Linear Programming Lecture Notes - Penn State Personal Web Server
Linear Programming Lecture Notes - Penn State Personal Web Server
Linear Programming Lecture Notes - Penn State Personal Web Server
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
connected set (as it is in a neighborhood in R n ) in order for this definition to be used or at<br />
least we must be able to define the concept of neighborhood on the set 1 .<br />
Exercise 2. Using analogous reasoning write a definition for a global and local minimum.<br />
[Hint: Think about what a minimum means and find the correct direction for the ≥ sign in<br />
the definition above.]<br />
In Example 1.1, we are constrained in our choice of x and y by the fact that 2x+2y = 100.<br />
This is called a constraint of the optimization problem. More specifically, it’s called an<br />
equality constraint. If we did not need to use all the fencing, then we could write the<br />
constraint as 2x+2y ≤ 100, which is called an inequality constraint. In complex optimization<br />
problems, we can have many constraints. The set of all points in R n for which the constraints<br />
are true is called the feasible set (or feasible region). Our problem is to decide the best values<br />
of x and y to maximize the area A(x, y). The variables x and y are called decision variables.<br />
Let z : D ⊆ R n → R; for i = 1, . . . , m, gi : D ⊆ R n → R; and for j = 1, . . . , l<br />
hj : D ⊆ R n → R be functions. Then the general maximization problem with objective<br />
function z(x1, . . . , xn) and inequality constraints gi(x1, . . . , xn) ≤ bi (i = 1, . . . , m) and<br />
equality constraints hj(x1, . . . , xn) = rj is written as:<br />
⎧<br />
max z(x1, . . . , xn)<br />
(1.5)<br />
⎪⎨<br />
⎪⎩<br />
s.t. g1(x1, . . . , xn) ≤ b1<br />
.<br />
gm(x1, . . . , xn) ≤ bm<br />
h1(x1, . . . , xn) = r1<br />
.<br />
hl(x1, . . . , xn) = rl<br />
Expression 1.5 is also called a mathematical programming problem. Naturally when constraints<br />
are involved we define the global and local maxima for the objective function<br />
z(x1, . . . , xn) in terms of the feasible region instead of the entire domain of z, since we<br />
are only concerned with values of x1, . . . , xn that satisfy our constraints.<br />
Example 1.4 (Continuation of Example 1.1). We can re-write the problem in Example<br />
1.1:<br />
⎧<br />
max A(x, y) = xy<br />
⎪⎨ s.t. 2x + 2y = 100<br />
(1.6)<br />
x ≥ 0<br />
⎪⎩<br />
y ≥ 0<br />
Note we’ve added two inequality constraints x ≥ 0 and y ≥ 0 because it doesn’t really make<br />
any sense to have negative lengths. We can re-write these constraints as −x ≤ 0 and −y ≤ 0<br />
where g1(x, y) = −x and g2(x, y) = −y to make Expression 1.6 look like Expression 1.5.<br />
1 Thanks to Bob Pakzad-Hurson who suggested this remark for versions after 1.1.<br />
3