NONLINEAR PROGRAMMING

Overview
■ Problems that fit the general linear programming format
but contain nonlinear functions are termed nonlinear
programming (NLP) problems.
■ Solution methods are more complex than linear
programming methods.
■ Determining an optimal solution is often difficult, if not
impossible.
■ Solution techniques generally involve searching a solution
surface for high or low points requiring the use of advanced
mathematics.

3
Optimal Value of a Single Nonlinear
Function Basic Model
Profit function, Z, with volume
independent of price:
Z = vp - cf - vcv
where v = sales volume
p = price
cf = unit fixed cost
cv = unit variable cost

Add volume/price relationship:

v = 1,500 - 24.6p

Figure Linear Relationship of Volume to Price

4
Optimal Value of a Single Nonlinear Function

With fixed cost (cf = $10,000) and variable cost (cv = $8):
Profit, Z = 1,696.8p - 24.6p2 - 22,000

Figure The Nonlinear Profit Function

5
Optimal Value of a Single Nonlinear Function
Maximum Point on a Curve
■ The slope of a curve at any point is equal to the derivative of
the curve’s function.
■ The slope of a curve at its highest point equals zero.

Figure Maximum profit for the profit function

6
Optimal Value of a Single Nonlinear Function
Solution Using Calculus
Z = 1,696.8p - 24.6p2 - 2,000
dZ/dp = 1,696.8 - 49.2p
=0
p = 1696.8/49.2
= $34.49
v = 1,500 - 24.6p
v = 651.6 pairs of jeans
Z = $7,259.45

7
Constrained Optimization in Nonlinear
Problems Definition

■ A nonlinear problem containing one or more constraints

becomes a constrained optimization model or a nonlinear
programming (NLP) model.
■ A nonlinear programming model has the same general
form as the linear programming model except that the
objective function and/or the constraint(s) are nonlinear.
■ Solution procedures are much more complex and no
guaranteed procedure exists for all NLP models.

8
Constrained Optimization in Nonlinear Problems
Graphical Interpretation (1 of 3)
Effect of adding constraints to nonlinear problem:

Figure Nonlinear Profit Curve for the Profit Analysis Model

9
Constrained Optimization in Nonlinear Problems
Graphical Interpretation (2 of 3)

Figure A Constrained Optimization Model

10
Constrained Optimization in Nonlinear Problems
Graphical Interpretation (3 of 3)

Figure

11
Constrained Optimization in Nonlinear Problems
Characteristics

■ Unlike linear programming, solution is often not on the

boundary of the feasible solution space.
■ Cannot simply look at points on the solution space
boundary but must consider other points on the surface
of the objective function.
■ This greatly complicates solution approaches.
■ Solution techniques can be very complex.

12
 NONLINEAR PROGRAMMING
If any or all of mathematical programming functions
(objective function and constraint functions) are nonlinear
then we deal with it as Nonlinear Programming Problem
(NLP).
General Formulation:
Max or Min f(xi) i=1,2,3….n
s.t. gj(xi) (, , or =) bj j=1,2,3….m
f(xi) and g(xi) are functions of the n variables.

Convex Function: Suppose f”(x) exists for all x. Then f(x) is a convex
function if and only if f”(x)  0 for all x.
Concave Function: Suppose f”(x) exists for all x. Then f(x) is a concave
function if and only if f”(x)  0 for all x.
Hessian Matrix: The Hessian of f(x1,x2,x3,….,xn) is the nxn matrix whose
ijth entry is
 2 f 
 
 xi x j 

13
 Examples of Convex and Concave Functions
1. f(x)= x2 is a convex function
f”(x)=2  0, so f(x) is convex

2. f(x) = x1/2 is a concave function for S(0,infinity)

f”(x)= -x-3/2/4  0, so f(x) is concave function on S(0,infinity)

3. f(x) = ax+b is both convex and concave function

f”(x) = 0, so f(x) is both convex and concave

4. Hessian for f(x1,x2) = x13 + 2x1x2 + x22

H (x1,x2)= 6x1 2
2 2

14
 local max and
global max

local max

f’’(x)<0 f’’(x)>0 f’’(x)<0 f‘’(x)>0

local min local min and global min

EM 505 Operations Research

15
 Global Optimum
A univariate function with a negative second derivative
everywhere guarantees a global maximum at the point
(if there is one) where f’(X)=0. These functions are
called “concave down” or sometimes just “concave.”

A univariate function with a positive second derivative

everywhere guarantees a global minimum (if there is
one) at the point where f’(X)=0. These functions are
called “concave up” or sometimes “convex.”

Multivariate Global Optimum: If the Hessian matrix is

positive definite (or negative definite) for all values of the
variables, then any optimum point found will be a global
minimum (maximum).
EM 505 Operations Research

16
Types of NLP:
f(x)
Unconstrained Optimization:
Max f(xi)
f’(xi) = 0
If it is a concave function.

x
Equality Constraint:
Max f(x)
f(x)
s.t. x=3

X=3 x

17
Lagrange Multipliers: Lagrange multipliers can be used to solve NLPs.
Consider NLP:

Max or Min f(xi) i=1,2,3….n

s.t. gj(xi) = bj j=1,2,3….m
f(xi) and g(xi) are functions of the n variables.

To solve, we associate a multiplier λj with the jth constraint and form

Lagrangean

L(x1,x2,…..xn, λ1, λ2…………. λm )

= f(x1,x2,…..xn)+Σj=1j=m λj [b j-g (x1,x2,…..xn)]

L L L L L L
  ......     ......... 0
x1 x2 xn 1 2 m

18
Solve the following nonlinear programming using Lagrangean method:

19