file.write(s)
file.close()
  1. 用 python 输出减少边之后最短路模型的邻接矩阵,存放在
    excel 文件中。
  • 分析:
    (1) 剪枝可以剪去任意点为起点到垂直校正点不符合垂直校正条件的枝
    (2)剪枝可以剪去任意点为起点到水平校正点不符合水平校正条件的枝
    (3) 剪枝可以剪去以B点为终点,距离超过 θ / δ \theta/\delta θ/δ的枝
  • 安装gurobi,参考另一篇(Gurobi9.0.3安装
  • 将距离矩阵dist转换为tupledict类型,用dict_dist存放邻接矩阵(要调用gurobi库:from gurobipy import *)
# 将距离矩阵dist转换为tupledict类型,用dict_dist存放邻接矩阵
dict_dist = {
    
    }
for i in range(shape):
    for j in range(shape):
        dict_dist[i, j] = dist[i][j]
dict_dist = tupledict(dict_dist)
  • 剪枝实现
# step1:剪去到垂直校正点不符合要求的边
for i in range(1, shape-1):
    for j in V:
        if dist[i][j] > min(alpha1, alpha2) / delta:
            dict_dist[i, j] = 0
# step2:剪去到水平校正点不符合要求的边
for i in range(1, shape-1):
    for j in H:
        if dist[i][j] > min(beta1, beta2) / delta:
            dict_dist[i, j] = 0
# step3:剪去距离终点B超过30000m的边
for i in range(shape-1):
    if dist[i][shape-1] > theta / delta:
        dict_dist[i, shape-1] = 0
# 定义边集
edge = []
for i in range(shape):
    for j in range(shape):
        if dict_dist[i, j] != 0:
            edge.append((i, j))
print("剪枝之后的边数:", len(edge))
  • 邻接矩阵存入excel表格
# 将 剪枝后的邻接矩阵写入到 dict_dist_save.xlsx 中
dict_dist_save = np.zeros(shape=(shape, shape))
for i in range(shape):
    for j in range(shape):
        dict_dist_save[i][j] = dict_dist[i, j]
dict_dist_save = pd.DataFrame(dict_dist_save)
writer = pd.ExcelWriter('dict_dist.xlsx')
dict_dist_save.to_excel(writer, 'sheet1')
writer.save()
  1. 结果
    在这里插入图片描述在这里插入图片描述

猜你喜欢

转载自blog.csdn.net/qq_40678163/article/details/109111996