This work describes a new methodology for robust identification (RI), meaning the identification of the parameters of a model and the characterization of uncertainties. The alternative proposed handles non-linear mode...
详细信息
This work describes a new methodology for robust identification (RI), meaning the identification of the parameters of a model and the characterization of uncertainties. The alternative proposed handles non-linear models and can take into account the different properties demanded by the model. The indicator that leads the identification process is the identification error (IE), that is, the difference between experimental data and model response. In particular, the methodology obtains the feasible parameter set (FPS, set of parameter values which satisfy a bounded IE) and a nominal model in a non-linear identification problem. To impose different properties on the model, several norms of the IE are used and bounded simultaneously. This improves the model quality, but increases the problem complexity. The methodology proposes that the RI problem is transformed into a multimodal optimization problem with an infinite number of global minima which constitute the FPS. For the optimization task, a special genetic algorithm (epsilon-GA), inspired by Multiobjective evolutionary algorithms, is presented. This algorithm characterizes the FPS by means of a discrete set of models well distributed along the FPS. Finally, an application for a biomedical model which shows the blockage that a given drug produces on the ionic currents of a cardiac cell is presented to illustrate the methodology. (C) 2008 Elsevier Ltd. All rights reserved.
Room temperature Fourier transform infrared spectra of the four-membered heterocycle trimethylene sulfide were collected with a resolution of 0.00096 cm(-1) using synchrotron radiation from the Canadian Light Source f...
详细信息
Room temperature Fourier transform infrared spectra of the four-membered heterocycle trimethylene sulfide were collected with a resolution of 0.00096 cm(-1) using synchrotron radiation from the Canadian Light Source from 500 to 560 cm(-1). The in-plane ring deformation mode (nu(13)) at similar to 529 cm(-1) exhibits dense rotational structure due to the presence of ring inversion tunneling and leads to a doubling of all transitions. Preliminary analysis of the experimental spectrum was pursued via traditional methods involving assignment of quantum numbers to individual transitions in order to conduct least squares fitting to determine the spectroscopic parameters. Following this approach, the assignment of 2358 transitions led to the experimental determination of an effective Hamiltonian. This model describes transitions in the P and R branches to J' = 60 and K-a' = 10 that connect the tunneling split ground and vibrationally excited states of the nu(13) band although a small number of low intensity features remained unassigned. The use of evolutionary algorithms (EA) for automated assignment was explored in tandem and yielded a set of spectroscopic constants that re-create this complex experimental spectrum to a similar degree. The EA routine was also applied to the previously well-understood ring puckering vibration of another four-membered ring, azetidine (Zaporozan et al., 2010). This test provided further evidence of the robust nature of the EA method when applied to spectra for which the underlying physics is well understood. (C) 2015 Elsevier Inc. All rights reserved.
In this paper, we propose novel methods to find the best relevant feature subset using fuzzy rough set-based attribute subset selection with biologically inspired algorithm search such as ant colony and particle swarm...
详细信息
In this paper, we propose novel methods to find the best relevant feature subset using fuzzy rough set-based attribute subset selection with biologically inspired algorithm search such as ant colony and particle swarm optimization and the principles of an evolutionary process. We then propose a hybrid fuzzy rough with K-nearest neighbor (K-NN)-based classifier (FRNN) to classify the patterns in the reduced datasets, obtained from the fuzzy rough bio-inspired algorithm search. While exploring other possible hybrid evolutionary processes, we then conducted experiments considering (i) same feature selection algorithm with support vector machine (SVM) and random forest (RF) classifier;(ii) instance based selection using synthetic minority over-sampling technique with fuzzy rough K-nearest neighbor (K-NN), SVM and RF classifier. The proposed hybrid is subsequently validated using real-life datasets obtained from the University of California, Irvine machine learning repository. Simulation results demonstrate that the proposed hybrid produces good classification accuracy. Finally, parametric and nonparametric statistical tests of significance are carried out to observe consistency of the classifiers.
In pattern recognition, the classification accuracy has a strong correlation with the selected features. Therefore, in the present paper, we applied an evolutionary algorithm in combination with linear discriminant an...
详细信息
In pattern recognition, the classification accuracy has a strong correlation with the selected features. Therefore, in the present paper, we applied an evolutionary algorithm in combination with linear discriminant analysis (LDA) to enhance the feature selection in a static image-based facial expressions system. The accuracy of the classification depends on whether the features are well representing the expression or not. Therefore the optimization of the selected features will automatically improve the classification accuracy. The proposed method not only improves the classification but also reduces the dimensionality of features. Our approach outperforms linear-based dimensionality reduction algorithms and other existing genetic-based feature selection algorithms. Further, we compare our approach with VGG (Visual Geometry Group)-face convolutional neural network (CNN), according to the experimental results, the overall accuracy is 98.67% for either our approach or VGG-face. However, the proposed method outperforms CNN in terms of training time and features size. The proposed method proves that it is able to achieve high accuracy by using far fewer features than CNN and within a reasonable training time.
In this article, the self-calibration problem of a planar two-degree-of-freedom (2-DoF) parallel manipulator with a redundant joint sensor is studied. By eliminating the passive joint positions, a new error function i...
详细信息
In this article, the self-calibration problem of a planar two-degree-of-freedom (2-DoF) parallel manipulator with a redundant joint sensor is studied. By eliminating the passive joint positions, a new error function is proposed. Furthermore, by decoupling the kinematic parameters of the error function, the minimization process is simplified. In order to obtain the global optimum, three evolutionary algorithms including genetic algorithm, particle swarm optimization, and differential evolution are applied to minimize the error function. In the application, the performances and effectiveness of the applied algorithms on this specific problem are compared under three different error functions, and the results show that the differential-evolution method under the decoupled error function produces the best result. Finally, actual calibration is carried out based on differential evolution under the decoupled error function, and the result demonstrates that all of the 12 parameters of the manipulator are calibrated with high accuracy.
A variety of important engineering and scientific tasks may be formulated as non-linear, constrained optimization problems. Their solution often demands high computational power. It may be reached by means of appropri...
详细信息
A variety of important engineering and scientific tasks may be formulated as non-linear, constrained optimization problems. Their solution often demands high computational power. It may be reached by means of appropriate hardware, software or algorithm improvements. The evolutionary algorithms (EA) approach to solution of such problems is considered here. The EA are rather slow methods;however, the main advantage of their application is observed in the case of non-convex problems. Particularly high efficiency is demanded in the case of solving large optimization problems. Examples of such problems in engineering include analysis of residual stresses in railroad rails and vehicle wheels, as well as the Physically Based Approximation (PBA) approach to smoothing experimental and/or numerical data. Having in mind such analysis in the future, we focus our current research on the significant EA efficiency increase. Acceleration of the EA is understood here, first of all, as decreasing the total computational time required to solve an optimization problem. Such acceleration may be obtained in various ways. There are at least two gains from the EA acceleration, namely i) saving computational time, and ii) opening a possibility of solving larger optimization problems, than it would be possible with the standard EA. In our recent research we have preliminarily proposed several new speed-up techniques based on simple concepts. In this paper we mainly develop acceleration techniques based on simultaneous solutions averaging well supported by a non-standard application of parallel calculations, and a posteriori solution error analysis. The knowledge about the solution error is used to EA acceleration by means of appropriately modified standard evolutionary operators like selection, crossover, and mutation. Efficiency of the proposed techniques is evaluated using several benchmark tests. These tests indicate significant speed-up of the involved optimization process. Further conce
For the classification of motor imagery brain-computer interface (BCI) based on electroencephalography (EEG), appropriate features are crucial to obtain a high classification accuracy. Considering the characteristics ...
详细信息
For the classification of motor imagery brain-computer interface (BCI) based on electroencephalography (EEG), appropriate features are crucial to obtain a high classification accuracy. Considering the characteristics of the EEG signals, the time-frequency-space three-dimensional features are extracted. Due to a considerable number of the extracted features, the performance of a classifier will degrade. Therefore, it is necessary to implement feature selection. However, existing feature selection methods are easy to fall into a local optimum of a high-dimensional feature selection problem. In this paper, a dimensionality reduction mechanism (called DimReM) is proposed, which gradually reduces the dimension of the search space by removing some unimportant features. In principle, DimReM transforms a high-dimensional feature selection problem into a low-dimensional one. DimReM does not introduce any additional parameters and its implementation is simple. To verify its effectiveness, DimReM is combined with different evolutionary algorithms and different classifiers to select features on various kinds of datasets. Compared with evolutionary algorithms without dimensionality reduction, their augmented versions equipped with DimReM can find feature subsets with higher classification accuracies while smaller numbers of selected features.
The search behavior of an evolutionary algorithm depends on the interactions between the encoding that represents candidate solutions to the target problem and the operators that act on that encoding. In this paper, w...
详细信息
The search behavior of an evolutionary algorithm depends on the interactions between the encoding that represents candidate solutions to the target problem and the operators that act on that encoding. In this paper, we focus on analyzing some properties such as locality, heritability, population diversity and searching behavior of various decoder-based evolutionary algorithm (EA) frameworks using different encodings, decoders and genetic operators for spanning tree based optimization problems. Although debate still continues on how and why EAs work well, many researchers have observed that EAs perform well when its encoding and operators exhibit good locality, heritability and diversity properties. We analyze these properties of various EA frameworks with two types of analytical ways on different spanning tree problems;static analysis and dynamic analysis, and then visualize them. We also show through this analysis that EA using the Edge Set encoding (ES) and the Edge Window Decoder encoding (EWD) indicate very good locality and heritability as well as very good diversity property. These are put forward as a potential explanation for the recent finding that they can outperform other recent high-performance encodings on the constrained spanning tree problems.
A distributed evolutionary algorithm is presented that is based on a hierarchy of (fitness or cost function) evaluation passes within each deme and is efficient in solving engineering optimization problems. Starting w...
详细信息
A distributed evolutionary algorithm is presented that is based on a hierarchy of (fitness or cost function) evaluation passes within each deme and is efficient in solving engineering optimization problems. Starting with non-problem-specific evaluations (using surrogate models or metamodels, trained on previously evaluated individuals) and ending up with high-fidelity problem-specific evaluations, intermediate passes rely on other available lower-fidelity problem-specific evaluations with lower CPU cost per evaluation. The sequential use of evaluation models or metamodels, of different computational cost and modelling accuracy, by screening the generation members to get rid of non-promising individuals, leads to reduced overall computational cost. The distributed scheme is based on loosely coupled demes that exchange regularly their best-so-far individuals. Emphasis is put on the optimal way of coupling distributed and hierarchical search methods. The proposed method is tested on mathematical and compressor cascade airfoil design problems.
This work presents a service oriented architecture for evolutionary algorithms, and an implementation of this architecture using a specific technology (called OSGiLiath). Service oriented architecture is a computation...
详细信息
This work presents a service oriented architecture for evolutionary algorithms, and an implementation of this architecture using a specific technology (called OSGiLiath). Service oriented architecture is a computational paradigm where users interact using services to increase the integration between systems. The presented abstract architecture is formed by loosely coupled, highly configurable and language-independent services. As an example of an implementation of this architecture, a complete process development using a specific service oriented technology is explained. With this implementation, less effort than classical development in integration, distribution mechanisms and execution time management has been attained. In addition, steps, ideas, advantages and disadvantages, and guidelines to create service oriented evolutionary algorithms are presented. Using existing software, or from scratch, researchers can create services to increase the interoperability in this area.
暂无评论