A new approach to field modelling

Language:
de dk es fr it cz ru cn


RSS Twitter Facebook Linkedin YouTube

>> >> >>

LabelMover optimization benchmark

Here are LabelMover optimization benchmarks.

Task

Theory

QuickField

%

Download linear1. Move square's right edge to the right in order to get the maximal cross-section area. The movement range is 0..1
The optimized value is linearly dependent upon the optimization parameter.

2

1.9981

0.1%

Download linear2. Move square's right and top edges right and up correspondingly in order to get the maximal cross-section area. The movement range for each edge is 0..1.
The optimized value is linearly dependent upon two optimization parameters.

4

3.9836

0.41%

Download linear3. Move square's 3 edges (right, left and top) in order to get the maximal cross-section area. The movement range for each edge is 0..1.
The optimized value is linearly dependent upon three optimization parameters.

6

5.9674

0.54%

Download linear1_2. Move square's right edge by its two vertices moving right independently in order to get the maximal cross-section area. The movement range for each vertex is 0..1.
The optimized value is linearly dependent upon two optimization parameters.

2

1.9967

0.17%

Download linear1_11. There are 11 squares (like the one in task linear1). Move each square's right edge in order to get the maximal total cross-section area. Each edge can be moved independently. All in all there are 11 optimization parameters.
The optimized value is linearly dependent upon 11 optimization parameters.

22

21.928

0.33%

Download square1. Move half-circle's right vertex in order to get the maximal cross-section area. The movement range is 0..1.
The optimized value is square dependent upon the optimization parameter.

3.5343

3.5271

0.2%

Download cubic1. The same as previous. Axisymmetric model represents sphere. The sphere volume is calculated. The movement range is 0..1.
The optimized value is cubic dependent on the optimization parameter

14.1372

14.094

0.31%

Download nonlinear1. Move common boundary 1-1 of rectangle and half-circle in order to get the minimal area.
Half circle cross section is S1=π/2·(x/2)2.
The rectangle cross section is S2=0.5·(1-x).
The optimal parameter value is x=2/π.

0.6366

0.6366

0.00%

Download nonlinear2. The same as previous but the volume of sphere and cylinder is minimized.
The sphere volume is V1=4π/3·(x/2)3. Rectangle in axisymmetric model represents a cylinder. Its volume can be calculated as V2=π·0.52·(1-x).
The optimal parameter value is x=1/sqrt(2).

0.7071

0.7059

0.17%

Download nonlinear3. Minimization of the total length of the lines connecting the internal point of the rectangle with its vertices. Optimal position of the internal point should be in the intersection of the diagonals.

5.6569

5.6569

0.00%