Multiple constrained optimization BUSINESS MATHEMATICS 1
CONTENTS More than two variables More than one constraint Lagrange method The Lagrange multiplier Old exam question Further study 2
MORE THAN TWO VARIABLES Recall the constrained optimization problem: α max π π¦, π§ s.t. π π¦, π§ = π βͺ function to be maximized depends on π¦ and π§ But sometimes there are more variables βͺ π π¦, π§, π¨ or π π¦ 1 , π¦ 2 , π¦ 3 3
MORE THAN TWO VARIABLES Problem formulation: α max π π¦, π§, π¨ s.t. π π¦, π§, π¨ = π Constrained optimization problem with more than two variables In vector notation: α max π π² s.t. π π² = π 4
EXERCISE 1 Write the following in standard notation: constraint π¦ = 2π§π¨ , objective function π¦π§ 2 + π¨ , purpose is maximization. 5
MORE THAN ONE CONSTRAINT Again recall the constrained optimization problem: α max π π¦, π§, π¨ s.t. π π¦, π§, π¨ = π βͺ constraint is π π¦, π§, π¨ = π But sometimes there are more constraints βͺ β π¦, π§, π¨ = π or π 2 π¦, π§, π¨ = π 2 7
MORE THAN ONE CONSTRAINT Problem formulation: max π π¦, π§, π¨ s.t. π 1 π¦, π§, π¨ = π 1 ΰ΅ and π 2 π¦, π§, π¨ = π 2 Constrained optimization problem with more than one constraint In vector notation: α max π π² s.t. π‘ π² = π 8
EXERCISE 2 Write the following in vector notation: constraints π¦ = 2π§π¨ and π¦ 2 β 5 = π¨ , objective function π¦π§ 2 + π¨ , purpose is maximization. 9
LAGRANGE METHOD How to solve the problem with three variables? α max π π¦, π§, π¨ s.t. π π¦, π§, π¨ = π Generalized trick: βͺ introduce Lagrange multiplier π βͺ define Lagrangian β π¦, π§, π¨, π β π¦, π§, π¨, π = π π¦, π§, π¨ β π π π¦, π§, π¨ β π Find the stationary points βͺ by solving four equations: πβ πβ πβ πβ ππ¦ = 0 , ππ§ = 0 , ππ¨ = 0 , and βͺ ππ = 0 11
LAGRANGE METHOD How to solve the with two constraints? max π π¦, π§, π¨ s.t. ΰ΅ π π¦, π§, π¨ = π and β π¦, π§, π¨ = π Another generalized trick: βͺ introduce two Lagrange multipliers π and π βͺ define Lagrangian β π¦, π§, π¨, π, π β π¦, π§, π¨, π, π = π π¦, π§, π¨ β π π π¦, π§, π¨ β π β π β π¦, π§, π¨ β π Find the stationary points βͺ by solving five equations: πβ πβ πβ πβ πβ ππ¦ = 0 , ππ§ = 0 , ππ¨ = 0 , ππ = 0 , and βͺ ππ = 0 12
LAGRANGE METHOD General constraint optimization problem α max π π² s.t. π‘ π² = π General solution with Lagrangian function β π², π = π π² β π β π‘ π² β π Stationary points of the Lagrangian function are candidate solutions to the optimization problem πβ π²,π πβ π²,π βͺ by solving many equations: = 0 and = 0 ππ¦ π ππ π 13
THE LAGRANGE MULTIPLIER We solve the constrained optimization problem by introducing an extra variable π (or extra variables π ) Setting all partial derivatives of β to 0 , we find the optimal value of π¦ and π§ and the optimal value of π ... ... but we also find the optimal value for π Does it have a meaning? Letβs go back to the simple case α max π π¦, π§ s.t. π π¦, π§ = π 14
THE LAGRANGE MULTIPLIER Suppose the constant in the constraint π changes βͺ recall constraint equation: π π¦, π§ = π βͺ example, the available budget changes Can we write the problem as a function of π ? βͺ the optimal values of π¦ and π§ depend on π βͺ so π¦ β = π¦ π and π§ β = π§ π βͺ and so does the optimal value of π = π β π βͺ can be written as a function of π : π π¦ π , π§ π Theorem ππ β π = π ππ 15
THE LAGRANGE MULTIPLIER Example Consider the production function optimization problem π πΏ, π , with πΏ capital and π labour: α max π πΏ, π = 120πΏπ s.t. 2πΏ + 5π = π Introduce the Lagrangian β πΏ, π, π = 120πΏπ β π 2πΏ + 5π β π 16
THE LAGRANGE MULTIPLIER Example (continued) Solving the first-order conditions (for given π ) πβ βͺ ππΏ = 120π β 2π = 0 πβ βͺ ππ = 120πΏ β 5π = 0 πβ βͺ ππ = β2πΏ β 5π + π = 0 yields 1 βͺ πΏ = πΏ π = 4 π 1 βͺ π = π π = 10 π βͺ π = π π = 6π 17
THE LAGRANGE MULTIPLIER Example (continued) The maximum value at these maximum points is βͺ π πΏ π , π π = 120πΏπ = 3π 2 This has been defined above as π β π ππ β π Now, consider = 6π ππ If π = 100 and if π is increased by Ξπ = 1 to π = 101 ; then π β π increases approximately by π π β π α Γ Ξπ = 6 Γ 100 Γ 1 = 600 ππ π=100 =π π 18
THE LAGRANGE MULTIPLIER Example (continued) βͺ So approximate change of π β is 600 βͺ Check approximation with exact result: βͺ π β 101 β π β 100 = 3 Γ 101 2 β 3 Γ 100 2 = 603 βͺ Good agreement Conclusion The value of a Lagrange multiplier gives the rate of change of the optimum value when the constraint constant is changed by a unit amount 19
EXERCISE 3 In a Lagrange problem, we have π = 1.5 and optimum value π π¦ β , π§ β = 30 . The budget increases by 4 . What happens? 20
OLD EXAM QUESTION 10 December 2014, Q3c 22
OLD EXAM QUESTION 27 March 2015, Q3d 23
FURTHER STUDY Sydsæter et al. 5/E 14.6 Tutorial exercises week 6 Lagrange method with three variables Lagrange method with two constraints 24
Recommend
More recommend