Invasive weed optimization for model order reduction of linear MIMO systems

In this work, a model order reduction (MOR) technique for a linear multivariable system is proposed using invasive weed optimization (IWO). This technique is applied with the combined advantages of retaining the dominant poles and the error minimization. The state space matrices of the reduced order system are chosen such that the dominant eigenvalues of the full order system are unchanged. The other system parameters are chosen using the invasive weed optimization with objective function to minimize the mean squared errors between the outputs of the full order system and the outputs of the reduced order model when the inputs are unit step. The proposed algorithm has been applied successfully, a 10th order Multiple-Input–Multiple-Output (MIMO) linear model for a practical power system was reduced to a 3rd order and compared with recently published work.     

An Arnoldi-Type Algorithm for Parametric Finite Element Modelling of Microwave Components

In this paper, we present an efficient model order reduction (MOR) technique for computing the solution of a linear equation system depending on a multivariate polynomial of scalar parameters. We propose a moment-matching projection method based on the concept of multivariate Krylov spaces of higher order. The new method exhibits greater stability and allows to build models of higher order than previous approaches. (© 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Index-aware model order reduction for differential-algebraic equations

We introduce a model order reduction (MOR) procedure for differential-algebraic equations, which is based on the intrinsic differential equation contained in the starting system and on the remaining algebraic constraints. The decoupling procedure in differential and algebraic part is based on the projector and matrix chain which leads to the definition of tractability index. The differential part can be reduced by using any MOR method, we use Krylov-based projection methods to illustrate our approach. The reduction on the differential part induces a reduction on the algebraic part. In this paper, we present the method for index-1 differential-algebraic equations. We implement numerically this procedure and show numerical evidence of its validity.     

A memory-efficient model order reduction for time-delay systems

Michiels et al. (SIAM J. Matrix Anal. Appl. 32(4):1399–1421, 2011) proposed a Krylov-based model order reduction (MOR) method for time-delay systems. In this paper, we present an efficient process, which requires less memory consumption, to accomplish the model reduction. Memory efficiency is achieved by replacing the classical Arnoldi process in the MOR method with a two-level orthogonalization Arnoldi (TOAR) process. The resulting memory requirement is reduced from quadratic dependency of the reduced order to linear dependency. Besides, this TOAR process can also be applied to reduce the original delay system into a reduced-order delay system. Numerical experiments are given to illustrate the feasibility and effectiveness of our method.     

Model order reduction methods and their possible application in compliant mechanisms

Model order reduction (MOR) is commonly used to approximate large-scale linear time-invariant dynamical systems. A new feed unit based on a compliant mechanism consisting of flexure hinges can be described by a discrete system of n ordinary differential equations. A projection framework using modal and Krylov subspace techniques is applied to reduce the order of the system to lower computational cost and make the model feasible for control, analysis and optimization. Single flexure hinges are investigated numerical, analytical and experimental and compared to reduced models via modal and tangential Krylov subspace methods regarding the first eigenfrequency. (© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Multi-step prediction of time series with random missing data

Multi-step prediction is still an open challenge in time series prediction. Moreover, practical observations are often incomplete because of sensor failure or outliers causing missing data. Therefore, it is very important to carry out research on multi-step prediction of time series with random missing data. Based on nonlinear filters and multilayer perceptron artificial neural networks (ANNs), one novel approach for multi-step prediction of time series with random missing data is proposed in the study. With the basis of original nonlinear filters which do not consider the missing data, first we obtain the generalized nonlinear filters by using a sequence of independent Bernoulli random variables to model random interruptions. Then the multi-step prediction model of time series with random missing data, which can be fit for the online training of generalized nonlinear filters, is established by using the ANN’s weights to present the state vector and the ANN’s outputs to present the observation equation. The performance between the original nonlinear filters based ANN model for multi-step prediction of time series with missing data and the generalized nonlinear filters based ANN model for multi-step prediction of time series with missing data is compared. Numerical results have demonstrated that the generalized nonlinear filters based ANN are proportionally superior to the original nonlinear filters based ANN for multi-step prediction of time series with missing data.     

Nonlinear dynamic analysis of a complex dual rotor-bearing system based on a novel model reduction method

In this paper, a dynamic model of a complex dual rotor-bearing system of an aero-engine is established based on the finite element method with three types of beam elements (rigid disc, cylindrical beam element and conical beam element), as well as taking into account the nonlinearities of all of the supporting rolling element bearings. To rapidly and accurately analyze dynamic behaviors of the complex dual rotor-bearing system, a two-level model order reduction (MOR) method is proposed by combining component mode synthesis (CMS) method and proper orthogonal decomposition (POD) technique. The first-level reduced-order model (ROM) of the dual rotors is obtained by CMS method with a high precision for the original system. Then, the POD method is applied to second-level model order reduction to further decrease the degrees of freedom (DOFs) of first-level ROM. Second-level ROM with mode expansion and direct second-level ROM are obtained, and the nonlinear displacement responses of the two ROMs are compared with the first-level ROM. The numerical results demonstrate that the proposed method has a higher computational efficiency and accuracy in terms of mode expansion than the direct model reduction by using POD method. In addition, the nonlinear vibration responses of the dual rotor-bearing system are studied by this second-level ROM in the case of different clearances of the inter-shaft bearing. The results indicate that the dynamic characteristics of the dual rotor-bearing system are very complicated for a large clearance.     

Model order reduction of finite element models: improved component mode synthesis

The finite element (FE) approach constitutes an essential methodology when modelling the elastic properties of structures in various research disciplines such as structural mechanics, engine dynamics and so on. Because of increased accuracy requirements, the FE method results in discretized models, which are described by higher order ordinary differential equations, or, in FE terms, by a large number of degrees of freedom (DoF). In this regard, the application of an additional methodology, referred to as the model order reduction (MOR) or DoF condensation, is rather compulsory. Herein, a reduced dimension set of ordinary differential equations is generated, i.e. the initially large number of DoF is condensed, while aiming to keep the dynamics of the original model as intact as possible. In the commercially available FE software tools, the static and the component mode syntheses (CMS) are the only available integrated condensation methods. The latter represents the state of the art generating well-correlated reduced order models (ROMs), which can be further utilized for FE or multi-body systems simulations. Taking into consideration the information loss of the CMS, which is introduced by its part-static nature, the improved CMS (ICMS) method is proposed. Here the algorithmic scheme of the standard CMS is adopted, which is qualitatively improved by adequately considering the advantageous characteristics of another MOR approach, the so-called improved reduction system method. The ICMS results in better correlated reduced order models in comparison to all the aforementioned methods, while preserving the required structural properties of the original FE model.     

基于IGRA-ISVD约减和RBF的森林火灾消防直升机需求预测模型

针对森林火灾消防直升机需求预测问题,提出了一种基于改进灰色关联分析(IGRA)和改进奇异值分解(ISVD)约简的径向基函数(RBF)神经网络预测模型.首先,基于既有研究梳理了森林火灾消防直升机需求预测指标体系;然后,在改进灰色关联分析和奇异值分解方法的基础上,分别对消防直升机需求预测数据信息进行属性约简和维度约简;最后,利用约简预测数据信息对RBF神经网络进行训练,进而构建消防直升机数量预测模型.案例分析和对比分析表明了本文所提方法的可行性和合理性.     

基于卷积神经网络气动力降阶模型的翼型优化方法

针对非线性大扰动翼型气动力优化问题,提出了基于卷积神经网络气动力降阶模型的优化方法.该方法用不同形状参数下翼型的气动力数据作为训练信号,训练卷积神经网络翼型气动力降阶模型.采用该气动力降阶模型,以最大升阻比为目标,对翼型进行优化,结果表明该方法可用于大扰动下翼型气动力的预测和优化.该文同时还讨论了池化法和径向基法的训练...     

预测水驱油藏中油、水流动问题的快速方法

以最佳正交分解(POD)技术为基础提出了一种快速预测油藏中油、水流动问题的方法.采用POD技术建立了水驱油藏中油、水两相流动的低阶模型.通过油藏数值模拟方法获得二维水驱油藏模型在时间0～500 d内的压力和含水饱和度的100个样本, 并从样本中提取出一组压力和含水饱和度的POD基函数.当注采参数不断变化后,采用已求得的POD基函数结合低阶模型对新的物理场进行预测.研究结果表明:POD方法能够快速、准确地预测出水驱油藏的压力和含水饱和度场,文中算例给出压力和含水饱和度场的预测误差分别不超过1.2%与1.5%,且计算速度比直接进行油藏数值模拟快50倍以上.     

AN ADAPTIVE LEARNING FRAMEWORK FOR FORECASTING SEASONAL WATER ALLOCATIONS IN IRRIGATED CATCHMENTS

Abstract This paper describes an adaptive learning framework for forecasting end‐season water allocations using climate forecasts, historic allocation data, and results of other detailed hydrological models. The adaptive learning framework is based on artificial neural network (ANN) method, which can be trained using past data to predict future water allocations. Using this technique, it was possible to develop forecast models for end‐irrigation‐season water allocations from allocation data available from 1891 to 2005 based on the allocation level at the start of the irrigation season. The model forecasting skill was further improved by the incorporation of a set of correlating clusters of sea surface temperature (SST) and the Southern oscillation index (SOI) data. A key feature of the model is to include a risk factor for the end‐season water allocations based on the start of the season water allocation. The interactive ANN model works in a risk‐management context by providing probability of availability of water for allocation for the prediction month using historic data and/or with the incorporation of SST/SOI information from the previous months. All four developed ANN models (historic data only, SST incorporated, SOI incorporated, SST‐SOI incorporated) demonstrated ANN capability of forecasting end‐of‐season water allocation provided sufficient data on historic allocation are available. SOI incorporated ANN model was the most promising forecasting tool that showed good performance during the field testing of the model.     

Reduced order modeling of nonlinear time periodic systems subjected to external periodic excitations

In this work, new methodologies for order reduction of nonlinear systems with periodic coefficients subjected to external periodic excitations are presented. The periodicity of the linear terms is assumed to be non-commensurate with the periodicity of forcing vector. The dynamical equations of motion are transformed using the Lyapunov–Floquet (L–F) transformation such that the linear parts of the resulting equations become time-invariant while the forcing and nonlinearity takes the form of quasiperiodic functions. The techniques proposed here construct a reduced order equivalent system by expressing the non-dominant states as time-varying functions of the dominant (master) states. This reduced order model preserves stability properties and is easier to analyze, simulate and control since it consists of relatively small number of states in comparison with the large scale system.Specifically, two methods are discussed to obtain the reduced order model. First approach is a straightforward application of linear method similar to the ‘Guyan reduction’. The second novel technique proposed here extends the concept of ‘invariant manifolds’ for the forced problem to construct the fundamental solution. Order reduction approach based on this extended invariant manifold technique yields unique ‘reducibility conditions’. If these ‘reducibility conditions’ are satisfied only then an accurate order reduction via extended invariant manifold approach is possible. This approach not only yields accurate reduced order models using the fundamental solution but also explains the consequences of various ‘primary’ and ‘secondary resonances’ present in the system. One can also recover ‘resonance conditions’ associated with the fundamental solution which could be obtained via perturbation techniques by assuming weak parametric excitation. This technique is capable of handling systems with strong parametric excitations subjected to periodic and quasi-periodic forcing. It is anticipated that these order reduction techniques will provide a useful tool in the analysis and control system design of large-scale parametrically excited nonlinear systems subjected to external periodic excitations.     

Learning patient-specific parameters for a diffuse interface glioblastoma model from neuroimaging data

Parameters in mathematical models for glioblastoma multiforme (GBM) tumour growth are highly patient specific. Here, we aim to estimate parameters in a Cahn–Hilliard type diffuse interface model in an optimised way using model order reduction (MOR) based on proper orthogonal decomposition (POD). Based on snapshots derived from finite element simulations for the full-order model (FOM), we use POD for dimension reduction and solve the parameter estimation for the reduced-order model (ROM). Neuroimaging data are used to define the highly inhomogeneous diffusion tensors as well as to define a target functional in a patient-specific manner. The ROM heavily relies on the discrete empirical interpolation method, which has to be appropriately adapted in order to deal with the highly nonlinear and degenerate parabolic partial differential equations. A feature of the approach is that we iterate between full order solvers with new parameters to compute a POD basis function and sensitivity-based parameter estimation for the ROM problems. The algorithm is applied using neuroimaging data for two clinical test cases, and we can demonstrate that the reduced-order approach drastically decreases the computational effort.     

Development and Evaluation of Decision-Making Model for Stock Markets

The paper introduces an intelligent decision-making model which is based on the application of artificial neural networks (ANN) and swarm intelligence technologies. The proposed model is used to generate one-step forward investment decisions for stock markets. The ANN are used to make the analysis of daily stock returns and to calculate one day forward decision for purchase of the stocks. Subsequently the Particle Swarm Optimization (PSO) algorithm is applied in order to select the “the best” ANN for the future investment decisions and to adapt the weights of other networks towards the weights of the best network. The experimental investigations were made considering different forms of decision-making model: different number of ANN, ANN inputs, sliding windows, and commission fees. The paper introduces the decision-making model, its evaluation results and discusses its application possibilities.     

Modal Substructuring of Geometrically Nonlinear Plates

This work considers model reduction of geometrically nonlinear finite element (FE) model of a plate structure developed in a commercial FE package. The structure is first divided into smaller substructures. Since there is normally no access to the nonlinear stiffness tensors in FE packages, a non-intrusive method is used in this paper to reduce the order of each substructure separately. In order to generate the nonlinear reduced order model (NLROM), the reduced substructures are assembled using the coupling procedure of the Component Mode Synthesis (CMS) method. As a linear basis, truncated free and fixed interface modes are used here to check the efficiency of the developed NLROM based on them. A plate structure subjected to large deflections is considered in this study to implement the substructuring method. For the sake of validation, Nonlinear Normal Modes (NNMs) are employed to check the convergence of NLROMs in a broadband frequency range. (© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Interpolation-based nonlinear parametric MOR for gas pipelines

This work deals with the model order reduction (MOR) of nonlinear, parametric systems of partial differential equations as they arise in gas pipeline modeling. We present an approach that is based on a linearization around parametric working points, linear modal reduction and interpolation. The choice of the working points as well as the interpolation strategy crucially determine the approximation quality. (© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

New approach to applying neural network in nonlinear dynamic model

In this work, radial basis function neural network (RBF-NN) is applied to emulate an extended Kalman filter (EKF) in a data assimilation scenario. The dynamical model studied here is based on the one-dimensional shallow water equation DYNAMO-1D. This code is simple when compared with an operational primitive equation models for numerical weather prediction. Although simple, the DYNAMO-1D is rich for representing some atmospheric motions, such as Rossby and gravity waves. It has been shown in the literature that the ability of the EKF to track nonlinear models depends on the frequency and accuracy of the observations and model errors. In some cases, just fourth-order moment EKF works well, but will be unwieldy when applied to high-dimensional state space. Artificial Neural Network (ANN) is an alternative solution for this computational complexity problem, once the ANN is trained offline with a high order Kalman filter, even though this Kalman filter has high computational cost (which is not a problem during ANN training phase). The results achieved in this work encourage us to apply this technique on operational model. However, it is not yet possible to assure convergence in high dimensional problems.     

A 3‐level discretisation strategy using substructuring

This contribution deals with a new three‐level discretisation strategy which enables to discretise a structure according to the required modelling accuracy. Correspondingly the structure is separated into a continuum, a structural and a black box model level. This note especially focuses on the transitions between continuum‐structure and structure‐substitute model. The first one is realised by means of a model adaptive concept based on the innovative finite element technology of the Q1SP family. The latter is carried out by an enhanced modal reduction method which is combined with a novel substructure technique. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Improving weighted information criterion by using optimization

Although artificial neural networks (ANN) have been widely used in forecasting time series, the determination of the best model is still a problem that has been studied a lot. Various approaches available in the literature have been proposed in order to select the best model for forecasting in ANN in recent years. One of these approaches is to use a model selection strategy based on the weighted information criterion (WIC). WIC is calculated by summing weighted different selection criteria which measure the forecasting accuracy of an ANN model in different ways. In the calculation of WIC, the weights of different selection criteria are determined heuristically. In this study, these weights are calculated by using optimization in order to obtain a more consistent criterion. Four real time series are analyzed in order to show the efficiency of the improved WIC. When the weights are determined based on the optimization, it is obviously seen that the improved WIC produces better results.