TOPICS
Search

Normal Cone


NormalCone

The normal cone to a closed convex set C subset= R^n at a point x in C is

 N_C(x)={v in R^n:v·(y-x)<=0 for every y in C}.

In the illustration, the point x is a vertex of the convex polygon C, and the red boundary rays of N_C(x) are generated by outward normal vectors v_1 and v_2 to the two incident sides.

For nonzero v and y-x, the defining inequality says that the angle between them is a right angle or an obtuse angle. The normal cone is a convex cone. By convention, N_C(x) is the empty set when x not in C.

At a point in the interior of a full-dimensional C, the normal cone is {0}. At a smooth boundary point of a full-dimensional C, it is the ray generated by the outward normal vector. At a vertex of a convex polyhedron, it is generated by the outward normal vectors of the facets incident to the vertex. Normal cones are fundamental in the optimality conditions of convex optimization theory.


See also

Convex Cone, Convex Optimization Theory, Convex Set, Normal Vector

Explore with Wolfram|Alpha

WolframAlpha

More things to try:

References

Rockafellar, R. T. Convex Analysis. Princeton, NJ: Princeton University Press, 1970.

Cite this as:

Weisstein, Eric W. "Normal Cone." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/NormalCone.html

Subject classifications