Mathematics and Statistics · Ch 14 — Linear Programming
Feasible Region — Bounded and Unbounded
Feasible Region — Bounded and Unbounded
Once an LPP is formulated, its constraints are graphed exactly as in Std XI: each inequation is a half-plane, and the region satisfying all of them at once is the set of allowable plans.
Feasible and infeasible solutions
Feasible solution and feasible region
A feasible solution is any point that satisfies all the constraints and the non-negativity restrictions simultaneously. The collection of every feasible solution is the feasible region — the common (overlapping) region of all the half-planes.
A point that violates even one constraint is an infeasible solution and lies outside the feasible region. Because appear in almost every commercial LPP, the feasible region normally lies in the first quadrant.
Corner (vertex) points
Corner point
A corner point (vertex) of the feasible region is a point where two of its boundary lines intersect. Corner points are found by solving the relevant boundary lines two at a time as simultaneous equations.
Corner points are the heart of the graphical method: the optimum of a linear objective always occurs at one of them (Section 3).
Bounded vs. unbounded regions
The two shapes a feasible region can take
- A feasible region is bounded if it can be enclosed inside some circle (or rectangle) of finite size — it does not run off to infinity. Bounded regions typically arise when the binding constraints are all of the type.
- A feasible region is unbounded if it extends infinitely far in at least one direction. Unbounded regions typically arise when constraints dominate.
Boundedness controls whether an optimum exists …
A point satisfying every constraint and the non-negativity restrictions at once. Any point violating a cons …
The set of all feasible solutions — the overlap of all the constraint half-planes, bounded by straight-line segments and (usually) lyin …
A feasible region that fits inside a finite circle. On it a linear objective always attains both a max …
A feasible region that extends infinitely in some direction. A linear objective on it may have no maximum (or no minimum), so an extra c …