With the development of machine learning technology, number algorithms grows rapidly and models become more complex. That causes two major problems in practice: selection hyperparameter optimization. In order to tackle these issues, this paper proposes a new method based on deep reinforcement learning. Long short-term memory (LSTM) network is used build an agent which automatically selects model optimizes hyperparameters for given dataset. The aims maximize accuracy selected validation At...

Artificial neural network (ANN) is a kind of bionic technology which stimulates the information processing mechanism transmission in human cerebral cortex, using this we can process complex non-linear system with multiple influencing factors, such as concrete system. Based on this, paper proposes prediction model high strength based radial basis artificial network. Through empirical analysis, it found that ANN has fast training speed, good effect, and accuracy. The accuracy RBF better than...

The development of renewable energy becomes increasingly important because exhaustion fossil energy. Hydropower is one the most ways to generate electricity from its relatively stable output among them. However, hydropower project with some inherent characteristics highly susceptible social and natural environments, which complicates investment process. For this purpose, paper proposes a feasible comprehensive optimization model portfolio perspective sustainable development, describing...

Hyperparameter optimization is a key part of AutoML. In recent years, there have been successful hyperparameter algorithms. However, these methods still face several challenges, such as high cost evaluating large models or datasets. this paper, we introduce new deep reinforcement learning architecture with model-based acceleration to optimize hyperparameters for any machine model. method, an agent constructed by Long Short-Term Memory Network aims at maximizing the expected accuracy model on...

We introduce a new deep reinforcement learning architecture - RPR-BP to optimize hyperparameter for any machine model on given data set. In this method, an agent constructed by Long Short-Term Memory Network aims at maximizing the expected accuracy of validation At each iteration, it selects set hyperparameters and uses as reward signal update its internal parameters. After multiple iterations, learns how improve decisions. However, computation requires significant time leads low sample...

The present work couples isogeometric analysis (IGA) and boundary element methods (BEM) for three dimensional steady heat conduction problems with variable coefficients. Computer-Aided Design (CAD) geometries are built by subdivision surfaces, meantime the basis functions of surfaces employed to discretize integral equations analysis. Moreover, radial integration method is adopted transform additional domain integrals caused coefficients integrals. Several numerical examples provided...

This work proposes a refreshing technique that utilizes the Taylor expansion to improve computational efficiency of multi-frequency acoustic scattering problem.The Helmholtz equation in problems is solved using boundary element method (BEM).In this work, utilized separate frequency-dependent terms from integrand function integral so wave number independent system, thereby avoiding time-consuming frequency sweep analysis.To conquer non-uniqueness solution for external field problem,...

A computer simulation approach known as the isogeometric (IGA) method may directly use surface information of geometric model. In 3D graphics, Loop subdivision surfaces are a common for creating complicated shapes. this study, we propose coupling algorithm that utilizes and direct differentiation computation acoustic-fluid-structure interaction performance structural-acoustic sensitivity analysis. This combines finite element (FEM) wideband fast multipole boundary (FMBEM). Because proposed...

Due to the intermittency of renewable energies, their acceptances in electricity market are lower than fossil energies. Therefore, it is necessary overcome development resistance through portfolio re-equilibrium. Balance supply and demand markets involving different stakeholders optimize return on investment. Compared with past, this paper proposes an integrated optimization combined meta-heuristic for wind-solar-hydro system. It investigates coordinate interactions power stations grid...

In many engineering challenges, the whole interaction between structural domain and acoustic must be taken into account, particularly for analysis of thin structures submerged in water. The fast multipole boundary element approach is used this work to simulate external finite method describe components. To improve coupling accuracy, discontinuous higher-order components are created domain. isogeometric (IGABEM) discretizes unknown physical fields by using CAD spline functions as basis...

Since the acoustic impedance in water cannot be neglected with respect to mechanical impedance, radiation caused by vibration of structures compressible fluid would react structure.Therefore, fluid-structure interaction needs considered.The finite element method is used for structure analysis and boundary analysis.Sound absorption materials are reduce scattering sound field reference region.The sensitivity a fully coupled structural-acoustic system proposed.Numerical tests verify correctness...

With rising power demand and stringent carbon emission regulations, renewable energy is gaining traction in the grid. However, its acceptance lower than that of fossil for inherent intermittent. This study motivated by such opposition seeks to overcome it enabling electricity consumption through coordinated operation strategies hydro-wind-photovoltaic system. To match maximize system output, non-priority output scheduling should be used. presents a methodical nonlinear programming strategy...

This paper presents a stochastic analysis method for linear elastic fracture mechanics using the Monte Carlo simulations (MCs) and scaled boundary finite element (SBFEM) based on proper orthogonal decomposition (POD) radial basis functions (RBF). The semianalytical solutions obtained by SBFEM enable us to capture stress intensity factors (SIFs) easily accurately. adoption of POD RBF significantly reduces model order increases computation efficiency, while maintaining versatility accuracy...