Upload
vutruc
View
227
Download
0
Embed Size (px)
Citation preview
Truss Design and ConvexOptimization
Robert M. Freund
April 23-25, 2002
Outline�Physical Laws of a Truss System
�The Truss Design Problem
�Second-Order Cone Optimization
�Truss Design and Second-Order Cone Optimization
�Some Computational Results
�Extensions of the Truss Design Problem
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 1
Outline�Semi-Definite Optimization
�Truss Design and Semi-Definite Optimization
�Truss Design and Linear Optimization
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 2
Physical Laws ofa Truss System
A Truss Structure
A truss is a structure in � � � or � � � dimensions,formed by � nodes and� bars joining these nodes.
2
3
4
5
6
1
Example of a truss in � � � dimensions,with � � � nodes and � � �� bars.
Examples of trusses include bridges, cranes, and the Eiffel Tower.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 3
Physical Laws ofa Truss System
Data Description of a Truss
The data used to describe a truss is:
� a set of nodes (given in physical space)
� a set of bars joining pairs of nodes with associateddata for each bar:
– the length �� of bar �– the Young’s modulus �� of bar �
– the volume �� of bar �
� an external force vector � on the nodes
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 4
Physical Laws ofa Truss System
Data Description of a Truss
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 5
Physical Laws ofa Truss System
Data Description of a Truss
Static and Free Nodes
Nodes can be:
� static nodes (fixed in place), or
� free nodes (movable when the truss is stressed).
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 6
Physical Laws ofa Truss System
Data Description of a Truss
Degrees of Freedom
The allowable movements of the nodes defines thedegrees of freedom (dof) of the truss.
We say that the truss has degrees of freedom.
� ��
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 7
Physical Laws ofa Truss System
Displacements and Forces
Movements of nodes in the truss will be repre-sented by a vector � of displacements � � �� .
The external force on the truss is given by avector � � �� .
Displacements of the nodes in the truss causeinternal forces of compression and/orexpansion to appear in the bars in the truss.
Let �� denote the internal force of bar �. Thevector � � �� is the vector of forces of thebars.
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 8
Physical Laws ofa Truss System
Example of a Truss Problem
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
There is a single external force appliedto node � in the direction indicated.
This will result in internal forces � alongthe bars in the truss and will simultane-ously cause small displacements � inall of the nodes.
Nodes � and � are fixed, and the othernodes are free.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 9
Physical Laws ofa Truss System
Example of a Truss Problem
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
We can label the 13 different bars by the nodes theylink. The bars will be: ��, ��, , ��.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 10
Physical Laws ofa Truss System
Example of a Truss Problem
Expansion and Compression
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
The internal force �� of bar � can be positive or negative.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 11
Physical Laws ofa Truss System
Forces on the Bars
Expansion
Lk∆ > 0
A bar under expansion.
If �� � �, bar � has been expanded and its internalforce counteracts the expansion with compression.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 12
Physical Laws ofa Truss System
Forces on the Bars
Compression
Lk
∆ < 0
A bar under compression.
If �� � �, bar � has been compressed and its internalforce counteracts the compression with expansion.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 13
Physical Laws ofa Truss System
Physical Laws
Force Balance Equations...
In a static truss the internal forces will balance the external forces in everydegree of freedom. This is a law of conservation of forces.
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
Consider the balance of forces on node 3:
� ��������� ���� ���� ������ � ���� ���� ������ � � ����
� ��������� ���� ������ � ���� ���� ������ � � ����
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 14
Physical Laws ofa Truss System
Physical Laws
...Force Balance Equations...
For the entire truss we write linear equations thatrepresent the balance of forces in each degree offreedom.
In matrix notation this is:
�� � ��
� is an �� matrix.
Each column of �, denoted as �, is the projection ofthe bar onto the degrees of freedom of the nodes thatbar � meets.c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 15
Physical Laws ofa Truss System
Physical Laws
...Force Balance Equations...
2
3
4
5
6
1f
f
F13
12
3
A truss problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 16
Physical Laws ofa Truss System
Physical Laws
...Force Balance Equations
In our example, we have:
�
�� �� � �� �� � �� � �� �� � � ��
�������������������� � � � ���
� ���
� � � � � �
�� � � � � ���
� � ���
� � � � � �
� �� � � � ���
� � � � ���
� � �
� � � � ���
� � � � ���
� � �
� � � ���
� � �� � � � � ���
� �
� � � ���
� � � � �� � � � ���
� �
� � � � ���
� � � � � � ���
� �� �
� � � � ���
� � � � � � ���
� � ���
�������������������� �
�� ��� �
�� � � �
� ��� �
�� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 17
Physical Laws ofa Truss System
Physical Laws
Constitutive Relationship Between Forces and Distortion of a Bar
If bar � has length �� and cross-sectional area ��,Young’s modulus ��, and its length is changed by ��,then the internal force �� is given by:
�� � ���
���
��
�� � �
� ��
We write:
�� ���
�������
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 18
Physical Laws ofa Truss System
Physical Laws
Distortion and Displacements ...
Displacements in the nodes will cause the lengths of the bars to change.
Suppose that ��� � ��� � ��� � ���, with all other barsmeasured proportionately.
Consider a displacement of: � � ����������� �
2
2
3
4
5
6
1
F3
Example of a small displacement in a node.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 19
Physical Laws ofa Truss System
Physical Laws
...Distortion and Displacements ...
� � �����������������
2
2
3
4
5
6
1
F3
Example of a small displacement in a node.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 20
Physical Laws ofa Truss System
Physical Laws
...Distortion and Displacements ...
Changes in the lengths of the bars under nodal displacement.Bar � New Length Linearized Length Linearized Change�� ����
��
12
�� � ����� �� �� � �� �
13 1 1 0 014 1 1 0 016
��
�� 0 0
23
�� � ���
�� 0 0
24
�� � ����� �� � � � � ��
25
�� � ����� ��
�� � ���� ���� � ����
34 1 1 0 035 1 1 0 036
��
�� 0 0
45
��
�� 0 046 1 1 0 056 1 1 0 0
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 21
Physical Laws ofa Truss System
Physical Laws
...Distortion and Displacements ...
If � is small, then the distortion of bar � is nicelyapproximated by the linear expression:
�� � � � �
We approximate the internal force on bar � due to afeasible displacement � as:
�� ���
���
���� � �������� � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 22
Physical Laws ofa Truss System
Physical Laws
...Distortion and Displacements
Define:
� ��
����
����
�
. . .
� ���
�����
� � ��� ��
�����
���
�
. . .
� ����
���
��
�� � �������� � �
can be written as: � � ����� � �which is �� �� � � � c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 23
Physical Laws ofa Truss System
Physical Laws
Equilibrium Conditions...
We write the physical laws of the truss system as:
�� � �� (conservation of forces)
�� �� � � � (forces, distortions, and displacements)
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 24
Physical Laws ofa Truss System
Physical Laws
...Equilibrium Conditions� � �� (conservation of forces)
�� �� � � (forces, distortions, and displacements)
Combining these yields: � � ���
� ����� �
� ��
where: � � �����
The matrix � is called the stiffness matrix of the truss.
Note that � is an SPSD matrix.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 25
Physical Laws ofa Truss System
Physical Laws
Solving the Equations� � �� (conservation of forces)
�� �� � � (forces, distortions, and displacements)
Form: � � �����
Solve for �: �� � �
Then compute: � � ����� �
This last expression is simply: �� � �������� � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 26
Physical Laws ofa Truss System
Physical Laws
Compliance
The compliance of the truss is the work (or energy)performed by the truss.
The compliance is the sum of the forces timesdisplacements:
� � � � �� � �
The compliance will generally be a positive quantity.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 27
Physical Laws ofa Truss System
Physical Laws
A View from Optimization...
Consider the following optimization problem:
�� minimize ��
�����
��
� ����
��
����
s.t.
������� ��� � �� �
The constraints can be written:
�� �
������� ��� � ��
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 28
Physical Laws ofa Truss System
Physical Laws
...A View from Optimization...
�� minimize ��
�����
���
�����
��
����
s.t.
������� ��� � �� �
This is a convex quadratic problem. The optimality conditions are:
��������� � ��
�������
��� ��
�� � � for � � �� � � � ���
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 29
Physical Laws ofa Truss System
Physical Laws
...A View from Optimization...
Optimality conditions:
��������� � ��
���
����� ��
�� � � for � � � � � � ��
Re-write as:
� � ��
�� �� � � �c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 30
Physical Laws ofa Truss System
Physical Laws
...A View from Optimization...
Optimality conditions of ���:
� � ��
�� �� � � �
These are the original equilibrium conditions of the truss.
The truss system is nature’s solution to an optimization problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 31
Physical Laws ofa Truss System
Physical Laws
...A View from Optimization
�� minimize ��
�����
��
���
�����
����
s.t.
������� ��� � �
The optimal objective function value is:
�����
���
�����
��
��� ��
����� � ������� � ������ ��
�� ��
The optimal objective value of ��� is the compliance of thetruss system.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 32
The TrussDesign Problem
We now consider the volumes �� of the bars � to be design parameters that we wish to determine.
� � ��
� �����
�
��������������
�
�
������������
�
���������
������������� � �����
where ���� �
�����
����
������� ����� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 33
The TrussDesign Problem
�� �
���������
��� � �
� � �� is a weighted sum of rank-one matrices(weighted by the volumes ��).
� � �� is a weighted sum of outer-productmatrices.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 34
The TrussDesign Problem
In designing the truss, we choose the volumes � of the barssubject to linear constraints:
� � � � � �
Typically, these constraints include upper and lower bounds onthe volumes of certain bars, as well as an overall cost or volumeconstraint:
����� � � � �
The criteria in truss design is to choose the volumes � of the barsso as to minimize the compliance of the truss, namely:
� �� �
Such a truss will be the most resistant to the external force � .c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 35
The TrussDesign Problem
Single-load Truss Design Problem
���� ����������� ���
���� �� � � �
� � �
� �
� � � � �� �
���� ����������� ���
����
���
��� ���
���
��� �����
� � �
� � �
� �
� � � � �� �c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 36
The TrussDesign Problem
Single-load Truss Design Problem
����� �������� ���
����
���
���
����
���
���� �����
� � �
�� � �
� � �� � �� � � � �� �
The decision variables are � and �.
The real decision variables are � only.
Once � is chosen, � will be determined by the solutionto the system of equations: ��� � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 37
The TrussDesign Problem
Single-load Truss Design Problem
� �� ����������� � �
��
���
�������
���
� � �
� � �
�� � �
� � �
� � ��� � � ��
Note that as written, the truss design problem TDP isnot a convex problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 38
The TrussDesign Problem
A Convex Version of TDP
Recall problem ���:
�� minimize ��
�����
���
�����
��
����
s.t.
������� ��� � �� �
Convert � � � � � � � to be decision variables and addconstraints
�� � �
� � � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 39
The TrussDesign Problem
A Convex Version of TDP
Write this as:
(TDP�� minimize���
�� �
��
�����
��
���
s.t.
��� ���� � ��
�� � �
� � �� � ��� � � ���
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 40
The TrussDesign Problem
A Convex Version of TDP
�� minimize ��
�����
��
���
�����
����
s.t.
������� ��� � �
(TDP� minimize����
�����
��
�����
��
���
s.t.
��������� � �
�� � �
� � �
� � ��� � � ��
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 41
The TrussDesign Problem
A Convex Version of TDP
Two modifications:
1. bars with zero volume (�� � �) are no longercounted
2. � instead of ��
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 42
The TrussDesign Problem
A Convex Version of TDP
(TDP� minimize����
�����
��
����
�����
s.t.
��������� � ��
� � �
� �
� � �� � ���
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 43
The TrussDesign Problem
A Convex Version of TDP
Re-write as:
(TDP�� minimize����
���
�����
s.t.
��������� � ��
�� � �
�������� � ���� for � � �� � � ��
� � �� � � �
�� �� � � ���c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 44
The TrussDesign Problem
A Convex Version of TDP
(TDP�� minimize����
���
�����
s.t.
��������� � ��
�� � �
�������� � ���� for � � �� � � ��
� � �� � � �
�� �� � � ���c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 45
The TrussDesign Problem
A Convex Version of TDP
Further re-write as:
(TDP�� minimize����
���
�����
s.t. � � ��
�� � �
�������� � ���� for � � �� � � ��
� � �� � � �
�� �� � � ���c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 46
The TrussDesign Problem
A Convex Version of TDP
(TDP�� minimize����
���
�����
s.t. � � ��
�� � �
���
����� � ���� for � � �� � � ��
� � �� � � �
�� �� � � ���c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 47
The TrussDesign Problem
A Convex Version of TDP
We have a linear objective function and almost all constraints arelinear.
It is pretty easy to show that the constraints:���
����� � ��� � � � �� � �
describe a convex region.
This is a convex optimization problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 48
Second-OrderCone Optimization
Definition of SOCP
A second-order cone optimization problem (SOCP) isan optimization problem of the form:
����� ���� � �
�� �� � �
����� ��� �
� � �� ��
� � � �� � �
In this problem, the norm ��� is the standard Euclidean norm:
��� ��� �
The norm constraints in SOCP are called“second-order cone” constraints.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 49
Second-OrderCone Optimization
Linear Optimization is an SOCP����� ���� � �
�� �� � �
����� ��� �
� � �� ��
� � � �� � �
SOCP is a convex problem, because
����� ��� �� � �� ��
is a convex function.
Linear optimization is a special case of SOCP: just set
�� � �, �� � �, �� � �, and �� to be the �th unit vector,
� � �� � �.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 50
Second-OrderCone Optimization
Quadratic Constraints and SOCP
���� ���� ���
���� � � �
���� ��� ����� � ��� � � � � � � � �
Suppose we have the constraint: ��� ��� � �� � � �
Factor � �� ������
���� ����� ���
�
���� � ����� �� �
�
Square both sides and collect terms.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 51
Second-OrderCone Optimization
Quadratic Objective Function and SOCP
�� � ���� ��
���� � � �
����� ��� ���� �� ��� � � �� � � � � � �
Suppose we have a convex quadratic objective:
��� � ������ ���
Create a new variable ���� and write:
�������� ����
����
...
������ ��� � ����
This can be further re-written as an SOCP.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 52
Truss Design and Second-Order Cone Optimization
CTDP
Recall:(TDP�� minimize���
��
�������
s.t. � � ��
�� � �
�������� � ���� for � � �� � � ��
� � �� � � �
�� �� � � ���c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 53
Truss Design and Second-Order Cone Optimization
CTDP
Make the simple change of variables:
�� �
�� ��
���
�� � ��� ��
���
Then �� � �� �� and � � �� � ��, and so:
��� � ��� � ����� �� � ��
� � ���
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 54
Truss Design and Second-Order Cone Optimization
CTDP
The constraint
�������� � ���
becomes:
�������� � ��� � ��� �
Rearrange this and take square roots:����
����� ��
� � ��
and re-write as: ����
�����
�������� � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 55
Truss Design and Second-Order Cone Optimization
CTDP
(TDP�� minimize����
���
�����
s.t. � � ��
�� � �
���
����� � ���� for � � �� � � ��
� � �� � � �
�� �� � � ���c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 56
Truss Design and Second-Order Cone Optimization
CTDP
can be re-written as:
(CTDP): minimize����
���� � ��
s.t. � � ��
�����
������
��������� � � � � � �� � � � ��
�� � �� � �
��� � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 57
Truss Design and Second-Order Cone Optimization
CTDP
(CTDP): minimize����
���� � ��
s.t. � � ��
�����
������
��������� � � � � � �� � � � ��
�� � �� � �
��� � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 58
Truss Design and Second-Order Cone Optimization
CTDP
CTDP is a second-order cone problem.The LHS is the norm of a 2-dimensional vector:
��!�� !��� ������
�� �����
���
������� �
and the RHS is the linear expression:
��
Proposition Suppose that ��� � �� is a feasible or optimal solution of CTDP.Let:
� � � � and � � � ��
Then ��� �� �� is the corresponding feasible or optimal solution of TDP�.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 59
Truss Design and Second-Order Cone Optimization
Other Formulations of TDP
We have seen TDP�, which is a convex problem.
We have also seen CTDP, which is a second-order cone problem.
We can also formulate the TDP as a “semidefinite optimization”(SDO for short) problem.
In the special case when the constraints on the volume variables
� � � � � are:
� �
����� � � �
TDP can actually be solved by linear optimization.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 60
SomeComputational Results
Three Truss Design Problems
Bridge Design Problem
0 2 4 6 8 10 12 14 16
0
1
2
3
4
5
6
7
The forces and nodes for the basic bridge design model.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 61
SomeComputational Results
Three Truss Design Problems
Possible Bars for the Bridge
0 2 4 6 8 10 12 14 16
0
1
2
3
4
5
6
7
The set of possible bars for the basic bridge design problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 62
SomeComputational Results
Hanging Sign Design Problem
0 5 10 15 20 25 30
0
2
4
6
8
10
12
14
16
18
20
The forces and nodes for the hanging sign design model.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 63
SomeComputational Results
Crane Design Problem
0 2 4 6 8 10 12 14 16 18 20
0
5
10
15
20
25
30
35
40
The forces and nodes for the crane design model.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 64
SomeComputational Results
Solution of the Basic Bridge Design Problem
0 2 4 6 8 10 12 14 16
0
1
2
3
4
5
6
7
Optimal solution to the basic bridge design problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 65
SomeComputational Results
Solution of the Hanging Sign Design Problem
0 5 10 15 20 25 30
0
2
4
6
8
10
12
14
16
18
20
Optimal solution to the hanging sign design problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 66
SomeComputational Results
Solution of the Crane Design Problem
0 2 4 6 8 10 12 14 16 18 20
0
5
10
15
20
25
30
35
40
Optimal solution to the crane design problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 67
SomeComputational Results
Solution of the Enhanced Bridge Design Problem
Lower Bounds on “Road-Surface” Bars
0 2 4 6 8 10 12 14 16
0
1
2
3
4
5
6
7
Optimal solution of the bridge design problem,with lower bounds on the “road surface” bar volumes.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 68
SomeComputational Results
Details of Problems Solved
Dimensions of the Truss Design Problems
Dimensions of the truss design problemsName Size of Nodes Arcs Degrees of Maximum
Grid � � Freedom � VolumeBridge-16�7 16�7 136 1,547 268 1,600Bridge-20�10 20�10 231 2,878 458 2,000Bridge-30�10 30�10 341 4,368 678 4,000
Sign-10�20 10�20 231 2,878 444 1,000Sign-20�30 20�30 651 9,058 1,280 2,000Sign-30�40 30�40 1,271 18,438 2,512 3,000
Crane-10�20 10�20 96 818 188 2,000Crane-20�40 20�40 291 3,188 578 5,000Crane-30�40 30�40 401 4,678 798 10,000
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 69
SomeComputational Results
Details of Problems Solved
Variables and Inequalities
Number of variables and inequalities in the truss design problemsLP Model SOCP Model
Problem Variables Inequalities Variables Inequalities
�� �� �� �� �� LBsBridge-16�7 4,641 1,564Bridge-20�10 8,634 2,899Bridge-30�10 13,104 4,399
Sign-10�20 5,756 5,756 8,634 2,879Sign-20�30 18,116 18,116 27,174 9,059Sign-30�40 36,876 36,876 55,314 18,439
Crane-10�20 1,636 1,636 2,454 819Crane-20�40 6,376 6,376 9,564 3,189Crane-30�40 9,356 9,356 14,034 4,679
� The linear optimization model cannot be used for the modified bridge design model.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 70
SomeComputational Results
Details of Problems Solved
Variables and Equations
Number of variables and equations in the truss design problemsLP Model SOCP Model SDO Model
Problem Variables Equations Variables Constraints Matrix Dimension
�� � �� � ��� �� LBs � ��� �
Bridge-16�7 4,641 1,832 1,685Bridge-20�10 8,634 3,357 3,111Bridge-30�10 13,104 5,077 4,711
Sign-10�20 5,756 444 8,634 3,323 3,111Sign-20�30 18,116 1,280 27,174 10,339 9,711Sign-30�40 36,876 2,512 55,314 20,951 19,711
Crane-10�20 1,636 188 2,454 1,007 916Crane-20�40 6,376 578 9,564 3,767 3,481Crane-30�40 9,356 798 14,034 5,477 5,081
� The linear optimization model cannot be used for the modified bridge design model.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 71
SomeComputational Results
Details of Problems Solved
Compliance of the Optimized Truss Design
Compliance of the optimized truss designfor the truss design problems
Problem LP SOCP SDPBridge-16�7 27.43365 27.15510Bridge-20�10 52.58314 52.10850Bridge-30�10 132.46038
Sign-10�20 0.77279 0.77293 0.77279Sign-20�30 2.52003 2.52056
Sign-30�40 4.08573 4.08681
Crane-10�20 28.67827 28.67840 28.67753Crane-20�40 286.24854 286.24954 286.19340Crane-30�40 596.73177 596.73324 596.63740
� The linear optimization model cannot be used for the for the modified bridge design model.
� The data to run the SDO model could not be prepared for this instance due to memoryrestrictions.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 72
SomeComputational Results
Details of Problems Solved
Iterations of Interior-Point Method
LOQO was used tosolve the LP and SOCPmodels.
���� was used tosolve the SDO models.
Number of iterations of the interior-pointalgorithm to solve the truss design problems
Problem LP SOCP SDOBridge-16�7 38 36Bridge-20�10 55 43Bridge-30�10 47
Sign-10�20 18 61 37Sign-20�30 21 47
Sign-30�40 24 53
Crane-10�20 17 52 34Crane-20�40 31 158 58Crane-30�40 31 144 60
� The linear optimization model cannot be used for the modified bridge design model.
� The data to run the SDO model could not be prepared for this instance due to memoryrestrictions.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 73
SomeComputational Results
Details of Problems Solved
Running Times (in seconds)
Running time (in seconds) of the interior-pointalgorithm to solve the truss design problems
Problem LP SOCP SDOBridge-16�7 93.90 257.01Bridge-20�10 460.96 1,088.09Bridge-30�10 904.86
Sign-10�20 3.33 513.37 1,081.54Sign-20�30 32.08 4,254.08
Sign-30�40 111.03 26,552.74
Crane-10�20 0.52 41.12 446.98Crane-20�40 6.46 1,299.12 33,926.20Crane-30�40 10.67 1,599.85 95,131.49
� The linear optimization model cannot be used for the modified bridge design model.
� The data to run the SDO model could not be prepared for this instance due to memoryrestrictions.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 74
Semi-DefiniteOptimization
Semi-definite Notation
If � is a � � � matrix, then � is a symmetric positivesemi-definite (SPSD) matrix if � is symmetric:
��" � �"� for any �� � � �� � �
and
� �� � � for any � � ��
If � is a � � � matrix, then � is a symmetric positivedefinite (SPD) matrix if � is symmetric and
� �� � for any � � ��� � � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 75
Semi-DefiniteOptimization
Semi-definite Notation
Let �� denote the set of symmetric � � � matrices.
Let ��� denote the set of symmetric positivesemi-definite (SPSD) � � � matrices.
Let ���� denote the set of symmetric positive definite(SPD) � � � matrices.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 76
Semi-DefiniteOptimization
Semi-definite Notation
Let � and! be any symmetric matrices.
We write “� � �” to denote that � is symmetric andpositive semi-definite.
We write “� �” to denote that � is symmetric andpositive definite.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 77
Semi-DefiniteOptimization
��� is a Convex Set
Remark 1 ��� � �� � �� � � � �� is a convex set in ���.
Proof : Suppose that ��! � ���. Pick any scalars
"�# � � for which "� # � �. For any � � ��, wehave:
� "� � #!� � "� ��� #� !� � � �
whereby "� � #! � ���. This shows that ��� is aconvex set. q.e.d.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 78
Semi-DefiniteOptimization
Definition of SDO
� �� ��������� � $
�� % �
�����$��� � �
�$ � �
The matrices %���� ��� are symmetric matrices.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 79
Semi-DefiniteOptimization
Interpretation of SDO
� �� ��������� � $
�� % �
�����$��� � �
�$ � �
The objective is to minimize the linear function:��
�����$�
of the� scalar variables $ � $�� � $�.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 80
Semi-DefiniteOptimization
Interpretation of SDO
� �� ��������� � $
�� % �
�����$��� � �
�$ � �
$ � $�� � $� must satisfy�$ � �.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 81
Semi-DefiniteOptimization
Interpretation of SDO
��� ��������� ���
����
�������� �
�� � � �
$ � $�� � $� must satisfy the condition that thematrix �, defined by:
� �� % �
�����$��� �
must be positive semi-definite. That is,
� �� % �
�����$��� � �c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 82
Semi-DefiniteOptimization
Illustration of SDO��� ����������� ����� �!��
����
��� � �
� ! �
� � "�
�� ���
�� � �
� � "
� " ��
�� ���
�� � #
� � �
# � �
� � # � �
���� "�� � ��
���� �� � � �
��� ������� ����� �!��
���� �� � � ���� ��� � � ���� ��� � � ���� #��� � ���� ��� ! � ���� ��� � � "��� ���
� � ���� #�� � � "��� ��� " � ���� ���
� � �
���� "�� � ��
���� �� � � �c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 83
Semi-DefiniteOptimization
SDO is Convex
� �� ��������� � $
�� % �
�����$��� � �
�$ � �
Remark 2 . SDO is a convex minimization problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 84
Semi-DefiniteOptimization
Further Remarks on SDO
Semi-definite optimization is a unifying model that includes:
linear optimization;
quadratic optimization; second-order cone optimization;
certain other convex optimization problems.
Semi-definite optimization has applications that span convexoptimization, discrete optimization, and control theory.
Semi-definite optimization may very well become the canonicalway that optimizers will think about optimization in the nextdecade.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 85
Truss Design and Semi-Definite Optimization
STDP�� �� ����������$ &
��
�& �
�
���
�������
��� � � ��
� �
�� � �
� � �
& � �� � � ��
Notice that STDP is a semi-definite optimization problem.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 86
Truss Design and Semi-Definite Optimization
TDP Yields Equivalent Solutions to STDP���� ����������� ��� ����� ����������� !
����
���
��� ���
�����������
� � � ����
�! � �
�
���
��� ���
������������
�
� � � � � �
� � � �
� � � � �� � ! � � � �� �
Proposition 3 . Suppose that � � is a feasible solution of TDP.Let:
! � � �� �Then � ! is a feasible solution of STDP, and ! � � ��.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 87
Truss Design and Semi-Definite Optimization
STDP Yields Equivalent Solutions to TDP���� ����������� ��� ����� ����������� !
����
���
��� ���
�����������
� � � ����
�! � �
�
���
��� ���
������������
�
� � � � � �
� � � �
� � � � �� � ! � � � �� �
Proposition 4 . Suppose that �� & is a feasiblesolution of STDP. Then there exists a vector � forwhich �� � is feasible for TDP, and in fact:
� � � & c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 88
Truss Design and Semi-Definite Optimization
Solving TDP via STDP���� ����������� ��� ����� ����������� !
����
���
��� ���
�����������
� � � ����
�! � �
�
���
��� ���
������������
�
� � � � � �
� � � �
� � � � �� � ! � � � �� �
Solve STDP for � ! and then solve the following linearequation system for �:�
����� ���
�������
��
� � �� �c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 89
Truss Design andLinear Optimization
Special Case of Truss Design
���� ����������� ���
����
���
��� ���
�����������
� � �
����� � � �
� �
� � � � �� �
The only constraint on the volumes of the bars � is a volumeconstraint limiting the total volume of the bars to not exceed thegiven value � .
This is a special case of TDP, but it is not too unusual.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 90
Truss Design andLinear Optimization
Special Case of Truss Design
����� �������� ���
����
���
�������
������������
� � �
������� � %
� � �� � �� � � � �� �
���� ��
���������
������������
� ����� �������� ���
���� ����� � �
������� � %
� � �� � �� � � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 91
Truss Design andLinear Optimization
A Dual of TDP
���� ����������� ���
���� �� � � �
����� � � �
� �
� � � � �� �
����� ����������� ��� �" � � #
����
����"�� � ���
��# � � � � � � �
� � � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 92
Truss Design andLinear Optimization
Weak Duality Property���� ����������� � �� ����� ����������� � �� �" � � #
���� �� � � � ����
����"�� � ���
��# � � � � � � �
����� � � � � � � �
� �
� � � � �� �
Proposition 5 . Suppose that �� � is feasible for TDPand that �� ' is feasible for DTDP. Then:
� � � ��� � � ( ' c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 93
Truss Design andLinear Optimization
A Pair of Linear Models
Consider the following pair of primal and dual linearoptimization models:
��� ������������
�����
��������� ����
���� ��� � �� � ��
�� � � �� � �
�� �� � �� �
��� ��������� �� ��
���� � ������ ���� � ���
�� � � � � � � �
� � � �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 94
Truss Design andLinear Optimization
Linear Models and Original Models
Dual Problems
�$�� �����������
�����
�����
�� � ��
�
�$�� ��%����� � � ��
���� �� � ��� � �� ���� � ������ ���� � ���
��� � � �� � � � ��
� � �� �� � � � � �� �
�� �� � �� �
����� �������� � �� ������ ��%������ � �� �! � % &
���� ����� � � ����
���!� � ���
��& � � � �� � � � ��
������� � % � � �� �
� � �� � �� � � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 95
Truss Design andLinear Optimization
Linear Model Solves the Original Problem
Proposition 6 . Suppose that � ��� ��� is an optimal solution ofLP and that �� is an optimal solution of LD, and consider thefollowing assignment of variables:
$ �
�����
���������� �����
� � �
��
���������� ������ � � � � � �
�� � �����
�" � ��
��
�# � ��
� �
Then �� �� solves TDP and ��" �# solves DTDP.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 96
Truss Design andLinear Optimization
Comparison of the Models�$�� �����������
�����
�����
�� � ��
�
�$�� ��%����� � � ��
���� �� � ��� � �� ���� � ������ ���� � ���
��� � � �� � � � ��
� � �� �� � � � � �� �
�� �� � �� �
����� �������� � �� ������ ��%������ � �� �! � % &
���� ����� � � ����
���!� � ���
��& � � � �� � � � ��
������� � % � � �� �
� � �� � �� � � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 97
Extensions of theTruss Design Problem
Multiple Loads
Conservative Model
We consider a finite set of external loads on the truss:
� � ���� ��� � �'� � )*�
A conservative strategy would be to minimize themaximum compliance:
��������� ���"�������'�� " �"�
�� � ��
� � � �� ���" � �"
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 98
Extensions of theTruss Design Problem
Multiple Loads
Averaging Model
We consider a finite set of external loads on the truss:
� � ��� �� � � � �� � %$ �
Let &� denote the relative frequency or importance associatedwith the truss structure undergoing the external force ��.
&� � � � &� � � �
�����&� � ��� �
���������
�����&���
� ��
� � �
� � ���� �� �� � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 99
Extensions of theTruss Design Problem
Self-Weights
We have assumed that the truss structure itself is not affected by its own weight.
Add an external force corresponding to the gravitational force associated withbar �, and linearly proportional to �� for � � �� � � � ��.
Let �� � ()� denote the vector that projects the gravitational force of bar �
onto the appropriate nodes.
����� ���������
� �
�����������
�
���� ����� ��
� �
����������
�� � �
� � �� � �� � � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 100
Extensions of theTruss Design Problem
Reinforcement
We are given an existing truss structure.
We must determine how to strengthen it.
We add lower bounds on the volumes of the existing bars equal tothe current volumes of the bars:
���� ����������� ���
���� �� � � �
� � �
� �
� � � � � � � � � �
� � � � �� �
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 101
Extensions of theTruss Design Problem
Robustness of Solutions
The optimal solution of the TDP might yield a trussdesign that is not rigid for a different external forcethan the one used.
To obtain a robust solution, we must consider multipleexternal loads that will contain the possible loads thatthe truss will be subject to.
c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 102
Extensions of theTruss Design Problem
Buckling Constraints
We have ignored the fact that if a bar is under greatcompression it might actually collapse instead ofcounteracting that external force with its internal force
For this we have to add lower bounds on the allowableinternal forces in terms of the design variables �� andthe geometry of the bars.
These constraints cause the resulting problem to loseits convex structure.c�2002 Massachusetts Institute of Technology. All rights reserved. 15.094 103