Genetic algorithms (GAs) have been extensively used as a means for performing global optimization in a simple yet reliable manner. However, in some realistic engineering design optimization domains a general purpose GA is often inefficient and unable to reach the global optimum. In this thesis we describe a GA for continuous designspace optimization that uses new GA operators and strategies tailored to the structure and properties of engineering design domains. Empirical results in several realistic engineering design domains as well as benchmark design domains demonstrate that using our system can greatly decrease the cost of design space search, and can also improve the quality of the resulting designs.

The interdisciplinary nature of complex systems design presents challenges associated with computational burdens and organizational barriers as these issues cannot be resolved with faster computers and more efficient optimization algorithms. There is a need to develop design methods that can model different degrees of collaboration and help resolve the conflicts between different disciplines. In this paper, an approach to providing flexibility in resolving the conflicts between the interests of multiple disciplines is proposed. We propose to integrate the robust design concept into game theory protocols, in particular the Stackelberg leader/follower protocol. Specifically, the solution for the design parameters which involve the coupled information between multiple players (disciplines) is developed as a range of solutions rather than a single point solution. This additional flexibility provides more freedom to the discipline that takes the role of follower, while also keeping the perf...

Optimal design of large engineering systems modeled as nonlinear programming problems remains a challenge because increased size reduces reliability and speed of numerical optimization algorithms. Decomposition of the original model into smaller coordinated submodels is desirable or even necessary. The article presents a methodology for optimal model-based decomposition of design problems, whether or not initially cast as optimization models. The overall model is represented by a hypergraph that is optimally partitioned into weakly-connected subgraphs satisfying partitioning constraints. The formulation is robust enough to account for computational demands and resources, and the strength of interdependencies between the design relations contained in the model. This decomposition methodology is applied to a vehicle powertrain system design model consisting of engine, torque converter, transmission, and wheel-tire assemblies, with 87 design relations and 119 design and state/behavior variables.

Collaborative optimization is a new design architecture specifically created for large-scale distributed-analysis applications. In this approach, a problem is decomposed into a user-defined number of subspace optimization problems that are driven towards interdisciplinary compatibility and the appropriate solution by a system-level coordination process. This decentralized design strategy allows domain-specific issues to be accommodated by disciplinary analysts, while requiring interdisciplinary decisions to be reached by consensus. The present investigation focuses on application of the collaborative optimization architecture to the multidisciplinary design of a single-stage-to-orbit launch vehicle. Vehicle design, trajectory, and cost issues are directly modeled. Posed to suit the collaborative architecture, the design problem is characterized by 95 design variables and 16 constraints. Numerous collaborative solutions are obtained. Comparison of these solutions demonstrates the influe...

Design engineers often employ numerical optimization techniques to assist in the evaluation and comparison of new aircraft configurations. While the use of numerical optimization methods is largely successful, the presence of numerical noise in realistic engineering optimization problems often inhibits the use of many gradient-based optimization techniques. Numerical noise causes inaccurate gradient calculations which in turn slows or prevents convergence during optimization. The problems created by numerical noise are particularly acute in aircraft design applications where a single aerodynamic or structural analysis of a realistic aircraft configuration may require tens of CPU hours on a supercomputer. The computational expense of the analyses coupled with the convergence difficulties created by numerical noise are significant obstacles to performing aircraft multidisciplinary design optimization. To address these issues, a procedure has been developed to create two types of noise-free mathematical models for use in aircraft optimization studies. These two methods use elements of statistical analysis and the overall procedure for using the methods is made computationally affordable by the application of parallel computing techniques. The first

... In this paper, an integrated methodology for propagating and mitigating the effect of uncertainties is proposed. Two approaches, namely, the extreme condition approach and the statistical approach, are developed to propagate the effect of uncertainties across a design system comprising interrelated subsystem analyses. Using the extreme condition approach, an interval of the output from a chain of simulations is obtained, while the statistical approach provides statistical estimates of the output. An uncertainty mitigation strategy based on the principles of robust design is proposed. The methodology is presented using an illustrative simulation chain and is verified using the case study of a six-link function-generator linkage design.

Several government/commercial off-the-shelf modular software packages have been combined to perform static aeroelastic analysis and aerodynamic optimization of a generic high-speed civil transport, with the eventual goal of performing high-fidelity multidisciplinary optimization (MDO). Results obtained for the aeroelastic analysis and aerodynamic optimization cases are consistent with expectations and serve as a benchmark for future results that will incorporate sensitivity derivatives from the computational fluids and structures solvers to reduce the wall-clock computational expense of high-fidelity MDO. 1 Introduction Multidisciplinary design optimization (MDO) methods developed in the past decade have not been widely adopted by the U.S. aerospace industry primarily because of the prohibitive computational expense incurred when MDO methods are applied to realistic aircraft design problems. Such MDO applications typically involve tens to hundreds of variables and hundreds of constrai...

Gradient-based 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 intelligence-based 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.

Rapid turn-around time for investigating new design concepts is a primary force driving design productivity initiatives across the industry. An integration framework focusing on the collaborative nature of rapid design automation at the preliminary and detailed design stage would ensure higher quality designs from the beginning of the product design cycle. As a result producing reliable, robust optimum designs from the preliminary design phase would enable companies to reduce the overall design cycle time.

Robust design has been gaining wide attention, and its applications have been extended to making reliable decisions when designing complex engineering systems in a multidisciplinary design environment. Though the usefulness of robust design is widely acknowledged for multidisciplinary design systems, its implementation is rare. One of the reasons is due to the complexity and computational burden associated with the evaluation of performance variations caused by the randomness (uncertainty) of a system. In this paper, a multidisciplinary robust design procedure that utilizes efficient methods for uncertainty analysis is developed. Different from the existing uncertainty analysis techniques, our proposed techniques bring the features of MDO framework into consideration. The system uncertainty analysis (SUA) method and the concurrent subsystem uncertainty analysis (CSSUA) method are developed to estimate the mean and variance of system performance subject to uncertainties associated with both design parameters and design models. As shown both analytically and empirically, compared to the conventional Monte Carlo simulation approach, the proposed techniques used for uncertainty analysis will significantly reduce the amount of design evaluations at the system level, and therefore improve the efficiency of robust design in the domain of MDO. A mathematical example and an electronic packaging problem are used as examples to verify the effectiveness of these approaches.