裝箱問題(Bin Packing Problem)
裝箱問題即搬家公司問題。一個搬家公司有無限多的箱子,每個箱子的承重上限為W,當搬家公司進入一個房間時,所有物品都必須被裝入箱子,每個物品的重量為wi (i=1,...,m),規划裝箱方式,使得使用的箱子最少。此文及所有本博客中的博文均為原創,本博客不轉發他人博文,特此聲明。
實例
一個海運公司有若干貨輪, 一個貨輪的最大載重量4000噸, 客戶貨物的重量是 1020T, 1930T, 3575T, 2861T, 4221T, 1541T, 2348T, and 1170T, 問如何分配貨物可以使總計需求的貨輪數量(航次) 最小。
建模
假設搬家公司帶來n個箱子,且n個箱子足夠裝入所有物品。設0-1變量x[i][j]表示第j個物品是否被安排裝入第i個箱子,0表示不裝入,1表示裝入。根據題意,任何物品必須被裝入某個箱子中,於是有約束:
sum{i=1,...,n} x[i][j] = 1 | j=1,...,m // (1)
如果箱子i有任何物品被裝入,則說該箱子被打開,並設0-1變量y[i]表示箱子i是否被打開(0-表示不打開,1-表示打開)。顯然目標是極小化打開箱子的數目,即:
min sum{i=1,...,n} y[i] //(2)
裝入箱子的物品重量和不能超過該箱子的承重,即:
sum{j=1,...,m} x[i][j] <= W*y[i] | i=1,...,n //(3)
上式表示當聚焦第i個箱子時,如果y[i]=0則任何x[i][j]都必須為0,亦即如果第i個箱子沒有被打開,則沒有物品可以裝入該箱子。反之,如果y[i]=1,則裝入該箱子的物品的重量和必須小於箱子的最大承重W。
綜合(1)-(3), 裝箱問題模型的核心部分如下:
//-------------------------------------------------------
min sum{i=1,...,n} y[i] //(2)
subject to
sum{i=1,...,n} x[i][j] = 1 | j=1,...,m // (1)
sum{j=1,...,m} x[i][j]w[i] <= W*y[i] | i=1,...,n //(3)
//-------------------------------------------------------
添加where段對模型中常量符號和變量符號的說明
//-------------------------------------------------------
where
m,n are integers
W is a number
w is a set
x[i][j] is a variable of binary|i=1,...,n;j=1,...,m
y[i] is a variable of binary|i=1,...,n
//-------------------------------------------------------
E、添加數據段
//-------------------------------------------------------
data
W=4000
w={1020, 1930,3575,2861,4221,1541,2348, 1170}
data_relation
m=_$(w) // <-- _\$(w) 函數給出集合w中的元素數。
n=m/2 // <-- 預備箱子數取為物體數的一半。
//-------------------------------------------------------
上面模型中,物品個數由求w中的元素數給出。預備箱子數給為物體數的一半。預備箱子數必須大於實際最優箱子數,否則問題無解。
CPLEX求解
求解模型,在Leapms環境中, 首先使用load命令調入並解析模型, 而后使用"cplex" 命令調用IBMC PLEX求解器完成求解.
如果你的leapms版本不支持cplex命令,可用savemps或者savelp保存成mps或lp模型,然后再用cplex求解(https://www.cnblogs.com/leapms/p/11846039.html)。
Leapms>load Current directory is "ROOT". ......... binpacking.leap ......... lease input the filename:binpacking =============================================================== : //------------------------------------------------------- : : min sum{i=1,...,n} y[i] //(2) : subject to : sum{i=1,...,n} x[i][j] = 1 | j=1,...,m // (1) : sum{j=1,...,m} x[i][j]w[i] <= W*y[i] | i=1,...,n //(3) : where : m,n are integers : W is a number 0: w is a set 1: x[i][j] is a variable of binary|i=1,...,n;j=1,...,m 2: y[i] is a variable of binary|i=1,...,n 3: data 4: W=4000 5: w={1020M, 1930M, 3575M, 2861M, 4221M, 1541M, 2348M, 1170} 6: //w={1020, 1930,3575,2861,4221,1541,2348, 1170} 7: data_relation 8: m=_$(w) 9: n=m 0: //------------------------------------------------------- =============================================================== >end of the file. arsing model: D R V O C S End. ................................. umber of variables=72 umber of constraints=16 ................................. Leapms>cplex You must have licience for Ilo Cplex, otherwise you will violate corresponding copyrights, continue(Y/N)? 你必須有Ilo Cplex軟件的授權才能使用此功能,否則會侵犯相應版權, 是否繼續(Y/N)?y Tried aggregator 1 time. MIP Presolve eliminated 1 rows and 9 columns. MIP Presolve modified 61 coefficients. Reduced MIP has 15 rows, 63 columns, and 119 nonzeros. Reduced MIP has 63 binaries, 0 generals, 0 SOSs, and 0 indicators. Presolve time = 0.02 sec. (0.14 ticks) Found incumbent of value 7.000000 after 0.08 sec. (0.32 ticks) Probing time = 0.00 sec. (0.06 ticks) Tried aggregator 1 time. Reduced MIP has 15 rows, 63 columns, and 119 nonzeros. Reduced MIP has 63 binaries, 0 generals, 0 SOSs, and 0 indicators. Presolve time = 0.02 sec. (0.08 ticks) Probing time = 0.00 sec. (0.06 ticks) Clique table members: 43. MIP emphasis: balance optimality and feasibility. MIP search method: dynamic search. Parallel mode: deterministic, using up to 4 threads. Root relaxation solution time = 0.00 sec. (0.05 ticks) Nodes Cuts/ Node Left Objective IInf Best Integer Best Bound ItCnt Gap * 0+ 0 7.0000 0.0000 100.00% * 0+ 0 4.0000 0.0000 100.00% 0 0 2.2824 5 4.0000 2.2824 8 42.94% * 0+ 0 3.0000 2.2824 23.92% 0 0 cutoff 3.0000 2.2824 8 23.92% Elapsed time = 0.20 sec. (0.72 ticks, tree = 0.00 MB, solutions = 3) Root node processing (before b&c): Real time = 0.22 sec. (0.72 ticks) Parallel b&c, 4 threads: Real time = 0.00 sec. (0.00 ticks) Sync time (average) = 0.00 sec. Wait time (average) = 0.00 sec. ------------ Total (root+branch&cut) = 0.22 sec. (0.72 ticks) Solution status = Optimal Solution value = 3 x1_1=1 x1_5=1 x1_8=1 x6_2=1 x6_7=1 x8_3=1 x8_4=1 x8_6=1 y1=1 y6=1 y8=1
求解結果為:文件分配方案是:第一航次運送1、5、8貨物;第二航次2、7貨物;第三航次3、4、6貨物。總計使用三個航次或三艘貨輪。