Optimal Model Structure Research Articles

A Sustainable Early Warning System Using Rolling Forecasts Based on ANN and Golden Ratio Optimization Methods to Accurately Predict Real-Time Water Levels and Flash Flood.

Remote monitoring sensor systems play a significant role in the evaluation and minimization of natural disasters and risk. This article presents a sustainable and real-time early warning system of sensors employed in flash flood prediction by using a rolling forecast model based on Artificial Neural Network (ANN) and Golden Ratio Optimization (GROM) methods. This Early Flood Warning System (EFWS) aims to support decision makers by providing reliable and accurate information and warning about any possible flood events within an efficient lead-time to reduce any damages due to flash floods. In this work, to improve the performance of the EFWS, an ANN forecast model based on a new optimization method, GROM, is developed and compared to the traditional ANN model. Furthermore, due to the lack of literature regarding the optimal ANN structural model for forecasting the flash flood, this paper is one of the first extensive investigations into the impact of using different exogenous variables and parameters on the ANN structure. The effect of using a rolling forecast model compared to fixed model on the accuracy of the forecasts is investigated as well. The results indicate that the rolling ANN forecast model based on GROM successfully improved the model accuracy by 40% compared to the traditional ANN model and by 93.5% compared to the fixed forecast model.

Comprehensive research on energy-saving green design scheme of crane structure based on computational intelligence

In the field of cranes, unreasonable structure design leads to high energy consumption. In order to solve the problems of heavy weight and serious steel consumption of a crane structure, a green energy-saving design method based on computational intelligence is proposed. For minimizing the weight of a structure, two optimization models are proposed. The specular reflection algorithm is used to make the green and lightweight design. A multi-objective optimization model for the green design is constructed. The minimum waste generated in the manufacturing process is the objective function of this model. Fuzzy mathematics theory is utilized to comprehensively evaluate the impact of crane structure weight and processing waste on the environment, and a structural optimization model with fuzzy comprehensive evaluation indicators for the green design is introduced. The results indicate that compared with the original design, the processing waste after fuzzy comprehensive optimization is 63.43% lower and the cross-sectional area of the main girder is reduced by 27.03%.

Performance Research and Structure Optimization of Labyrinth Screw Pump.

The labyrinth screw pump is a new type of low-flow rotor pump with a simple structure and good sealing performance. It is suitable for the transport of high-viscosity, high-gas-content, and particle-containing media. In this study, a rectangular labyrinth screw pump was used as the research object. The effect of the medium viscosity on the performance of the labyrinth pump was studied through numerical simulations, and the correctness of the simulation method was verified using existing test data. The efficiency and head of the labyrinth screw pump were selected as the optimization objectives, and the pump structural parameters were selected as the optimization parameters. A structural optimization model of the labyrinth screw pump based on response surface theory was established. The structural parameters of the labyrinth pump were reasonably simplified through size correlations, and then parameter sensitivity analysis was performed to determine the important structural parameters that needed to be optimized. The OSFD (optimized space-filling design) was used to combine the optimized parameters and generate the sample space. The response surface theory was combined with a neural network prediction model and a multi-objective genetic algorithm to perform optimization calculations. The results showed that there was an interactive influence between the structural parameters of the stator and rotor of the labyrinth screw pump. Compared with the original model, the optimized model pump had an efficiency increase of 13.55% and a lift increase of 19.53% when conveying a medium with a viscosity of 133 cp.

Ore image classification based on small deep learning model: Evaluation and optimization of model depth, model structure and data size

The ore image classification technology based on deep learning is an effective way to improve the image sensor-based ore sorting classification capability. However, in practice, the image sensor-based ore sorting technique often has the problem of insufficient data, and has not systematically considered the impact of model structure and dataset size on the modeling efficiency and classification performance of deep learning. Therefore, this paper attempts to explore a more suitable small deep learning model for ore image classification by considering the model depth, model structure, and dataset size. Six Convolutional Neural Networks (CNNs) models are established with different depths based on Alex Net and VGG Net and the model structure is optimized by adding BN layer. Taking the gas-coal image dataset as case study, we systematically explore the influence of model depth, model structure, dataset size on the training process efficiency and classification accuracy. Meanwhile, the operational process of coal image classifiers is analyzed visually through the ways of Channel Visualization maps, Heatmaps, Grad-CAM map, and Guided Backpropagation maps.

Environmental taxes in ensuring national security: A structural optimization model

Environmental taxes in ensuring national security: A structural optimization model

A parameter design method for multifrequency perfect sound-absorbing metasurface with critical coupled Helmholtz resonator

The low-frequency harmonic components of urban substation noise are easy to annoy the residents. Multi-frequency perfect sound-absorbing metasurface based on the Helmholtz resonator (HR) is an alternative solution to suppress the low-frequency harmonic noise. This paper proposes an efficient design method of structural parameter for the multi-frequency perfect sound-absorbing metasurface. Taking the perfect sound absorption at the target frequency as objective and the structural parameters of HR as optimization variables, the structural parameter optimization model of multi-frequency perfect sound-absorbing metasurface is established and solved by the sequential quadratic programming algorithm. The proposed design method effectively overcomes the deterioration of sound absorption performance caused by the combined design of multiple perfect sound absorption units. Utilizing the proposed method, we designed a multi-frequency perfect sound-absorbing metasurface to absorb the four harmonic components of an urban substation noise simultaneously. The finite element simulation results and the experimental results of the physical sample indicate that the designed multi-frequency perfect sound-absorbing metasurface can satisfy critical coupling to achieve perfect sound absorption at all target frequencies.

Simultaneous Calibration of Hydrologic Model Structure and Parameters Using a Blended Model

AbstractThe advent of hydrological modeling frameworks that support multiple model structures using the same software enables both model structure and model parameters to be calibrated and assessed. To date, the identification of optimal model structure has typically been performed manually. Here, a continuous (rather than discrete) treatment of model structure is used, which enables simultaneous automatic calibration of model structure and parameters using a conventional real‐valued decision variable optimization algorithm (the dynamically dimensioned search algorithm, DDS). The method, referred to herein as blended model structure calibration (BMSC), relies upon the calculation of each hydrologic flux (e.g., for infiltration) as a weighted average of fluxes generated from multiple process algorithm options. This method is applied to 12 lumped MOPEX catchment models and compared to the calibration of 108 fixed model structures, representing all possible permutations of fixed model structures with the given process options in this study. The BMSC method consistently identified near‐optimal model structure (as evaluated using average model rank performance) at significantly lower computational cost than calibrating the collective of fixed structure models. The BMSC method also provides a useful tool in identifying dominant processes and model structures in catchments.

Моделі спостережуваного руху космічних об'єктів на основі часових рядів TLE-елементів

Timely detection of changes in the characteristics of space hardware objects during their long-term operation is one of the main tasks in the development and study of onboard systems that maintain the efficiency of their operation. This paper presents a statistical method for simulating the motion of space objects (spacecraft and used launch vehicle stages) in the class of autoregressive models. The method allows one to improve the quality of description and prediction of the motion of space objects based on simulating time series of their TLE-elements (two-line orbital element sets). The purpose of this work is to increase the accuracy of mathematical models of the observed motion of space objects in the problems of deorbit time determination, satellite collision prediction, and space debris cataloging. The paper presents a system for simulating the motion of space objects, which allows one to determine an optimal amount of learning samples in simulating time series of TLE elements, determine the order of autoregression and find an optimal model structure for each variable element, identify model parameters in conditions of unequally spaced observations, identify features of the time behavior of the root-mean-square errors of the constructed autoregressive models on the basis of dividing the initial time series of TLE-elements into successive learning intervals, and obtain predictive estimates of the values of variable elements. The proposed statistical method of space object motion simulation can be recommended to describe and predict the motion of spacecraft and used launch vehicle stages represented as time series of TLE-elements (which are publicly available and regularly updated). The application of the proposed statistical method will increase the accuracy of mathematical models of the observed motion of space objects in the problems of deorbit time determination, satellite collision prediction, and space debris cataloging.

Methods for Identifying Microbial Natural Product Compounds that Target Kinetoplastid RNA Structural Motifs by Homology and De Novo Modeled 18S rRNA.

The development of novel anti-infectives against Kinetoplastids pathogens targeting proteins is a big problem occasioned by the antigenic variation in these parasites. This is also a global concern due to the zoonosis of these parasites, as they infect both humans and animals. Therefore, we need not only to create novel antibiotics, but also to speed up the development pipeline for these antibiotics. This may be achieved by using novel drug targets for Kinetoplastids drug discovery. In this study, we focused our attention on motifs of rRNA molecules that have been created using homology modeling. The RNA is the most ambiguous biopolymer in the kinetoplatid, which carries many different functions. For instance, tRNAs, rRNAs, and mRNAs are essential for gene expression both in the pro-and eukaryotes. However, all these types of RNAs have sequences with unique 3D structures that are specific for kinetoplastids only and can be used to shut down essential biochemical processes in kinetoplastids only. All these features make RNA very potent targets for antibacterial drug development. Here, we combine in silico methods combined with both computational biology and structure prediction tools to address our hypothesis. In this study, we outline a systematic approach for identifying kinetoplastid rRNA-ligand interactions and, more specifically, techniques that can be used to identify small molecules that target particular RNA. The high-resolution optimized model structures of these kineoplastids were generated using RNA 123, where all the stereochemical conflicts were solved and energies minimized to attain the best biological qualities. The high-resolution optimized model’s structures of these kinetoplastids were generated using RNA 123 where all the stereochemical conflicts were solved and energies minimized to attain the best biological qualities. These models were further analyzed to give their docking assessment reliability. Docking strategies, virtual screening, and fishing approaches successfully recognized novel and myriad macromolecular targets for the myxobacterial natural products with high binding affinities to exploit the unmet therapeutic needs. We demonstrate a sensible exploitation of virtual screening strategies to 18S rRNA using natural products interfaced with classical maximization of their efficacy in phamacognosy strategies that are well established. Integration of these virtual screening strategies in natural products chemistry and biochemistry research will spur the development of potential interventions to these tropical neglected diseases.

An Economic Equilibrium Model for Optimizing Passenger Transport Corridor Mode Structure Based on Travel Surplus

Proper corridor configuration is the basic prerequisite for meeting demand to the maximum extent and optimally configuring transport resources. As the existing studies only pursue quantitative balance, without good understanding of the decision-making mechanism of passenger travel, they can hardly be used to guide effective infrastructure planning. The aim of this study was to put forward a new planning concept from the perspective of economics. Firstly, the decision-making mechanism of travel behavior was considered based on the demand subject, and the travel demand was classified. Next, the travel surplus of the demand subject was analyzed, and the travel decision-making criterion of maximizing the travel surplus was put forward. Then, a discrete economic equilibrium model for structural optimization of the passenger transport corridor was constructed and solved by the GlobalSearch algorithm. Finally, the effectiveness of the model and algorithm was verified through examples. The results indicate the good convergence of the algorithm. Different corridor travel demand, time value distribution and fixed cost of service mode all have great influence on the service mode configuration of the comprehensive transport passenger corridor and basically conform to the internal mechanism of supply and demand. The results show that the model and algorithm proposed in this paper are valid and can provide effective reference for the design and policy making of passenger transport corridor mode supply.

The influence of carbon-glass blade structure on the performances of the whole wind turbine

ABSTRACT The ever-increasing size of wind turbines has given rise to a need for design of the carbon-glass hybrid blades. From the perspective of extreme loads, aero-elastic characteristics and cost, this study investigates the influence of carbon-glass blade structure on the performances of the whole wind turbine. With an integrated tool for analyzing ultimate loads and deformations of wind turbine components, the structural optimization model for the carbon-glass hybrid blade was built, and the new structural blade was optimized. By using a quantitative cost model, the performance of the wind turbine equipped with the optimized blades was analyzed; it shows that the use of glass-carbon hybrid materials and the reasonable structural design can effectively reduce the loads and deformation of the blade in the edge-wise and torsional directions. The use of carbon fiber will increase the manufacturing cost of the blade, but from an overall perspective, the cost of the whole wind turbine will have a slightly reduction of 0.8%.

Research on Multi-objective Optimization Model of the Planting Structure Based on TOPSIS

The problem of planting structure optimization is a multi-objective optimization decision problem. To solve this problem, this paper proposed a bi-level multi-objective optimization model of planting structure based on TOPSIS to optimize the agricultural planting structure. The optimal planting structure could be obtained by integrating the decision from both the upper-level and lower-level in the study area. The upper-level DM maximum total planting income with the minimum of total irrigation water could be obtained and the lower-level DMs maximum planting income with the minimum planting cost in sub-region. Finally, the constructed model was applied to optimize ternary planting structure of “food, cash and feed” in a part of the “Bohai Granary” of Science and Technology Demonstration Areas in Shandong Province in China. The optimal results of planting structure were obtained, and the comparison between the optimal results and the actual results was also made. It shows that the optimized planting structure has a better performance in terms of the increase in total planting income and the decrease in total irrigation water.

Research on Weapon Equipment Storage Structure Optimization Model Based on Improved ABC Classification Method

Weapon equipment storage is an important part of weapon equipment supply support. In this paper, we analysed the related concepts of weapon equipment storage, introduced the ABC classification method, proposed the steps to construct the optimization model of the weapon equipment storage structure, studied the influencing factors of the weapon equipment storage structure, and constructed a storage structure optimization model based on the improved ABC classification. The optimal weapon equipment storage structure and strategy were obtained, and a case analysis was carried out. The research conclusions can provide reference for the practical work of weapon equipment storage.

The Impact of Internet on the Trade-Off between Investment Return and Asset Structure

The Internet has an important impact on optimizing the asset structure of enterprises. Restricted by internal and external factors, the optimization of enterprise asset structure is the result of multi-party game. Based on factor analysis, this paper constructs the production technology constraint model of asset structure and optimizes it with capital and income as constraints. In this paper, we study the agent constrained optimization model and discuss the possible solutions of asset structure optimization. This paper finds that the asset structure allocation model with production technology constraints is the basic model of asset structure optimization and has anchoring effect.

Structural optimization of H-type VAWT blade under fluid-structure interaction conditions

To reduce the errors caused by the rigid body hypothesis in the aerodynamics-structure coupling calculation and improve the structural performance, an optimum structure design with the consideration of the fluid-structure interaction are performed for the H-type vertical axis wind turbine (VAWT) blade. Based on the ANSYS Workbench platform, the geometric model, computational domain and grids of the wind wheel are constructed, the turbulence model, boundary conditions and composite material layers are set up, and the fluid and solid domains are solved in a coupled way. The single-objective structural optimization model in which the thicknesses of glass clothes, foam and gel coat, and the positions of two webs are taken as design variables is solved using the response surface optimization method to minimize the wind wheel mass. The frequencies and vibration modes of original and optimized blades with and without pre-stress and the transient characteristics of wind wheels in different wind speeds are investigated. The results indicate that after the blade optimization, the first-order frequency and critical speed become larger and other frequencies reduce for the static, single pre-stress and multiple pre-stresses states, and the maximum displacement, stress and strain of the wind wheel decrease under rated and extreme wind speeds, confirming significant performance improvements. The research provides useful guidance for the integrated design of structure and aerodynamics of wind turbine blades.

IMPLEMENTATION OF THE INDICATOR SYSTEM IN MODELING OF COMPLEX TECHNICAL SYSTEMS

Context. The problem of determining the optimal topology of a neuromodel, which is characterized by a high level of logical transparency in modeling complex technical systems, is considered. The object of research is the process of applying an indicator system to simplify and select the topology of neuromodels.
 Objective of the work is to develop and use a system of indicators to determine the level of complexity of the modeling problem and gradually select the optimal logically transparent topology of the neuromodel.
 Method. A method is proposed for selecting an optimal, logically transparent neural network topology for modeling complex technical systems using a system of corresponding indicators. At the beginning, the method determines the overall level of complexity of the modeling task and, using the obtained estimate, determines the method for further optimization of the neuromodel. Then, using Task data and input data characteristics, the method allows to obtain the most optimal structure of the neural model for further modeling of the system. The method reduces trainingvtime and increases the level of logical transparency of neuromodels, which significantly expands the practical use of such models, without using neuroevolution methods, which may not be justified by resource-intensive tasks.
 Results. The developed method is implemented and investigated in solving the problem of modeling the dynamics of pitting processes of steel alloys. Using the developed method made it possible to reduce the training time of the model by 22%, depending on the computing resources used. The method also increased the level of logical transparency of the model by reducing the number of computing nodes by 50%, which also indicates faster and more efficient use of resources.
 Conclusions. The conducted experiments confirmed the operability of the proposed mathematical support and allow us to recommend it for use in practice in the design of topologies of neuromodels for further solving modeling, diagnosis and evaluation problems. Prospects for further research may consist in the development of methods for structural optimization of previously synthesized models and the development of new methods for feature selection.

Structural reliability optimization model with credibilistic safety constraint

Some traditional fuzzy reliability models quantify the system uncertainty based on possibility measure metric without self-duality, which leads to the undesired result that the sum of possibility of any one fuzzy set/event and its complementary set/event is not equal to one. To overcome this drawback, this paper formulates a new reliability metric tool based on self-dual credibility measure. Firstly, we introduce credilistic safety constraint to formulate credibilistic reliability optimization model on the basis of limit-state function. Then, several theorems are provided to show that the feasible domain generated by credibilistic safety constraint can be reformulated as the equivalent convex domain under mild assumptions, which enables credibilistic safety constraint to reformulate as the equivalent reliability constraint. Finally, one numerical example on the optimal design of three-bar truss structure is provided to show the applicability of the credibilistic reliability optimization model. By using out-of-sample back-testing, these comparison results are stated on the provided reliability level and the back-testing reliability level, which evidence the effectiveness of the optimal solution achieved by the proposed model for practical structural engineering.

Efficient cross-validatory algorithm for identifying dynamic nonlinear process models

Kernel partial least squares (KPLS) is an effective nonlinear modeling technique for control engineering applications, including model predictive control, process monitoring or general system diagnosis. It can deal with small sample sizes and variable sets that are noisy and highly correlated. Kernel partial least squares maps the input (or cause) variables to a feature space and carries out the task of producing an optimal prediction model for the process output (or effect) variables using the standard linear partial least squares (PLS) approach. Resulting from the typically large size of the feature space, the kernel partial least squares procedure can be computational intensive. In particular, if the optimal model structure is estimated using cross-validation, KPLS is not efficient in handling large data sets. This paper first modifies the conventional kernel partial least squares procedure in order to embed it within a leave-one-out cross-validation (LOOCV) framework. The proposed efficient kernel partial least squares (EKPLS) is able to reduce the computational complexity by an order of magnitude compared to the conventional approach, which is proven both analytically and through modeling applications to three industrial data sets.

Recent advances in knowledge‐based model structure optimization and extrapolation techniques for microwave applications

AbstractArtificial neural network modeling techniques have been recognized as important vehicles in the microwave computer‐aided design (CAD) area in addressing the growing challenges of designing next generation microwave device, circuits, and systems. This article provides an overview of recent advances in knowledge‐based neural network model generation and extrapolation techniques for microwave applications. We first introduce the unified knowledge‐based neural network structure optimization technique. Using the distinctive property for feature selection of l1 optimization, this unified modeling technique efficiently determines the type and topology of the mapping structure in a knowledge‐based model. This knowledge‐based model structure optimization technique is more flexible and systematic, and can further speed up the knowledge‐based neural model development. As a further advancement, we also discuss the advanced multi‐dimensional extrapolation technique for neural‐based microwave modeling. The purpose is to make the neural network model can be reliably used not only inside the training range but also outside the training range. Multi‐dimensional cubic polynomial extrapolation formulation and optimization over grids outside the training range are utilized to make neural models more robust and reliable when they are used outside the training range. The validity of these techniques is demonstrated by microwave modeling examples.

Auto-VirtualNet: Cost-adaptive dynamic architecture search for multi-task learning

Multi-task learning (MTL) improves learning efficiency by solving multiple tasks simultaneously compared to multiple instances of individual learning. However, despite its benefits, there still remain several major challenges: first, negative interference can reduce the learning efficiency when the number of tasks is high or the tasks are of limited relevance. Second, exploring an optimal model structure manually is quite restricted. Last but not least, offering cost-adaptive solutions has not been addressed in the MTL regime. In spite of its notable merits, the combined problem has not been well discussed. In this work, we propose a novel MTL approach to address the combinatorial problem while minimizing memory consumption. The proposed method discovers multiple network models dynamically from a pool of candidate models, and produces a set of widely distributed solutions with respect to different computational costs for each task. For the diversity of candidate models, we modularize the given backbone architecture that generates basic building blocks and then construct a hierarchical structure based on the building blocks. The proposed method is trained to optimize both task performance and computational costs of selected models. The proposed method dynamically generates optimal networks for each task and offers significant performance improvements over existing MTL approaches in a range of experiments.