cplex - Cplex:无法获取决策变量的值
问题描述
我正在解决 Cplex 的医院工作人员调度问题,而且我是 Cplex 的新手。
但 Cplex 无法配置决策变量的答案。
我想可能是关于我的模型设计问题。
这是一个长模型。
非常感谢我能得到帮助。
++型号
- 套:
{字符串} E=...; // 一组经验级别(高级、新秀)
{字符串} 我=...; // 所有医生的集合
{字符串} 是=...; //一组经验水平为S的医生
{string} Ir=...;//经验等级R的医师集合
{字符串} K=...; //班次类型 (o & t)
{字符串} A1=...; //一组班次(白天,晚上,晚上)
{字符串} A2=...; //一组班次t工作(白天,晚上,晚上1,晚上2)
- 参数
诠释 D=...;
range Day=1..D;//计划范围的范围
int Co[Is]=...; //具有高级经验水平的医师 i 的雇用成本
诠释 Ct[Ir]=...; //新人经验级别的医生 i 的雇佣成本
诠释 Pku=...; //最大限度。医生轮班的天数
诠释PKD = ...; //分钟。医生轮班的天数
int Pkuh=...;//最大。医生轮班的天数
int Pkdh=...;//分钟。医生轮班的天数
诠释 s=...; //最大限度。连续工作日
诠释dosd = ...; //轮班“高级”医师的日常工作需求
诠释剂量=...; //轮班“高级”医师的晚间工作需求
int dosn=...; //“高级”医师的夜班需求
诠释多德=...;//轮班“菜鸟”医师的日常工作需求
诠释多尔=...; //轮班“菜鸟”医师的晚间工作需求
诠释多恩=...;//“菜鸟”医师的夜班需求
诠释dtsd = ...; //“高级”医师轮班 t 天的工作需求
诠释 dtse=...; //“高级”医师的晚班工作需求
诠释 dtsn1=...; //轮班t night1“高级”医师的工作需求
诠释 dtsn2=...; //值班t night2“高级”医师的工作需求
诠释dtrd = ...; //“菜鸟”医师轮班 t 天工作需求
诠释dtre = ...; //“菜鸟”医师的夜班工作需求
诠释 dtrn1=...; //“菜鸟”医师轮班 t night1 工作需求
诠释dtrn2 = ...; //“菜鸟”医师轮班t night2工作需求
--二进制决策变量
dvar int x1[I][Day][A1] in 0..1; //1、if Senior+ shift o+ 其中一项工作
dvar int x2[I][Day][A2] in 0..1; //1、如果senior+ shift t+其中一项工作
dvar int y1[I][Day][A1] in 0..1; //1、如果菜鸟+ shift o+ 之一的工作
dvar int y2[I][Day][A2] in 0..1; //1、如果rookie+ shift t+ 之一的工作
--决策变量的表达
dexpr float cost =sum(i in Is, d in Day, a in A1)Co[i]*x1[i][d][a]
+ sum(i in Is, d in Day, a in A2)Co[i]*x2[i][d][a]
+ sum(i in Ir, d in Day, a in A1)Ct[i]*y1[i][d][a]
+ sum(i in Ir, d in Day, a in A2)Ct[i]*y2[i][d][a] ;
--目标函数
最小化成本;
--约束
subject to
{
forall(d in Day, Day in A1)
{
sum(i in Is) x1[i, d ,"Day"] == dosd; //Daily demand:Senior+shift o+day work
}
forall(d in Day, Evening in A1)
{
sum(i in Is ) x1[i, d, "Evening"] == dose; //Daily demand:Senior+shift o+evening work
}
forall(d in Day, Night in A1)
{
sum(i in Is ) x1[i, d, "Night"] == dosn; //Daily demand:Senior+shift o+night work
}
forall(d in Day, Day in A2)
{
sum(i in Is) x2[i, d, "Day"] == dtsd; //Daily demand:Senior+ shift t +day work
}
forall(d in Day, Evening in A2)
{
sum(i in Is) x2[i, d, "Evening"] == dtse; //Daily demand:Senior+shift t+evening work
}
forall(d in Day, Night1 in A2)
{
sum(i in Is) x2[i, d, "Night1"] == dtsn1; //Daily demand:Senior+shift t+night1 work
}
forall(d in Day, Night2 in A2)
{
sum(i in Is) x2[i, d, "Night2"] == dtsn2; //Daily demand:Senior+shift t+night2 work
}
forall(d in Day, Day in A1)
{
sum(i in Ir) y1[i, d, "Day"] == dord; //Daily demand:Rookie+shift o+day work
}
forall(d in Day, Evening in A1)
{
sum(i in Ir) y1[i, d, "Evening"] == dore; //Daily demand:Rookie+shift o+evening work
}
forall(d in Day, Night in A1)
{
sum(i in Ir) y1[i, d, "Night"] == dorn; //Daily demand:Rookie+shift o+night work
}
forall(d in Day, Day in A2)
{
sum(i in Ir) y2[i, d, "Day"] == dtrd; //Daily demand:Rookie+shift t+day work
}
forall(d in Day, Evening in A2)
{
sum(i in Ir) y2[i, d, "Evening"] == dtre; //Daily demand:Rookie+shift t+evening work
}
forall(d in Day, Night1 in A2)
{
sum(i in Ir) y2[i, d, "Night1"] == dtrn1; //Daily demand:Rookier+shift t+night1 work
}
forall(d in Day, Night2 in A2)
{
sum(i in Ir) y2[i, d, "Night2"] == dtrn2; //Daily demand:Rookie+shift t+night2 work
}
forall(i in Is, d in Day:d<D, Evening in A1)
{
x1[i, d, "Evening"]+ x1[i,(d+1), "Day"]<=1; //(Senior)For shift type o,if previous day is evening
work, can't have day work the following day'
}
forall(i in Is, d in Day:d<D, Evening in A2)
{
x2[i, d, "Evening"]+ x2[i,(d+1), "Day"]<=1; //(Senior)For shift type t,if previous day is evening
work, can't have day work the following day'
}
forall(i in Ir, d in Day:d<D, Evening in A1)
{
y1[i, d, "Evening"]+ y1[i,(d+1), "Day"]<=1; //(Rookie)For shift type o,if previous day is evening
work, can't have day work the following day'
}
forall(i in Ir, d in Day:d<D, Evening in A2)
{
y2[i, d, "Evening"]+ y2[i,(d+1), "Day"]<=1; //(Rookie)For shift type t,if previous day is evening
work, can't have day work the following day'
}
forall(i in Is, d in Day:d<D, Night in A1)
{
x1[i, d,"Night"]+ x1[i,(d+1),"Day"]<=1; //(Senior)For shift type o,if previous day is night work,
can't have day work the following day'
}
forall(i in Is, d in Day:d<D, Night1 in A2)
{
x2[i, d,"Night1"]+ x2[i,(d+1),"Day"]<=1; //(Senior)For shift type t,if previous day is night1 work,
can't have day work the following day'
}
forall(i in Is, d in Day:d<D, Night2 in A2)
{
x2[i, d,"Night2"]+ x2[i,(d+1),"Day"]<=1; //(Senior)For shift type t,if previous day is night2 work,
can't have day work the following day'
}
forall(i in Ir, d in Day:d<D, Night in A1)
{
y1[i, d,"Night"]+ y1[i,(d+1),"Day"]<=1; //(Rookie)For shift type o,if previous day is night work,
can't have day work the following day'
}
forall(i in Ir, d in Day:d<D, Night1 in A2)
{
y2[i, d,"Night1"]+ y2[i,(d+1),"Day"]<=1; //(Rookie)For shift type t,if previous day is night1 work,
can't have day work the following day'
}
forall(i in Ir, d in Day:d<D, Night2 in A2)
{
y2[i, d,"Night2"]+ y2[i,(d+1),"Day"]<=1; //(Rookie)For shift type t,if previous day is night2 work,
can't have day work the following day'
}
forall (i in Is, d in Day)
{
sum(a in A1)x1[i,d,a] + sum(a in A2)x2[i,d,a] <=1; //(Senior)each day can only have one work with
one shift
}
forall (i in Ir, d in Day)
{
sum(a in A1)y1[i,d,a] + sum(a in A2)y2[i,d,a] <=1; //(Rookie)each day can only have one work with
one shift
}
forall(i in Is, a in A1)
{
sum(d in Day)x1[i,d,a] <=Pku; //(Senior) can’t have shift type o that more than upper bound. (in
days)
}
forall(i in Ir, a in A1)
{
sum(d in Day)y1[i,d,a] <=Pku; //(Rookie) can’t have shift type o that more than upper bound. (in
days)
}
forall(i in Is, a in A1)
{
sum(d in Day)x1[i,d,a]>=Pkd; //(Senior) can’t have shift type o that less than lower bound. (in
days)
}
forall(i in Ir, a in A1)
{
sum(d in Day)y1[i,d,a]>=Pkd; //(Rookie) can’t have shift type o that less than lower bound. (in
days)
}
forall(i in Is, a in A2)
{
sum(d in Day)x2[i,d,a]<=Pkuh; //(Senior) can’t have shift type t that more than upper bound. (in
days)
}
forall(i in Ir, a in A2)
{
sum(d in Day)y2[i,d,a]<=Pkuh; //(Rookie) can’t have shift type t that more than upper bound. (in
days)
}
forall(i in Is, a in A2)
{
sum(d in Day)x2[i,d,a]>=Pkdh; //(Senior) can’t have shift type t that less than lower bound. (in
days)
}
forall(i in Ir, a in A2)
{
sum(d in Day)y2[i,d,a]>=Pkdh; //(Rookie) can’t have shift type t that less than lower bound. (in
days)
}
forall(i in Is)
{
sum(d in Day:d<=(D-s), a in A1)x1[i,d,a]+sum(d in Day:d<=(D-s), a in A2)x2[i,d,a]<=s; //(Senior)
can’t have consecutive working days (regardless of shift type) that more than upper bound. (in days)
}
forall(i in Ir)
{
sum(d in Day:d<=(D-s), a in A1)y1[i,d,a]+sum(d in Day:d<=(D-s), a in A2)y2[i,d,a]<=s; //(Rookie)
can’t have consecutive working days (regardless of shift type) that more than upper bound. (in days)
}
++数据
E={"高级","菜鸟"};
我={"A","B","C","D","E","F","G","H","I","J","K","L" ,"M","N","O","P","Q","R","S","T","W","X","Y","Z"};
是={"A","B","C","D","E","F","G","H"};
Ir={"I","J","K","L","M","N","O","P","Q","R","S","T" ,"W","X","Y","Z"};
K={"o","t"};
A1={"白天","晚上","夜晚"};
A2={"白天","晚上","夜晚 1","夜晚 2"};
D=7;
钴=[1200,1150,1100,1050,1000,950,900,850];
Ct=[800,750,750,750,700,700,700,700,700,700,650,650,650,650,650,650];
Pku=10;
PKd=5;
Pkuh=8;
PKDH=4;
s=5;
解决方案
您的模型可能不可行。如果您标记约束,那么 CPLEX 会给您一些放松和冲突,这将帮助您理解原因。我的建议是标记所有约束。
例如,您可以更改
forall(d in Day, Day in A1)
{
sum(i in Is) x1[i, d ,"Day"] == dosd; //Daily demand:Senior+shift o+day work
}
进入
forall(d in Day, Day in A1)
ct1:sum(i in Is) x1[i, d ,"Day"] == dosd; //Daily demand:Senior+shift o+day work
推荐阅读
- slack - 是否可以为 Slack 定制 ServiceNow?
- java - 多对多关系的条件查询
- java - 当我点击验证码然后点击登录它不起作用
- angular - Highcharts 将左列添加到分组数据
- python-requests - Python 调用 PUT 请求:不支持“PUT”
- android - Flutter:ListView 在键盘弹出时处理小部件
- django - Django3多个文件提交到数据库
- arm - 在 ARM64 v8 上找不到真正的函数地址
- java - Mybatis 自定义类型处理程序被调用两次
- android - 房间数据库在 mac book pro m1 中不起作用