1. Introduction
Jan 26, 2016
1. Introduction
Introduction
• Modeling
Use a set of mathematical relations to represent mathematically a real life situation
Compromise between a closer image of the reality and the difficulty of solving the model
Mathematical model
• Three items to identify
i) The set of actions (activities) of the decision maker (variables)
ii) The objective of the problem specified in terms of a mathematical fonction (objective fonction)
iii) The context of the problem specified in terms of mathematical relations (contraint functions)
Solving the problem
• Three items to identify
i) The set of actions (activities) of the decision maker (variables)
ii) The objective of the problem specified in terms of a mathematical fonction (objective fonction)
iii) The context of the problem specified in terms of mathematical relations (contraint functions)
• Solving method
Use a procedure (algorithm) to determine
the values of the variables indicating how the different activities are used
to optimise the objective fonction (to reach the objective)
and satisfying the contraints
Linear model
Two specific properties
1. Additivity of the values of the variables: the total effect of any set of actions (variables) is equal to the sum of the individual effect of each action (variable) in the set.
There is no cross action of the variables
2. The variables are always non negative
Exemple 1: diet problem
• 3 types of grains are available to feed an herb: g1, g2, g3
• Each kg of grain includes 4 nutritional elements: ENA, ENB, ENC, END
• The weekly quantity required for each nutritional element is specified
• The price per kg of each grain is also specified.
• Problem: Determine the quantity (in kg) of each grain to specify the minimum cost diet for the herb satisfying the nutritional requirements of the diet
Problem data
• 3 types of grains are available to feed the herb: g1, g2, g3
• Each kg of grain includes 4 nutritional elements: ENA, ENB, ENC, END
• The weekly quantity required for each nutritional element is specified
• The price per kg of each grain is also specified.
• Problem: Determine the quantity (in kg) of each grain to specify the minimum cost diet for the herb satisfying the nutritional requirements of the diet
quantité g1 g2 g3 hebd.
ENA 2 3 7 1250 ENB 1 1 0 250ENC 5 3 0 900END 0.6 0.25 1 232.5
$/kg 41 35 96
Problem data
• 3 types of grains are available to feed the herb: g1, g2, g3
• Each kg of grain includes 4 nutritional elements: ENA, ENB, ENC, END
• The weekly quantity required for each nutritional element is specified
• The price per kg of each grain is also specified.
• Problem: Determine the quantity (in kg) of each grain to specify the minimum cost diet for the herb satisfying the nutritional requirements of the diet
weekly g1 g2 g3 quantity
ENA 2 3 7 1250 ENB 1 1 0 250ENC 5 3 0 900END 0.6 0.25 1 232.5
$/kg 41 35 96
Problem variables
• 3 types of grains are available to feed the herb: g1, g2, g3
• Each kg of grain includes 4 nutritional elements: ENA, ENB, ENC, END
• The weekly quantity required for each nutritional element is specified
• The price per kg of each grain is also specified.
• Problem: Determine the quantity (in kg) of each grain to specify the minimum cost diet for the herb satisfying the nutritional requirements of the diet
i) Activities or actions of the model Actions variables
# kg de g1 x1
# kg de g2 x2
# kg de g3 x3
Objective function and constraints
ii) Objective function Weekly cost of the diet = 41x1 + 35x2 + 96x3 to minimise
iii) Contraints
ENA: 2x1 + 3x2 +7x3 ≥ 1250ENB: 1x1 + 1x2 ≥ 250ENC: 5x1 + 3x2 ≥ 900END: 0.6x1 + 0.25x2 + x3 ≥ 232.5
Non negativity constraints: x1 ≥ 0, x2 ≥ 0, x3 ≥ 0
weekly g1 g2 g3 quantity
ENA 2 3 7 1250 ENB 1 1 0 250ENC 5 3 0 900END 0.6 0.25 1 232.5
$/kg 41 35 96
Mathematical model
ii) Objective function Weekly cost of the diet = 41x1 + 35x2 + 96x3 to minimize
iii) Contraints
ENA: 2x1 + 3x2 +7x3 ≥ 1250ENB: 1x1 + 1x2 ≥ 250ENC: 5x1 + 3x2 ≥ 900END: 0.6x1 + 0.25x2 + x3 ≥ 232.5Non negativity constraints: x1 ≥ 0, x2 ≥ 0, x3 ≥ 0
min z = 41x1 + 35x2 + 96x3 s.t. 2x1 + 3x2 +7x3 ≥ 1250 1x1 + 1x2 ≥ 250 5x1 + 3x2 ≥ 900 0.6x1 + 0.25x2 + x3 ≥ 232.5 x1 ≥ 0, x2 ≥ 0, x3 ≥ 0
Restaurant owner problem
• Seafoods available: 30 sea-urchins 24 shrimps 18 oysters
• Two types of seafood plates to be offered: $8 : including 5 sea-urchins, 2 shrimps et 1 oyster $6 : including 3 sea-urchins, 3 shrimps et 3 oysters
• Problem: determine the number of each type of plates to be offered by the owner in order to maximize his revenue according to the seafoods available
Problem variables
• Seafoods available: 30 sea-urchins 24 shrimps 18 oysters
• Two types of seafood plates to be offered:
$8 : including 5 sea-urchins, 2 shrimps et 1 oyster
$6 : including 3 sea-urchins, 3 shrimps et 3 oysters
• Problem: determine the number of each type of plates to be offered by the owner in order to maximize his revenue according to the seafoods
available
i) Activities or actions Actions variables # plates $8 x # plates $6 y
Objective function and contraints
• Seafoods available: 30 sea-urchins 24 shrimps 18 oysters
• Two types of seafood plates to be offered:
$8 : including 5 sea-urchins, 2 shrimps et 1 oyster
$6 : including 3 sea-urchins, 3 shrimps et 3 oysters
• Problem: determine the number of each type of plates to be offered by the owner in order to maximize his revenue according to the seafoods
available
i) Activities or actionsii) Objective function owner’s revenue = 8x + 6y to maximise iii) Contraints
sea-urchins: 5x + 3y ≤ 30 shrimbs: 2x + 3y ≤ 24 oysters: 1x + 3y ≤ 18 Non negative variables: x,y ≥ 0
Mathematical model
i) Activities or actionsii) Objective function owner’s revenue = 8x + 6y to maximise iii) Contraints
sea-urchins: 5x + 3y ≤ 30 shrimbs: 2x + 3y ≤ 24 oysters: 1x + 3y ≤ 18 Non negative variables: x,y ≥ 0
max 8x + 6y s.t. 5x + 3y ≤ 30 2x + 3y ≤ 24 1x + 3y ≤ 18 x,y ≥ 0
Maximal Open Pit problem: to determine the maximal gain expected from the extraction
the net value of extracting block i
ib
otherwise. 0
extracted is block if 1 ixi
objective function .iNi
i xb
Notation: : if 0 (including ore)
waste bloore blo
c k: if ck
0 i
i
bb
Maximal pit slope constraintsto identify the set Bi of predecessor blocks that have to be removed before block i
Maximal pit slope constraintsto identify the set Bi of predecessor blocks that have to be removed before block i
Maximal pit slope constraintsto identify the set Bi of predecessor blocks that have to be removed before block i
The maximal pit slope constraints:
0 , j i ix x j B i N
(MOP) Max
Subject to 0 , (1)
0 or 1 . (2)
i ii N
j i i
i
b x
x x j B i N
x i N
Maximal pit slope constraintsto identify the set Bi of predecessor blocks that have to be removed before block i