Hybrid surrogate-model-based multi-fidelity efficient global optimization applied to helicopter blade design

A multi-fidelity optimization technique by an efficient global optimization process using a hybrid surrogate model is investigated for solving real-world design problems. The model constructs the local deviation using the kriging method and the global model using a radial basis function. The expected improvement is computed to decide additional samples that can improve the model. The approach was first investigated by solving mathematical test problems. The results were compared with optimization results from an ordinary kriging method and a co-kriging method, and the proposed method produced the best solution. The proposed method was also applied to aerodynamic design optimization of helicopter blades to obtain the maximum blade efficiency. The optimal shape obtained by the proposed method achieved performance almost equivalent to that obtained using the high-fidelity, evaluation-based single-fidelity optimization. Comparing all three methods, the proposed method required the lowest total number of high-fidelity evaluation runs to obtain a converged solution.     

Efficient global optimization applied to wind tunnel evaluation-based optimization for improvement of flow control by plasma actuators

A kriging-based genetic algorithm called efficient global optimization (EGO) was employed to optimize the parameters for the operating conditions of plasma actuators. The aerodynamic performance was evaluated by wind tunnel testing to overcome the disadvantages of time-consuming numerical simulations. The proposed system was used on two design problems to design the power supply for a plasma actuator. The first case was the drag minimization problem around a semicircular cylinder. In this case, the inhibitory effect of flow separation was also observed. The second case was the lift maximization problem around a circular cylinder. This case was similar to the aerofoil design, because the circular cylinder has potential to work as an aerofoil owing to the control of the flow circulation by the plasma actuators with four design parameters. In this case, applicability to the multi-variant design problem was also investigated. Based on these results, optimum designs and global design information were obtained while drastically reducing the number of experiments required compared to a full factorial experiment.     

Experiment and computation multi-fidelity multi-objective airfoil design optimization of fixed-wing UAV

Journal of Mechanical Science and Technology - Optimization with a surrogate model is a popular method used in design to avoid a high computational cost problem and is often encountered in...     

Distribution of epithelial-specific antigen in uterine cervix with endocervical glandular dysplasia

The human orosomucoid (ORM) is controlled by two closely linked loci, ORM1 and ORM2, and two tandem genes, AGP1 and AGP2, encoding the proteins produced by the two loci, have been cloned. In this study the molecular basis of ORM1 polymorphism was investigated. For the detection of mutations the products of the six exons of each gene, amplified by the polymerase chain reaction (PCR), were screened by single-strand conformation polymorphism analysis. Subsequently, the exons with an altered migration pattern were gene-specifically amplified by nested PCR. Sequencing of the gene-specific PCR products showed that the three common ORM1 alleles result from A-->G transitions at the codons for amino acid positions 20 in exon 1 and 156 in exon 5 of the AGP1 gene: ORM1*F1 was characterized by CAG (Gln) and GTG (Val), ORM1*F2, by CAG (Gln) and ATG (Met), and ORM1*S, by CGG (Arg) and GTG (Val). The phylogenesis of the genes encoding these three ORM1 alleles is discussed.     

Solid-liquid interfacial energy on crystallization in Fe-Si-B alloy glass

Isothermal crystallization on ageing is investigated in Fe-10at% Si-15at% B alloy glass. Based on homogeneous nucleation theory, the solid-liquid interfacial energy is estimated by use of the nucleation frequency of crystallization. The energy obtained is quite close to that estimated by extending Turnbull's assumption for alloys.     

Prediction of Aircraft＇s Longitudinal Motion Based on Aerodynamic Coefficients and Derivatives by Surrogate Model Approach

The accuracy of a flight simulation is highly dependent on the quality of the aerodynamic database and prediction accuracies of the aerodynamic coefficients and derivatives. A surrogate model is an approximation method that is used to predict unknown functions based on the sampling data obtained by the design of experiments. This model can also be used to predict aerodynamic coefficients/derivatives using several measured points. The objective of this paper is to develop an efficient digital flight simulation by solving the equation of motion to predict the aerodynamics data using a surrogate model. Accordingly, there is a need to construct and investigate aerodynamic databases and compare the accuracy of the surrogate model with the exact solution, and hence solve the equation of motion for the flight simulation analysis. In this study, sample datas for models are acquired from the USAF Stability and Control DATCOM, and a database is constructed for two input variables （the angle of attack and Mach number）, along with two derivatives of the X-force axis and three derivatives for the Z-force axis and pitching moment. Furthermore, a comparison of the value predicted by the Kriging model and the exact solution shows that its flight analysis prediction ability makes it possible to use the surrogate model in future analyses.