Pyomo代码无法创建glpk求解器的问题咨询

问题重现 #

此前正常使用Pyomo，现在运行建模代码时出现以下错误：

PS C:\Users\oiste\Downloads>  & 'c:\Users\oiste\anaconda3\python.exe' 'c:\Users\oiste\.vscode\extensions\ms-python.debugpy-2024.2.0-win32-x64\bundled\libs\debugpy\adapter/../..\debugpy\launcher' '55445' '--' 'c:\Users\oiste\Downloads\Assignment7_BaseCase.py'
ERROR: Unknown IO type: python
Traceback (most recent call last):
  File "c:\Users\oiste\Downloads\Assignment7_BaseCase.py", line 140, in <module>
    results = opt.solve(model,tee=True)
              ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\Users\oiste\anaconda3\Lib\site-packages\pyomo\opt\base\solvers.py", line 102, in solve
    self._solver_error('solve')
  File "c:\Users\oiste\anaconda3\Lib\site-packages\pyomo\opt\base\solvers.py", line 119, in _solver_error
    raise RuntimeError(
RuntimeError: Attempting to use an unavailable solver.

The SolverFactory was unable to create the solver "glpk"
and returned an UnknownSolver object.  This error is raised at the point
where the UnknownSolver object was used as if it were valid (by calling
method "solve").

The original solver was created with the following parameters:
        solver_io: python
        type: glpk
        _args: ()
        options: {}

对应的核心代码片段：

# -------------- SOLVING THE SHORT TERM LP PROBLEM -----------------------

model.dual = pyo.Suffix(direction=pyo.Suffix.IMPORT)
opt = SolverFactory('glpk', solver_io="python")
results = opt.solve(model,tee=True)

完整建模代码：

# -*- coding: utf-8 -*-
"""
Created on Wed Jan 29 08:27:38 2020

@author: dqpinel

Renate Berge, 2021
"""

import pyomo.environ as pyo
import numpy as np
from pyomo.environ import ConcreteModel,Set,RangeSet,Param,Suffix,Reals,NonNegativeReals,NonPositiveReals,Binary,Objective,minimize,maximize,value
from pyomo.core import Constraint,Var,Block,ConstraintList
from pyomo.opt import SolverFactory, SolverStatus, TerminationCondition



""" BASE CASE MODEL """

# Creation of a Concrete Model
model = pyo.ConcreteModel()


#------------------- PARAMETERS -------------------------------------

# Reservoir level limits [Mm^3]
Vstart = {'Upper':420, 'Lower':18}
Vmin = {'Upper':0, 'Lower':4}
Vmax = {'Upper':500, 'Lower':20}

# Discharge [m^3/s]
Qmin = {'Upper':2, 'Lower':4}
Qamax = {'Upper':12, 'Lower':12}
Qbmax = {'Upper':4, 'Lower':5}
Qcmax = {'Upper':100, 'Lower':0}

# Production [MW]
Pmin = {'Upper':12, 'Lower':6}
Pmax = {'Upper':200, 'Lower':200}

# Load [MW]
L = {1:60,2:120,3:90,4:80}

# Inflow [m^3/s]
Q = {'Upper':0.5, 'Lower':0.1}

# Constants and conversion factors
beta_a = {'Upper':6, 'Lower':1.5} # MW/m^3/s (conversion factor between production and discharge for generator 1 (Upper) and 2 (Lower), segment A)
beta_b = {'Upper':5, 'Lower':1.3} # MW/m^3/s (conversion factor between production and discharge for generator 1 and 2, segment B)


fm3 = 3600/1000000 # Conversion factor between m^3/s and Mm^3/hour. 1h = 3600s and M = 10^6.
l = 6              # Length of each time step (6 hours)
wv = {'Upper':650000, 'Lower':120000 }   # Water value for reservoir 2 given from the "Seasonal Model" to maintain long-term strategy (NOK/Mm^3)


#----------------------- SETS -------------------------------------

# Set of periods
model.periods = pyo.Set(initialize=[1,2,3,4])
model.periods_ = pyo.Set(initialize=[2,3,4])

# Set of reservoirs
model.reservoirs = pyo.Set(initialize=['Upper','Lower'])


# -------------- VARIABLES AND THEIR BOUNDS ------------------------


# the variables are defined for each reservoir and each period

def qA_bounds(model,i,j):
    return (0,Qamax[i])
model.qA = pyo.Var(model.reservoirs,model.periods, bounds=qA_bounds)

def qB_bounds(model,i,j):
    return (0,Qbmax[i])
model.qB = pyo.Var(model.reservoirs,model.periods, bounds=qB_bounds)

def qC_bounds(model,i,j):
    return (0,Qcmax[i])
model.qC = pyo.Var(model.reservoirs,model.periods, bounds=qC_bounds)

def q_bounds(model,i,j):
    return (Qmin[i],np.inf)
model.q = pyo.Var(model.reservoirs,model.periods, bounds=q_bounds)

def v_bounds(model,i,j):
    return (Vmin[i],Vmax[i])
model.v = pyo.Var(model.reservoirs,model.periods, bounds=v_bounds)

def p_bounds(model,i,j):
    return (Pmin[i],Pmax[i])
model.p = pyo.Var(model.reservoirs,model.periods, bounds=p_bounds)

# ----------------- CONSTRAINTS ------------------------

# Discharge
def disch_rule(model,i,j):
    return model.qA[i,j] + model.qB[i,j] + model.qC[i,j] + Qmin[i] == model.q[i,j]
model.disch_cons = pyo.Constraint(model.reservoirs,model.periods, rule=disch_rule)

# Production
def prod_rule(model,i,j):
    return model.qA[i,j]*beta_a[i] + model.qB[i,j]*beta_b[i] + Pmin[i] == model.p[i,j]
model.prod_cons = pyo.Constraint(model.reservoirs,model.periods, rule=prod_rule)

# Load Balance
def load_rule(model,j):
    return model.p['Upper',j] + model.p['Lower',j] == L[j]
model.load_cons = pyo.Constraint(model.periods, rule=load_rule)

# Reservoir Balance
def res_start_rule(model,i):
    if i=='Upper':
        return model.v[i,1] == Vstart[i] + Q[i]*fm3*l - model.q[i,1]*fm3*l
    elif i == 'Lower':
        return model.v[i,1] == Vstart[i] + Q[i]*fm3*l - model.q[i,1]*fm3*l + model.q['Upper',1]*fm3*l
model.res_start_cons = pyo.Constraint(model.reservoirs, rule=res_start_rule)

def res_rule(model,i,j):
    if i=='Upper':
        return model.v[i,j] == model.v[i,j-1] + Q[i]*fm3*l - model.q[i,j]*fm3*l
    elif i=='Lower':
        return model.v[i,j] == model.v[i,j-1] + Q[i]*fm3*l - model.q[i,j]*fm3*l + model.q['Upper',j]*fm3*l
model.res_cons = pyo.Constraint(model.reservoirs,model.periods_, rule=res_rule)


# -------------------- OBJECTIVE FUNCTION --------------------------------

def ObjRule(model):
    return sum(model.q[i,j]*wv[i]*fm3*l for i in model.reservoirs for j in model.periods)
model.obj= pyo.Objective(rule=ObjRule, sense=pyo.minimize)


# -------------- SOLVING THE SHORT TERM LP PROBLEM -----------------------

model.dual = pyo.Suffix(direction=pyo.Suffix.IMPORT)
opt = SolverFactory('glpk', solver_io="python")
results = opt.solve(model,tee=True)

print('\n')

for v in model.component_data_objects(pyo.Var):
  print('%s:   %s'%(str(v), v.value))

model.display()
model.dual.display()

错误原因 #

• 指定的solver_io="python"参数无效，GLPK不支持该IO类型，正确可选值为'lp'或'glpk'
• 若尝试使用GLPK的Python绑定，可能未正确安装pyomo.glpk模块，导致Pyomo无法找到有效接口

解决方案 #

1. 移除错误的solver_io参数 #

修改SolverFactory初始化代码，让Pyomo自动适配正确的IO接口：

opt = SolverFactory('glpk')

2. 确保GLPK及Pyomo接口安装完整 #

• 使用conda安装（推荐）：

  conda install -c conda-forge glpk pyomo.glpk
  

• 使用pip安装：

  pip install pyomo[glpk]
  

3. 验证GLPK系统路径 #

在终端执行以下命令，确认GLPK已添加到系统PATH：

glpsol --version

若命令无响应，需手动将GLPK的安装目录（通常包含glpsol.exe）添加到系统环境变量PATH中。

内容的提问来源于stack exchange，提问作者Fiskemann