Results 1 
7 of
7
Using Modeling Knowledge to Guide Design Space Search
 In Fourth International Conference on Artificial Intelligence in Design '96
, 1996
"... Automated search of a space of candidate designs seems an attractive way to improve the traditional engineering design process. To make this approach work, however, the automated design system must include both knowledge of the modeling limitations of the method used to evaluate candidate designs an ..."
Abstract

Cited by 31 (7 self)
 Add to MetaCart
Automated search of a space of candidate designs seems an attractive way to improve the traditional engineering design process. To make this approach work, however, the automated design system must include both knowledge of the modeling limitations of the method used to evaluate candidate designs and also an effective way to use this knowledge to influence the search process. We suggest that a productive approach is to include this knowledge by implementing a set of model constraint functions which measure how much each modeling assumptions is violated, and to influence the search by using the values of these model constraint functions as constraint inputs to a standard constrained nonlinear optimization numerical method. We test this idea in the domain of conceptual design of supersonic transport aircraft, and our experiments indicate that our model constraint communication strategy can decrease the cost of design space search by one or more orders of magnitude. To appear: Artificial ...
The Use Of Artificial Intelligence To Improve The Numerical Optimization Of Complex Engineering Designs
, 1996
"... Gradientbased numerical optimization of complex engineering designs promises to produce better designs rapidly. However, such methods generally assume that the objective function and constraint functions are continuous, smooth, and defined everywhere. Unfortunately, realistic simulators tend to vio ..."
Abstract

Cited by 8 (3 self)
 Add to MetaCart
Gradientbased numerical optimization of complex engineering designs promises to produce better designs rapidly. However, such methods generally assume that the objective function and constraint functions are continuous, smooth, and defined everywhere. Unfortunately, realistic simulators tend to violate these assumptions. We present several artificial intelligencebased techniques for improving the numerical optimization of complex engineering designs in the presence of such pathologies in the simulators. We have tested the resulting system in several realistic engineering domains, and have found that using our techniques can greatly decrease the cost of design space search, and can also increase the quality of the resulting designs.
Intelligent gradientbased search of incompletely defined design spaces
 Artificial Intelligence for Engineering Design, Analysis and Manufacturing
, 1996
"... Gradientbased numerical optimization of complex engineering designs o ers the promise of rapidly producing better designs. However, such methods generally assume that the objective function and constraint functions are continuous, smooth, and dened everywhere. Unfortunately, realistic simulators te ..."
Abstract

Cited by 5 (3 self)
 Add to MetaCart
Gradientbased numerical optimization of complex engineering designs o ers the promise of rapidly producing better designs. However, such methods generally assume that the objective function and constraint functions are continuous, smooth, and dened everywhere. Unfortunately, realistic simulators tend to violate these assumptions. We present a rulebased technique for intelligently computing gradients in the presence of such pathologies in the simulators, and show how this gradient computation method can be used as part of a gradientbased numerical optimization system. We tested the resulting system in the domain of conceptual design of supersonic transport aircraft, and found that using rulebased gradients can decrease the cost of design space search by one or more orders of magnitude.
A Computational Environment for Exhaust Nozzle Design
 In Proceedings, Computing in Aerospace 10
, 1995
"... The Nozzle Design Associate (NDA) is a computational environment for the design of jet engine exhaust nozzles for supersonic aircraft. NDA may be used either to design new aircraft or to design new nozzles that adapt existing aircraft so they may be reutilized for new missions. NDA was developed in ..."
Abstract

Cited by 4 (1 self)
 Add to MetaCart
(Show Context)
The Nozzle Design Associate (NDA) is a computational environment for the design of jet engine exhaust nozzles for supersonic aircraft. NDA may be used either to design new aircraft or to design new nozzles that adapt existing aircraft so they may be reutilized for new missions. NDA was developed in a collaboration between computer scientists at Rutgers University and exhaust nozzle designers at General Electric Aircraft Engines and General Electric Corporate Research and Development. The NDA project has two principal goals: to provide a useful engineering tool for exhaust nozzle design, and to explore fundamental research issues that arise in the application of automated design optimization methods to realistic engineering problems. Introduction The Nozzle Design Associate (NDA) is a computational environment for the design of jet engine exhaust nozzles for supersonic aircraft. NDA may be used either to design exhaust nozzles for new aircraft or to design new nozzles that adapt existi...
Learning When Reformulation is Appropriate for Iterative Design
 IJCAI95 Workshop of Machine Learning in Engineering
, 1995
"... It is well known that searchspace reformulation can improve the speed and reliability of numerical optimization in engineering design. We argue that the best choice of reformulation depends on the design goal, and present a technique for automatically constructing rules that map the design goal int ..."
Abstract

Cited by 4 (4 self)
 Add to MetaCart
It is well known that searchspace reformulation can improve the speed and reliability of numerical optimization in engineering design. We argue that the best choice of reformulation depends on the design goal, and present a technique for automatically constructing rules that map the design goal into a reformulation chosen from a space of possible reformulations. We tested our technique in the domain of racingyachthull design, where each reformulation corresponds to incorporating constraints into the search space. We applied a standard inductivelearning algorithm, C4.5, to a set of training data describing which constraints are active in the optimal design for each goal encountered in a previous design session. We then used these rules to choose an appropriate reformulation for each of a set of test cases. Our experimental results show that using these reformulations improves both the speed and the reliability of design optimization, outperforming competing methods and approaching t...
An Innovative Steiner Tree Based Approach for Polygon Partitioning
, 2008
"... As device technology continues to scale past 65nm, the heavy application of resolution enhancement techniques (RET) makes the complexity, run time and quality issues in mask data preparation (MDP) grow severely. As one major and core step in MDP, polygon partitioning converts the complex layout sha ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
As device technology continues to scale past 65nm, the heavy application of resolution enhancement techniques (RET) makes the complexity, run time and quality issues in mask data preparation (MDP) grow severely. As one major and core step in MDP, polygon partitioning converts the complex layout shapes into trapezoids suitable for mask writing. The partitioning run time and quality of the resulting polygon partitions directly impacts the cost, integrity, and quality of the written mask. In this work, we introduce an innovative approach to solve the polygon partition problem by constructing a variant Steiner minimal tree: minimal partition tree (MPT). We prove the equivalence between MPT and the optimal polygon partition. Also, the solution search space for MPT is further reduced for the efficiency of the MPT algorithms. Finally, a generic MPT algorithm flow and a lineartime heuristic algorithm based on it are proposed. Experiments show that MPT solves the polygon partitioning with very promising and high quality results.
Using Feedback To Improve VLSI Designs
, 1995
"... VLSI HLS design is a typical design optimization problem with a focus on generating good solutions. Typical expert design systems incorporate a large amount of domain knowledge to generate good initial solutions. These systems are unable to use information gleaned from analysis of the solutions (fee ..."
Abstract
 Add to MetaCart
(Show Context)
VLSI HLS design is a typical design optimization problem with a focus on generating good solutions. Typical expert design systems incorporate a large amount of domain knowledge to generate good initial solutions. These systems are unable to use information gleaned from analysis of the solutions (feedback)to generate better solutions. This paper describes a new technique called ConstrainedRedo that uses feedback to improve both the power and coverage of an existing design system, the DAA system. 1 Introduction The VLSI High Level Synthesis (HLS) problem consists of creating a register level transfer (RTL) description of a chip that satisfies a given behavioral specification, while simultaneously maximizing the speed of the chip and minimizing area usage. The input to an HLS system is an algorithmic (functional) description in a high level language (e.g., VHDL) of a chip, either a microprocessor or a digital filter. The output is a register transfer level (RTL) description of the chip ...