Large Number Of Geometrical Parameters Research Articles

On demand design of multi-freedom metamaterial absorbers based on recurrent prediction networks

Metamaterial absorbers (MMAs) have received a lot of attention due to their wide range of promising applications. In general, metamaterials involve a large number of geometric parameters, so the traditional simulation requires researchers to have rich experience and sufficient computational resources. In this paper, we demonstrate the forward and on demand design of multi-freedom MMAs based on the recurrent prediction network (RPN). We used the variable contribution analysis (VCA) algorithm to eliminate 55 absorption points whose contribution score was less than 1.0, the maximum fitting coefficient (R2) was increased by 7.0%, and the average calculation time was reduced by 20.326s. multi-freedom design is achieved by quantifying the design material into binary features as geometric parameter inputs into the RPN model. We solve the prediction peak error problem by many-to-one RPN, compared with the many-to-many prediction of classical deep neural network (DNN) model, the RPN error range is 0-0.12 compared to the 0-0.25 error range of the DNN. The ultra-wideband MMA design with the absorption of 0.2-4.2µm broadband is more than 90%, and the highest absorption rate is 99.2%. This method can be used in zoom imaging, metamaterials filters and other fields.

Optimizing broadband metamaterial absorber using deep reinforcement learning

Optimization of the geometry of broadband metamaterial absorbers is crucial for improving the performance of optoelectronic devices. However, a large number of geometric parameters should be considered to achieve broad absorption, which is time-consuming. Herein, we propose a rapid and simple method for optimizing metamaterial absorbers dedicated to thermal radiation absorption using deep reinforcement learning. Deep reinforcement learning generated an ideal geometry for a broadband metamaterial absorber after 4 h, demonstrating the effectiveness of this technique for the rapid and effective optimization of metamaterial absorbers.

Optimized Design of Plasma Metamaterial Absorber Based on Machine Learning

Metamaterial absorbers have become a popular research direction due to their broad application prospects, such as in radar, infrared imaging, and solar cell fields. Usually, nanostructured metamaterials are associated with a large number of geometric parameters, and traditional simulation designs are time consuming. In this paper, we propose a framework for designing plasma metamaterial absorbers in both a forward prediction and inverse design composed of a primary prediction network (PPN) and an auxiliary prediction network (APN). The framework can build the relationship between the geometric parameters of metamaterials and their optical response (reflection spectra, absorption spectra) from a large number of training samples, thus solving the problem of time-consuming and case-by-case numerical simulations in traditional metamaterial design. This framework can not only improve forward prediction more accurately and efficiently but also inverse design metamaterial absorbers from a given required optical response. It was verified that it is also applicable to absorbers of different structures and materials. Our results show that it can be used in metamaterial absorbers, chiral metamaterials, metamaterial filters, and other fields.

A holistic optimization tool for the design of power take-off systems in realistic wave climates: The Wells turbine case

The growing demand of energy coupled with the worldwide commitments in favor of renewable technologies turns marine renewable sources especially attractive, including wave energy. In particular, the oscillating water column technology is shown to be one of the most promising technologies, for which the design of efficient air turbines is crucial. The large number of geometrical parameters that such turbines own, together with the fact that the wave climate conditions affect the overall behavior, call for a tool that combines both aspects for optimizing the turbine design. A genetic-algorithm-based approach is proposed, which runs over a blade-element-method coupled with a stochastic modeling of the sea-state. The optimization method computes an overall fitness function given as a weighted average of the stochastic turbine efficiency over a set of most relevant sea-states. The suggested methodology is proven capable of finding the global fitness maximum upon a 6-dimensional parametric space. Additionally, it allows identifying and discarding the geometrical parameters that are shown to be negligible for the computation of the fitness function. Hence, a reduced parametric space results in a 2-dimensional region comprised by the hub-to-tip ratio and the tip solidity, which permits refining the search and analyzing their individual effects upon the turbine design. It is shown that the optimal configuration corresponds to trade-off solutions between the hub-to-tip ratio and the tip solidity, avoiding both an excessive rotational speed and an exceeding blockage of the turbine.

Electromagnetic design and optimization of the multi-segment dielectric-loaded accelerating tube using genetic algorithm

A low-energy dielectric loaded accelerator with a non-uniform, multi-segment structure is studied and optimized. So far, no analytical solution is provided for such structures. Also, due to the existing nonlinear behavior and a large number of geometric parameters, the problem of numerical optimizations is complex. For this reason, a method is presented to design and optimize such structures using the Genetic Algorithm (GA). Moreover, the GA output results are compared with Trust Region (TR) and Nelder-Mead Simplex (NMS) methods. Comparative results show that the GA is more efficient in achieving optimization goals and also has a higher speed than the two other methods. Finally, an optimized accelerating tube is integrated into a proper coupler. Then, the accelerator is simulated for full electromagnetic investigations using the CST suite of codes. This design leads to a structure with a power of about 80 kW in the X-band, which delivers electrons to the output energy in the range of 300–459 kV. The length and outer diameter of the accelerating tube obtained are 10 cm and 1 cm, respectively.

Neural network enabled metasurface design for phase manipulation.

The phase of electromagnetic waves can be manipulated and tailored by artificial metasurfaces, which can lead to ultra-compact, high-performance metalens, holographic and imaging devices etc. Usually, nanostructured metasurfaces are associated with a large number of geometric parameters, and the multi-parameter optimization for phase design cannot be possibly achieved by conventional time-consuming simulations. Deep learning tools capable of acquiring the relationship between complex nanostructure geometry and electromagnetic responses are best suited for such challenging task. In this work, by innovations in the training methods, we demonstrate that deep neural network can handle six geometric parameters for accurately predicting the phase value, and for the first time, perform direct inverse design of metasurfaces for on-demand phase requirement. In order to satisfy the achromatic metalens design requirements, we also demonstrate simultaneous phase and group delay prediction for near-zero group delay dispersion. Our results suggest significantly improved design capability of complex metasurfaces with the aid of deep learning tools.

Hierarchically structured metamaterials with simultaneously negative mass density and Young’s modulus by using dynamic homogenization

We propose one-dimensional hierarchical metamaterials (HMMs) that have simultaneously negative mass density and Young’s modulus in the sense of effective properties. Frequency-dependent effective properties of the HMMs are evaluated by using a dynamic homogenization theory. By considering three necessary conditions for dynamic homogenizability, we verify that our proposed HMMs can have physically meaningful double negativity, whereas the conventional periodic structures cannot. In addition, HMMs yield the broader ranges of the frequency and the effective properties for double negativity due to the large number of geometrical parameters. By exploring all the possible effective properties of HMMs, we discuss whether we can design metamaterials with the opposite sign of material properties in nature, ultimately aiming at perfect impedance matching.

Numerical approach based design of centrifugal pump volute

Centrifugal pumps are widely used in many industrial sectors. However the experimental characterisation processus of pumps is very complex and expensive task due to the large number of geometrical parameters in consideration. The complexity is increased due to their elaboration by molding process dicted by the pump components (volute, wheel, and rotor). It is then worthful to minimizing the costs associated to design and manufacturing processes of pumps. Ever though the tremendous efforts made in this domain the experimentation does not represent the most reliable solution. A precise numerical simulation of complex liquid flow in centrifugal pump can be rapid and reliable solution for improving the pump performance. A large number of numerical or experimental researches are focussed on rotor dimensurring of radial volute. In this study three different forms of tangential volutes (circular, Bezier and trapezoidal) with same dimensions and the same rotor are used. The behaviour of the fluid either Newtonian or Non-Newtonian is taken into consideration. The objective of the study is to know the effect of the form of the volute and nature of the fluid on the properties of the internal fluid flow. Finite volume method (FVM) is used under the commercial code Ansys-CFX. The numerical simulation is carried out to study the velocity distribution and pressure field distribution in the three volutes. The nature of Non-Newtonian fluid behaviour is integrated in the code via a subprogram. The turbulence of flows in the centrifugal pump is taken into account using k-Ɛ model. The obtain results show that the form of the volute and the nature of fluid flow have a sensitive effects on the velocity, the pressure as well as on the shear stress. DOI: http://dx.doi.org/10.5755/j01.mech.21.4.8930

Classification of airborne laser scanning point clouds based on binomial logistic regression analysis

This article presents a newly developed procedure for the classification of airborne laser scanning (ALS) point clouds, based on binomial logistic regression analysis. By using a feature space containing a large number of adaptable geometrical parameters, this new procedure can be applied to point clouds covering different types of topography and variable point densities. Besides, the procedure can be adapted to different user requirements. A binomial logistic model is estimated for all a priori defined classes, using a training set of manually classified points. For each point, a value is calculated defining the probability that this point belongs to a certain class. The class with the highest probability will be used for the final point classification. Besides, the use of statistical methods enables a thorough model evaluation by the implementation of well-founded inference criteria. If necessary, the interpretation of these inference analyses also enables the possible definition of more sub-classes. The use of a large number of geometrical parameters is an important advantage of this procedure in comparison with current classification algorithms. It allows more user modifications for the large variety of types of ALS point clouds, while still achieving comparable classification results. It is indeed possible to evaluate parameters as degrees of freedom and remove or add parameters as a function of the type of study area. The performance of this procedure is successfully demonstrated by classifying two different ALS point sets from an urban and a rural area. Moreover, the potential of the proposed classification procedure is explored for terrestrial data.

Electron-Diffraction Investigation of the Fluorofullerene C60F48

The structure of the fullerene C60F48 has been investigated in the gas phase by electron diffraction from a sample volatilized at 360 °C. The analysis was carried out under two assumptions: (1) the molecules have either D3 or S6 symmetry as suggested by NMR spectroscopy and verified by an X-ray study of the crystal, and only one of these is present in the gas; (2) all carbon−fluorine bonds have the same length. With the named symmetries, the structure of the carbon skeleton may be defined by the positions of 10 atoms forming two pentagons, one near the top of the molecule and one near the equator, and the locations of the fluorine atoms obtained as the resultant of three vectors originating from carbons not involving a double bond. Simultaneous refinement of the large number of geometrical parameters (30 for the carbon skeleton and 17 for the fluorines) either failed to converge or yielded implausible values, but successive refinements of small groups of four or five parameters were successful. Dozens of groups were tested and all of the resulting models gave satisfactory fits to the observed diffraction patterns. Although values of individual parameters in these models might differ appreciably, the values obtained as averages from the many refinements have good precision. Some of these averaged results (ra/Å, ∠/deg) for the D3/S6 models, with estimated standard deviations, are the following: r(C−F) = 1.368(1)/1.368(1); r(CC) = 1.327(3)/1.326(4); r(Csp2−Csp3) = 1.503(15)/1.500(11); r(Csp3−Csp3) = 1.585(44)/1.585(41); ∠(C−CC) = 113.7(4)/113.6(4) and ∠(C−C−C) = 105.5(1)/105.5(2) within pentagons; and ∠(C−CC) = 124.2(3)/124.0(4) and ∠(C−C−C) = 116.6(3)/116.5(3) within hexagons. The average distances from the center of the cage (spherical radii) are quite different for the three types of carbon atoms (those in a double bond, those adjacent to a double bond, and those not adjacent to a double bond) and quite different from the C60 value of 3.555Å for all atoms. For symmetries D3/S6 these radii (R/Å) are 3.937(23)/3.937(17) for sp3 atoms not bonded to sp2 ones and 3.781(18)/3.778(20) for sp3 atoms bonded to sp2 ones. The average radii to the sp2 atoms are much shorter than those to the other atoms. These radii fall into two groups for each symmetry: for symmetry D3 they are 3.018(14) and 3.190(15) Å, and for S6, 3.017(11) and 3.180(15) Å. The surprising length of some of the carbon−carbon bonds and other features of the structures relative to the structure of C60 are discussed.

Experimental observations on the general trends of the steady state and stability behaviour of single-phase natural circulation loops

The steady state flow in uniform or non-uniform diameter single-phase natural circulation loops (NCLs) can be expressed as Re ss= C[ Gr m/ N G] r where the constants C and r depend on the nature of the flow (i.e. laminar or turbulent). Testing these correlations with experimental data showed reasonable agreement. The above equation suggests that simulation of the steady state flow in single-phase NCLs can be achieved by simulating the non-dimensional parameter ( Gr m/ N G). The stability behaviour of a non-uniform diameter loop is found to depend on a large number of geometric parameters in addition to the modified Stanton number ( St m) and the ratio ( Gr m) b/(3− b) /( N G) 3/(3− b) . The transient and stability behaviour of a uniform diameter loop can be simulated by simulating the product ( Gr m) b/(3− b) ×( D ̄ / L ̄ t ) 3/(3− b) and St m if the length scales are preserved.

Surface structure analysis of the domain-wall phase of S/Ru(0001) using an efficient parameter optimization method

A structural analysis of the striped domain wall phase of S/Ru(0001) is performed using LEED-IV techniques. For handling the large number of geometrical parameters, a new technique for multi-dimensional data optimization designed for the analysis of LEED-IV data was developed. It is based on the Newton-Raphson method, which includes all parameter couplings to first order in the fit. However, instead of using first- and second-order derivatives to evaluate the required quadratic expansion of the R-factor hypersurface, we use a weighted least-squares fit to adapt a paraboloid to the R-factor data (PLEED). The structural analysis, which requires the determination of 27 geometrical parameters, reveals that the local structure of the domain walls resembles that of the c(2 × 4) structure, but additional relaxations have been found which can be identified as the origin of the effective lateral repulsion between the domain walls. The main mechanisms of local structure formation in this system are discussed. The formation of a strong local bond is modified by lateral relaxations due to effective repulsive interactions. They involve tilting of the tetrahedron formed by the S atom and its three nearest neighbors and lateral shifts.