首页期刊导航|Applied Soft Computing
期刊信息/Journal information
Applied Soft Computing
Elsevier Science, B.V.
Applied Soft Computing

Elsevier Science, B.V.

1568-4946

Applied Soft Computing/Journal Applied Soft ComputingEIISTPSCIAHCI
正式出版
收录年代

    A population-based algorithm with the selection of evaluation precision and size of the population

    Cpalka, KrzysztofSlowik, AdamLapa, Krystian
    21页
    查看更多>>摘要：In this paper, a new nature-inspired hybrid population-based algorithm is proposed. Firstly, during its operation, it changes the size of the population to reduce the number of processed individuals. For this purpose, dedicated functions that determine the size of population for each algorithm step are used. Secondly, for each individual of the population, the algorithm selects and changes an operator for its modification. This provides a balance between searching for new solutions and fine-tuning of those already found. Thirdly, the algorithm can control the sampling period of the optimized (dynamic) systems, reducing the complexity of the fitness function for individuals. This makes it easier to use the algorithm to optimize even complex systems, which is of great practical importance. Finally, the algorithm allows to solve problems consisting in choosing the structure of the solution and the parameters of this structure. The control problems considered in the simulations, where both the parameters and the structure of the PID-based controller have to be selected, are exactly this type of problem. The results obtained for the proposed algorithm are significantly better than the results obtained with the use of other methods. (C) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Data-driven optimization for mitigating tunnel-induced damages

    Guo, KaiZhang, Limao
    18页
    查看更多>>摘要：Along with the rapid development of urban metro systems, the tunnel-induced damage becomes one of the most critical problems closely related to the safety of tunneling projects. It is urgent to perform an in-depth analysis, identify the key factors influencing the damage, and look for the strategies that could optimize the tunneling process to realize the tunnel-induced damage mitigation. To achieve this, a hybrid data-driven approach with the integration of random forest and non-dominant sorting genetic algorithm-II (NSGA-II) is proposed to perform the multi-objective optimization for mitigating tunnelinduced damages under uncertainty. The random forest is used to construct the meta-model between identified influential factors and objectives. NSGA-II is used to perform the optimization process based on the proposed optimization principle. A total of 16 input variables are identified, and two key factors (i.e., the accumulative settlement and building tilt rate) are determined as the optimization objectives related to the mitigation of the tunnel-induced damage. A case study is conducted to test the applicability and effectiveness of the proposed approach. Through the case study, it is found that: (1) An average damage mitigation improvement degree of 20.9% can be achieved through the proposed optimization process; (2) The optimization can gain the highest improvement degree 32.6% for the tunnel-induced damage mitigation problem when adjusting 3 influential variables; (3) The proposed approach is applicable for the damage mitigation optimization with more objectives, but the consideration of a third objective degrades the optimization improvement for the first two by 2.2% and 6.5%, respectively. The novelty lies in that: (1) The random forest algorithm is incorporated into the model to represent the complex relationship between the identified objectives and the influential factors; (2) Multi-objectives are identified for the mitigation of the tunnel-induced damages, and the optimization of the multi-objectives is realized by the integration of NSGA-II. This research enriches the area of the safety management of tunneling projects by the integration of the random forest and NSGA-II algorithms. With the proposed hybrid approach, the complex relationship between desired objectives and the influential factors could be represented, and the damage mitigation and project optimization could be realized, even potential conflict between objectives may exist. (c) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    FDN-ADNet: Fuzzy LS-TWSVM based deep learning network for prognosis of the Alzheimer's disease using the sagittal plane of MRI scans

    Sharma, RahulGoel, TriptiTanveer, M.Murugan, R....
    11页
    查看更多>>摘要：Alzheimer's disease (AD) is the most pervasive form of dementia, resulting in severe psychosocial effects such as affecting personality, reasoning, emotions, and memory. Several neuroimaging techniques are available to correctly identify the structural changes in the brain, out of which the most popular is structural T-1 weighted Magnetic Resonance Imaging (MRI). From 3D MRI, sagittal plane slices provide more clear information related to the hippocampus, amygdala, corpus callosum, and several vital regions of the brain, which defines the extent of degeneration of the AD. Although diverse analysis of machine learning (ML) and deep learning (DL) based algorithm is already proposed for diagnosis of AD, still there is scope of research for early prediction so that treatment can be started either by medication or by improving the lifestyle. This paper proposed a DL model for all level feature extraction and fuzzy hyperplane based least square twin support vector machine (FLS-TWSVM) for the classification of the extracted features for early diagnosis of AD (FDN-ADNet) using extracted sagittal plane slices from 3D MRI images. Model is trained over the online available ADNI dataset and triangular fuzzy function is applied for the construction of hyperplane for classification. The proposed model attains the highest accuracy of 97.15%, 97.29% and 95% for CN vs AD, CN vs MCI and AD vs MCI classification, respectively when compared with the several state of the art networks. (C) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Intuitionistic fuzzy twin support vector machines with the insensitive pinball loss

    Liang, ZhizhengZhang, Lei
    14页
    查看更多>>摘要：Due to the use of membership and nonmembership functions of samples from intuitionistic fuzzy sets(IFSs), intuitionistic fuzzy twin support vector machines (IFTSVMs) can effectively suppress noise in the data. However, the objective function of IFTSVMs partially considers score values of samples and employs the hinge loss function which leads to the sensitivity to feature noise and instability to re-sampling. To enhance the performance of IFTSVMs, we propose novel IFTSVMs with the insensitive pinball loss function. In the proposed convex optimization models, a simple strategy is devised to achieve the score value of each training sample and score values of samples in both classes are defined by using IFSs. Unlike previous methods, we introduce two groups of slack variables to derive the dual formulations of convex models which make them have compact representations. Some properties of the proposed models including geometric properties and noise insensitivity are theoretically analyzed. We also explain the proposed models in terms of the idea of the weighted scatter minimization, which provides theoretical foundations for the proposed models. Experiments on a series of data sets are performed and experimental results demonstrate that the proposed convex models are superior to some existing learning models in the presence of feature noise or label noise. (c) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Multi-modal feature fusion for 3D object detection in the production workshop

    Hou, RuiChen, GuangzhuHan, YinheTang, Zaizuo...
    16页
    查看更多>>摘要：3D object detection technology is of great significance to realize intelligent perception and ensure the production safety of a workshop. Existing 3D object detection relies on large-scale, high-quality 3D annotation data and is unsuitable for actual workshop scenes' perception. This paper proposes a multi-modal feature fusion 3D object detection method (MFF3D) for a production workshop. The design of MFF3D includes the following steps: (1) Improved YOLOv3 attains the 2D prior region of an object, and RGB-D saliency detection obtains the object image pixels in that region. (2) Depth image pixels corresponding to the object are projected to generate the object's frustum point cloud, and a multi-modal feature fusion strategy simplifies the object's frustum point cloud, so as to remove outlier points and reduce the number of point clouds. This can replace the 3D object reasoning process based on deep neural networks; (3) An axis-aligned bounding box algorithm is used to generate the object's 3D bounding box, and principal component analysis algorithm (PCA) is used to calculate the object's pose information. MFF3D is applied in the workshop, and experiments verify the feasibility and detection accuracy. We set up a production workshop object dataset (PWOD) for experimental evaluation. In the case of a small amount of 2D annotation data and no 3D annotation data, experimental results show that when the threshold value of intersection over union of 3D object (IoU(3D)) is 0.5, the mean average precision value of 3D object (mAP(3D)) reaches 60.31, and the detection speed reaches 3 FPS. MFF3D does not rely on 3D annotation data and can effectively detect objects of a production workshop. (C) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Application of integrated factor evaluation-analytic hierarchy process-T-S fuzzy fault tree analysis in reliability allocation of industrial robot systems

    Bai, BinXie, ChuxiongLiu, XiangdongLi, Wei...
    17页
    查看更多>>摘要：Aiming at the defects (only two kinds of state, i.e., normal or fault) of existed reliability allocation methods without considering the intermediate degradation process, a methodology named integrated factor evaluation-analytic hierarchy process-T-S fuzzy fault tree analysis (IFE-AHP-T-S fuzzy FTA) is proposed to allocate the reliability index of industrial robot systems (IRSs) which have multiple fault states. Firstly, the reliability model of IRSs is established and the allocation principle of reliability index is discussed. Secondly, two-layer IFE model is established considering the degradation of mechanical structure and multi-state fault of IRSs to evaluate the technical merit of different subsystems. The hesitant fuzzy language set is adopted to reduce the subjectivity of expert evaluation process, which have the ability of dealing with uncertain information. Then, the AHP is proposed to allocate different weights for influence factors, and the T-S fuzzy FTA is presented to calculate fault probability and mean time between failures (MTBF) in the process of weight allocation of reliability index for IRSs and six subsystems. Finally, multi-state reliability index allocation of IRSs is completed. This investigation has a significance for reducing the fault probability and unsafe factors, and provides a theoretical basis for the whole life cycle design of IRSs. (C) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Object localization through a single multiple-model switching CNN and a superpixel training approach

    Lotfi, FarazFaraji, FarnooshTaghirad, Hamid D.
    17页
    查看更多>>摘要：Object localization has a vital role in any object detector and tracker, and therefore, has been the focus of attention by many researchers. In this article, a special training approach is proposed for a light convolutional neural network (CNN) to determine the region of interest (RoI) in an image while effectively reducing the number of probable anchor boxes. Almost all CNN based detectors utilize a fixed input size image, which may yield poor performance when dealing with various object sizes. In this paper, a different CNN structure is proposed taking three different input sizes, to enhance the performance. To demonstrate the effectiveness of the proposed method, two common data set are used for training while tracking by localization application is considered to demonstrate its final performance. The promising results indicate the applicability of the presented structure and the training method in practice. (C) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Soft measurement of effluent index in sewage treatment process based on overcomplete broad learning system

    Chang, PengZhao, LuLuMeng, FanChaoXu, Ying...
    14页
    查看更多>>摘要：To measure whether the sewage treatment meets the standards, biochemical oxygen demand (BOD5) is often used to determine, but the measurement of this indicator often has a long time lag and difficult to observe the real-time changes of BOD5, which brings inconvenience to the industrial process. The soft measurement technology based on neural network can realize BOD5 prediction at every moment by means of auxiliary variables, which has attracted people's attention. However, there are still two problems with soft measurement technology, neural network-based soft measurement technology has high computational complexity and a certain time delay in measurement; and it cannot handle non-Gaussian data well. To solve them, this paper introduces an over-complete broad learning system (OBLS) based on feature fusion to deal with the problems of real-time measurement of BOD5 in sewage treatment industrial process. In view of the data characteristics, the feature extraction ability of the BLS is improved, the non-Gaussian characteristic of sewage data is captured by the method of Overcomplete Independent Component Analysis (OICA), and the OBLS is used to deal with the real-time soft measurement. Compared with state-of-the-art methods on the sewage standard test platform, the measurement accuracy of the proposed algorithm is found to be higher and the performance is more stable. (C) 2021 Published by Elsevier B.V.
      原文链接:
    • NSTL
    • Elsevier

    A multi-objective cross-entropy optimization algorithm and its application in high-speed train lateral control

    Tang, QichaoMa, LeiZhao, DuoLei, Jieyu...
    16页
    查看更多>>摘要：In this paper a novel improved multi-objective cross-entropy optimization (MOCEO+) algorithm is proposed. We seek to provide a low-cost, fast and effective solution for complex multi-objective problems. Firstly, a population size segmentation mechanism is adopted to reduce the computational cost. This also helps to increase the exploration ability of the algorithm to the optimal Pareto front and the convergence rate is accelerated as well. Then, an individual selection mechanism based on hyper volume (HV) sorting strategy is proposed to retain the elite individuals in the evolution process. Finally, a recombination mechanism is provided to increase diversity of the population individuals and to avoid local optimum. The test results of 2-, 3-, 5-objective WFG test functions indicate that MOCEO+ offers better performance and faster convergence compared with six optimization algorithms. In order to verify feasibility and effectiveness of the MOCEO+ in engineering practice, it is applied to the parameter optimization of the repetitive learning controller for the high-speed train lateral suspension system. The simulation demonstrates that the suspension system optimized by MOCEO(+ )has better lateral stability compared with three other algorithms. Particularly, the lateral vibration is significantly decreased in the sensitive frequency range [1,2] Hz of human body. (C) 2021 Elsevier B.V. All rights reserved.
      原文链接:
    • NSTL
    • Elsevier

    Adaptive neighborhood size and effective geometric features selection for 3D scattered point cloud classification

    Gunen, Mehmet Akif
    14页
    查看更多>>摘要：Classification of 3D scatter and unorganized point cloud (PC) is an ongoing hard problem due to high redundancy, unbalanced sampling density, and large data structure of PC. Geometric and spectral features derived from the PC are generally used for classification. In this paper, an Omnivariance based adaptive neighborhood size selection method and a new feature set composed of 14 features are proposed for extraction of geometric features for each individual point within the local neighborhood. Performance of 8 modern classifiers with different strategies (i.e., boosting, ensemble, and deep learning etc.) were evaluated on the Oakland, Vaihingen, and ISPRS datasets. These 3 datasets are identified by 5, 9, and 2 distinct object classes, respectively. The results were compared with different neighborhood size selection methods (i.e., eigenentropy based, fixed number of the k-nearest neighbors) and feature set (i.e., 21 features). Only 3D local features were employed to classify datasets with varying characteristics and properties. The proposed optimum neighbor selection method and feature set provided the best statistical results with Auto-Encoder classifier (the overall accuracies are over 85%, 60% and, 90% the Oakland, Vaihingen, and ISPRS datasets, respectively). Especially for the ISPRS dataset, the Auto-Encoder obtained over 94%, 90%, and 93% precision, recall, and f-score, respectively. (C) 2021 Published by Elsevier B.V.
      原文链接:
    • NSTL
    • Elsevier