Surrogate-based Analysis and Optimization

https://eprints.soton.ac.uk/65935/01/Forr_09.pdf

Recent advances in surrogate-based optimization

Evolutionary algorithms often have to solve optimization problems in the presence of a wide range of uncertainties. Generally, uncertainties in evolutionary computation can be divided into the following four categories. First, the fitness function is noisy. Second, the design variables and/or the environmental parameters may change after optimization, and the quality of the obtained optimal solution should be robust against environmental changes or deviations from the optimal point. Third, the fitness function is approximated, which means that the fitness function suffers from approximation errors. Fourth, the optimum of the problem to be solved changes over time and, thus, the optimizer should be able to track the optimum continuously. In all these cases, additional measures must be taken so that evolutionary algorithms are still able to work satisfactorily. This paper attempts to provide a comprehensive overview of the related work within a unified framework, which has been scattered in a variety of research areas. Existing approaches to addressing different uncertainties are presented and discussed, and the relationship between the different categories of uncertainties are investigated. Finally, topics for future research are suggested.

/pdf/evolutionary-optimization-in-uncertain-environments-a-survey-1cgcnwsuec.pdf

Evolutionary optimization in uncertain environments-a survey

Evolutionary algorithms (EAs) have received increasing interests both in the academy and industry. One main difficulty in applying EAs to real-world applications is that EAs usually need a large number of fitness evaluations before a satisfying result can be obtained. However, fitness evaluations are not always straightforward in many real-world applications. Either an explicit fitness function does not exist, or the evaluation of the fitness is computationally very expensive. In both cases, it is necessary to estimate the fitness function by constructing an approximate model. In this paper, a comprehensive survey of the research on fitness approximation in evolutionary computation is presented. Main issues like approximation levels, approximate model management schemes, model construction techniques are reviewed. To conclude, open questions and interesting issues in the field are discussed.

https://epubs.surrey.ac.uk/7610/2/SC2005.pdf

A comprehensive survey of fitness approximation in evolutionary computation

Recent progress in flapping wing aerodynamics and aeroelasticity

A sensitivity analysis is done for turbulent cavitating flows using a pressure -based Navier-Stokes solver coupled with a phase volume fraction transport model and non-equilibrium k-e turbulence closure. Four modeling parameters are adopted for evaluation, namely, Ce1 and Ce2, which directly influences the production and dissipation of turbulence kinetic energy, and Cdest and Cprod, which regulate the evaporation and condensation of the phases. Response surface methodology along with design of experiments is used for the sensitivity studies. The difference between the computational and experimental results is used to judge the model fidelity. Under non-cavitating conditions, the best selections of Ce1 and Ce2, exhibit a linear combination with multiple optima. Using this information, cavitating flows around an axi -symmetric geometry with a hemispherical fore-body and the NACA66(MOD) airfoil are assessed. Analysis of the cavitating model shows that the favorable combinations of Cdest and Cprod, are inversely proportional to each other for the geometries considered. A set of cavitation numbers is selected for each of the geometries to demonstrate the predictive capability of the present modeling approach for attached, turbulent cavitating flows.

Sensitivity Evaluation of a Transport-Based Turbulent Cavitation Model

This paper presents sensitivity and trade-off analyses for the design of a gaseous injector used for liquid rocket propulsion whose geometry (hydrogen flow angle, hydrogen and oxygen flow areas and oxygen post tip thickness) is sought to optimize performance (combustion length) and life/survivability (temperatures at three different locations). The analyses are conducted using data available from surrogate models of the design objectives, and the multi-objective optima (Pareto optimal front, POF) generated with the aid of multi-objective genetic algorithm and local search method. The trade-offs among objectives are estimated in different clusters (identified through a hierarchical clustering algorithm) along the POF. Sensitivity analyses are conducted using a variance-based non-parametric approach over the whole design space as well as on the clusters at the POF. The former analysis help identify the contribution of the design variables to the objective variability. The latter analysis highlights: the variability of the design variables and the objectives within a cluster of the POF (using box plots), and the relationship between the design variables and the objectives (using partial correlation coefficients). Additionally, the trade-offs between selected pairs of objectives are also analyzed.

Global Sensitivity and Trade-Off Analyses for Multi-Objective Liquid Rocket Injector Design

Neural Network-Based and Response Surface-Based Optimization Strategies for Rocket Engine Component Design

An injector optimization methodology, method i, is used to investigate optimal design points for gaseous oxygen/gaseous hydrogen (GO2/GH2) injector elements. A swirl coaxial element and an unlike impinging element (a fuel-oxidizer-fuel triplet) are used to facilitate the study. The elements are optimized in terms of design variables such as fuel pressure drop, APfi oxidizer pressure drop, ADo, combustor length, Lco,,b, and full cone swirl angle, 8, (for the swirl element) or impingement half-angle, a, (for the impinging element) at a given mixture ratio and chamber pressure. Dependent variables such as energy release efficiency, ERE, wall heat flux, Qw, injector heat flux, Q,,j, relative combustor weight, W,¢t, and relative injector cost, Cr¢l, are calculated and then correlated with the design variables. An empirical design methodology is used to generate these responses for both element types. Method i is then used to generate response surfaces for each dependent variable for both types of elements. Desirability functions based on dependent variable constraints are created and used to facilitate development of composite response surfaces representing the five dependent variables in terms of the input variables. Three examples illustrating the utility and flexibility of method i are discussed in detail for each element type. First, joint response surfaces are constructed by sequentially adding dependent variables. Optimum designs are identified after addition of each variable and the effect each variable has on the element design is illustrated. This stepwise demonstration also highlights the importance of including variables such as weight and cost early in the design process. Secondly, using the composite response surface that includes all five dependent variables, unequal weights are assigned to emphasize certain variables relative to others. Here, method i is used to enable objective trade studies on design issues such as component life and thrust to weight ratio. Finally, combining results from both elements to simulate a trade study, thrustto-weight trends are illustrated and examined in detail.

/pdf/an-optimization-based-approach-to-33vaeqp9y6.pdf

Rajkumar Vaidyanathan

Papers

Sensitivity Evaluation of a Transport-Based Turbulent Cavitation Model

Global Sensitivity and Trade-Off Analyses for Multi-Objective Liquid Rocket Injector Design

Neural Network-Based and Response Surface-Based Optimization Strategies for Rocket Engine Component Design

An Optimization-Based Approach To

CFD code verification using least square extrapolation method