Nonlinear model for min rectangular area I'm trying to slove one problem.
Find the rectangle of minimum area that encloses each of these 4 points 
(2,3), (4,4), (4,2), (6,2).
I need for the Nonlinear problem (model) formulation.
I think, using  the rotation (for the angle arctg(0.5) in this case) 
we reduce the problem of findig such rectangle whose pair-sides are paralel to x (and y) axis.
With the rotation, it sems for me that this problem made little bit easyer for doing the model, but I 
have problems with constructing the NL model also in this case. So, plese
Please, help.
 A: Not sure what you mean by "Nonlinear problem formulation" but this paper describes several algorithms for finding the answer, in particular the Rotating Calipers algorithm. 
Basically, you first determine the convex hull of the set of points (this will be a polygon). Then, as the paper says, "...the rotating calipers algorithm starts with a bounding rectangle having an edge coincident with a polygon edge and a supporting set of polygon vertices for the other polygon edges. The rectangle axes are rotated counterclockwise by the smallest angle that leads to the rectangle being coincident with another polygon edge." 
Using this algorithm I found the minimum area rectangle for your points. 

The blue points are your points and the blue lines give the convex hull. The initial bounding rectangle is the green one, coincident with the edge $AB$. Rotating the green rectangle, the subsequent rectangles are purple (coincident with $CD$), red (coincident with $BC$) and finally black (coincident with $AD$). The black rectangle is the minimum area rectangle with $Area \approx 1.79 \times 4.02 = 7.20$.
