$ontext Column generation for work force scheduling Erwin Kalvelagen erwin@amsterdamoptimization.com References: Annemieke van Dongen, "Personeelsplanning en kolomgeneratie", Vrije Universiteit Amsterdam, 2005 $offtext set t 'periods' /t1*t48/; parameter req(t) 'required number of employees' / t1 2, t2 5, t3 6, t4 5, t5 7, t6 7, t7 4, t8 6, t9 7, t10 7 t11 7, t12 7, t13 6, t14 5, t15 6, t16 6, t17 5, t18 4, t19 7, t20 6 t21 6, t22 4, t23 5, t24 5, t25 6, t26 6, t27 5, t28 7, t29 7, t30 6 t31 9, t32 6, t33 7, t34 6, t35 6, t36 7, t37 7, t38 4, t39 5, t40 5 t41 4, t42 5, t43 3, t44 4, t45 5, t46 3, t47 5, t48 4 /; set slen 'shift lengths' /short, medium, long/ ; table shiftlen(slen,*) 'lengths of shifts in 0.25 hours' minwork maxwork sumbreak maxnumbreak short 16 22 1 1 medium 23 32 2 2 long 33 36 3 2 ; scalar maxwlen 'max periods of contiguous work' /13/; *----------------------------------------------------- * Gilmore-Gomory column generation algorithm *----------------------------------------------------- sets s 'possible shifts' /shift1*shift1000/ iter 'maximum iterations' /iter1*iter1000/ ; *----------------------------------------------------- * Master model *----------------------------------------------------- parameters a(t,s) 'matrix growing in dimension s' c(s) 'cost vector growing in dimension s' ; variables x(s) 'shifts used' z 'objective variable' ; integer variable x; x.up(s) = sum(t, req(t)); set ds(s) 'dynamic subset'; equations masterobj 'cost: total number of periods worked' edemand(t) 'meet required manpower' ; masterobj.. z =e= sum(ds,c(ds)*x(ds)); edemand(t).. sum(ds, a(t,ds)*x(ds)) =g= req(t); model master /masterobj,edemand/; * reduce amount of information written to the listing file master.solprint = 2; master.limrow = 0; master.limcol = 0; master.solvelink = 2; *----------------------------------------------------- * Sub model *----------------------------------------------------- variables sigma 'reduced cost' work(t) 'new shift: quarters worked' shift(t) 'operator is present' break(t) 'operator is on break' totwork 'number of periods worked' totbreak 'break time' startshift(t) 'shift starts' startbreak(t) 'break starts' tshift(slen) 'type of shift: short, medium, long' ; binary variables work, present, break, startshift, startbreak, tshift; positive variables totwork, totbreak, totpresent; * relax: these variables are automatically integer break.prior(t) = INF; startbreak.prior(t) = INF; startshift.prior(t) = INF; equations subobj 'objective function' eshift(t) 'shift = work or break' etotwork 'calculate total quarter hours worked' etotbreak 'calculate total quarter hours used for breaks' estshift(t) 'start of shift' estbreak(t) 'start of break' estartsh1 'start shift once' etshift1 'limit work for short and long shifts' etshift2 'limit work for short and long shifts' etshift3 'limit breaks for short and long shifts' ebreak1 'required break time' ebreak2 'limits on number of breaks (forces longer breaks)' emaxcont(t) 'max contiguous working time' ; subobj.. sigma =e= sum(t, (1-edemand.m(t))*work(t)); eshift(t).. shift(t) =e= work(t) + break(t); etotwork.. totwork =e= sum(t, work(t)); etotbreak.. totbreak =e= sum(t, break(t)); estshift(t).. startshift(t) =g= shift(t) - shift(t-1); estbreak(t).. startbreak(t) =g= break(t) - break(t-1); estartsh1.. sum(t, startshift(t)) =e= 1; etshift1.. totwork =g= sum(slen, shiftlen(slen,'minwork')*tshift(slen)); etshift2.. totwork =l= sum(slen, shiftlen(slen,'maxwork')*tshift(slen)); etshift3.. sum(slen, tshift(slen)) =e= 1; ebreak1.. totbreak =e= sum(slen, shiftlen(slen,'sumbreak')*tshift(slen)); ebreak2.. sum(t, startbreak(t)) =l= sum(slen, shiftlen(slen,'maxnumbreak')*tshift(slen)); alias (t,tt); set tmaxw(t,tt); tmaxw(t,tt)$(ord(tt)>=ord(t) and ord(tt)<=ord(t)+maxwlen and ord(t)<=card(t)-maxwlen) = yes; display tmaxw; emaxcont(t)$sum(tmaxw(t,tt),1).. sum(tmaxw(t,tt), work(tt)) =l= maxwlen; model sub / subobj, eshift, etotwork, etotbreak, estshift, estbreak, estartsh1, etshift1, etshift2, etshift3, ebreak1, ebreak2, emaxcont /; equation edummy; variable dummy; edummy.. dummy =e= 0; model subcheck / edummy, eshift, etotwork, etotbreak, estshift, estbreak, estartsh1, etshift1, etshift2, etshift3, ebreak1, ebreak2, emaxcont /; * reduce amount of information written to the listing file sub.solprint = 2; sub.limrow = 0; sub.limcol = 0; sub.solvelink = 2; sub.optcr=0; subcheck.solprint = 2; subcheck.limrow = 0; subcheck.limcol = 0; subcheck.solvelink = 2; *----------------------------------------------------- * initial feasible solution * invent some candidate shifts *----------------------------------------------------- table init(t,s) '1=work,2=break' shift1 t1 1 t2 1 t3 1 t4 1 t5 1 t6 1 t7 1 t8 1 t9 1 t10 1 t11 1 t12 2 t13 1 t14 1 t15 1 t16 1 t17 1 t18 1 t19 1 t20 1 t21 1 t22 1 t23 1 ; set s2(s) /shift2*shift26/; loop(s2(s), init(t,s) = init(t-1,s-1); ); display init; * * check initial solution. * loop(s$sum(t$init(t,s),1), shift.fx(t)$init(t,s) = 1; work.fx(t)$(init(t,s)=1) = 1; break.fx(t)$(init(t,s)=2) = 1; solve subcheck minimizing dummy using mip; abort$(subcheck.modelstat<>1 and subcheck.modelstat<>8) "Initial solution not correct"; shift.lo(t) = 0; shift.up(t) = 1; work.lo(t) = 0; work.up(t) = 1; break.lo(t) = 0; break.up(t) = 1; ); ds(s)$sum(t$init(t,s),1) = 1; a(t,ds)$(init(t,ds)=1) = 1; c(ds) = sum(t$(init(t,ds)=1),1); set nexts(s); nexts(s)$(ord(s)=card(ds)+1) = yes; option a:0; option c:0; display "------------------------------------------------" "initial solution", "------------------------------------------------", a,c,ds,nexts; *----------------------------------------------------- * column generation algorithm starts here *----------------------------------------------------- scalar done /0/; scalar iteration; loop(iter$(not done), * * solve master problem * solve master using rmip minimizing z; * * solve knapsack problem * solve sub using mip minimizing sigma; * * new column found? * if(sigma.l < -0.001, a(t,nexts) = work.l(t); c(nexts) = sum(t, work.l(t)); ds(nexts) = yes; iteration = ord(iter); display "--------------------------------------------------------------", iteration, "--------------------------------------------------------------", ds,a; nexts(s) = nexts(s-1); else done = 1; ); ); abort$(not done) "Too many iterations."; * * solve final mip * display "--------------------------------------------------------------", "Final MIP", "--------------------------------------------------------------"; master.optcr=0; solve master using mip minimizing z; display z.l,x.l; *-------------------------------------------------------------- * reporting *-------------------------------------------------------------- scalars totreq 'total required periods' totusd 'total allocated periods' ; totreq = sum(t, req(t)); totusd = z.l; display totreq, totusd; alias(s,s3); set curs(s3) /shift1/; scalar k; parameter schedule(s,t); loop(s$(x.l(s)>0.5), for(k=1 to round(x.l(s)), schedule(curs,t) = a(t,s); curs(s3) = curs(s3-1); ); ); option schedule:0; display schedule; |
Can you please explain this in AMPL
ReplyDeleteUnfortunately I charge a fee for those type of jobs.
ReplyDeleteI would like to know how much it would cost for getting to know the above code in AMPL.
DeleteContact me by email.
Delete