使用 pyomo 和 gurobi 尋找解決方案

Question

我有一個使用 pyomo 和 gurobi 的代碼，但是當我嘗試運行它時，它並沒有給我最佳解決方案，只是說它們存在。 我能做些什么？

我有這個：

TransformationFactory('gdp.bigm').apply_to(model)
opt = SolverFactory("gurobi", solver_io="python") 
opt.options['NonConvex'] = 2
opt.solve(model).write()

它給了我這個：

# ==========================================================
# = Solver Results                                         =
# ==========================================================
# ----------------------------------------------------------
#   Problem Information
# ----------------------------------------------------------
Problem:
- Name: unknown
  Lower bound: 3722.006655208413
  Upper bound: 3722.0163664746015
  Number of objectives: 1
  Number of constraints: 98859
  Number of variables: 57239
  Number of binary variables: 13996
  Number of integer variables: 13996
  Number of continuous variables: 29247
  Number of nonzeros: 200378
  Number of solutions: 10
# ----------------------------------------------------------
#   Solver Information
# ----------------------------------------------------------
Solver:
- Name: Gurobi 9.52
  Status: ok
  Wallclock time: 56.91826629638672
  Termination condition: optimal
  Termination message: Model was solved to optimality (subject to tolerances), and an optimal solution is available.
 ----------------------------------------------------------
   Solution Information
 ----------------------------------------------------------
Solution:
- number of solutions: 0
  number of solutions displayed: 0

Answer 1

當pyomo從求解器接收到解時，它會保留加載到所有變量中的最優值。 所以，你有很多選擇。

如果您想查看整個 model，您可以display哪些替換值：

model.display()

或僅針對一個變量

model.x.display()

或者您可以查詢對象的結果，例如您的目標 function

value(model.obj)   # assumes your objective function is named obj

這里的value是pyomo function，所以如果你導入和重命名（推薦），比如：

import pyomo.environ as pyo

那么它將是

pyo.value(model.obj)