检索结果-内蒙古大学图书馆

Rule extraction based on PROMETHEE-assisted multi-objective genetic algorithm for generating interpretable neural networks

引用

APPLIED SOFT COMPUTING 2024年 151卷

作者： Yedjour, Dounia Yedjour, Hayat Amri, Mohammed Bilel Senouci, Adlania Univ Sci & Technol Oran Mohamed Boudiaf USTO MB Fac Math & Comp Sci Dept Comp Sci BP 1505Mnaouer Oran 31000 Algeria

Rule extraction from artificial neural networks (ANNs) refers to the process of identifying the underlying decision-making logic used by the network in its predictions. This involves converting the complex, non-linear relationships between inputs and outputs in the network into a set of interpretable rules. The main objective of rule extraction is to enhance the interpretability of decision-making in artificial neural networks (ANNs) and make it more understandable for human experts, especially in critical domains like healthcare, where clear decision-making is essential. Despite the ability of machine learning methods to interpret ANNs, interpreting complex neural networks remains limited, as most studies tend to sacrifice one criterion, related to rule quality, in favor of another. Rule extraction methods applied directly to complex neural networks may produce overly specific rules that are tailored to the training data. Such rules can suffer from overfitting and may not generalize well to unseen data. Also, The use of multiobjective genetic algorithms for rule extraction can result in a large number of rules, making it challenging for decision-makers to interpret them. This study proposes an innovative approach to address challenges in rule extraction from complex artificial neural networks (ANNs) by integrating a multiobjective genetic algorithm with the PROMETHEE method. By integrating these techniques, the study aims to identify meaningful relationships between inputs and outputs of ANNs, generating concise and reliable rules that are ranked based on support, confidence, and lift values. The use of PROMETHEE's rule ranking mechanism enabled the consideration of rule priority during candidate selection, guiding the genetic algorithm towards discovering high-quality rule sets and efficiently exploring the search space. This approach not only enhances the quality of extracted rules but also facilitates efficient decision-making by striking a balance between rule

关键词： Artificial neural network Classification rules multiobjective genetic algorithm PROMETHEE method

来源：评论

学校读者我要写书评

暂无评论

Multi-performance optimization in wire EDM of Inconel 617 using GRA and genetic algorithm 2nd

Multi-performance optimization in wire EDM of Inconel 617 us...

引用

2nd International Conference on Functional Material, Manufacturing and Performances (ICFMMP)

作者： Nair, Anish Kumanan, S. Shanavas, K. P. Kalasalingam Acad Res & Educ Dept Mech Engn Krishnankoil India Natl Inst Technol Tiruchirappalli Dept Prod Engn Tiruchirappalli India Coll Engn Vadakara Dept Mech Engn Kozhikode Kerala India

Inconel 617 superalloy is a candidate material for various high temperature components employed in ultra-supercritical power plants. Superalloys are hard to machine materials that easier to process using nontraditional methods such as wire electrical discharge machining. This paper discusses multi perfor-mance optimization in machining of Inconel 617 using grey relational analysis and genetic algorithm. The performance characteristics of wire cut EDM are surface roughness, material removal rate, wire wear rate, form and orientation tolerances. Parameters under consideration are pulse on time, pulse off time, current, voltage, flushing pressure, wire tension, table feed and wire speed. The performance measures are consolidated as three performance indices using the grey relational grade. The mathematical models to estimate the performance objectives are developed using regression analysis. multiobjective optimiza-tion is applied through genetic algorithm and a pareto front containing the optimized responses is com-puted. The optimized parameters are validated experimentally and the results are reported. Copyright (c) 2021 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the 2nd International Con-ference on Functional Material, Manufacturing and Performances

关键词： Wire electrical discharge machining Inconel 617 Ultra-supercritical power plants Grey relational analysis multiobjective genetic algorithm

来源：评论

学校读者我要写书评

暂无评论

An optimization model for the impervious surface spatial layout considering differences in hydrological unit conditions for urban waterlogging prevention in urban renewal

引用

ECOLOGICAL INDICATORS 2024年 158卷

作者： Zhao, Juchao Ke, Entong Wang, Bin Zhao, Yaolong South China Normal Univ Sch Geog Guangzhou 510631 Peoples R China Minist Nat Resources Key Lab Nat Resources Monitoring Trop & Subtrop Ar Guangzhou 510631 Peoples R China South China Normal Univ Beidou Res Inst Foshan 528225 Peoples R China

The expansion of impervious surfaces during urbanization, as well as the irrationality of their spatial layout, is one of the main causes of urban waterlogging. Optimizing the spatial layout of impervious surfaces through urban renewal is important for urban waterlogging prevention. The differences in hydrological conditions among different hydrological units affect the optimization efficiency of the spatial layout of impervious surfaces in different hydrological units. Therefore, this study proposes an effective optimization model for impervious surface spatial layout, employing a hybrid approach that couples the Soil Conservation Service Curve Number (SCS-CN) model, nondominated sorting genetic algorithm 2 (NSGA2), and multiple linear programming (MLP) algorithm. This approach is applied in a case study of Guangzhou, China. The results show that (1) compared to the MLP-SCS model, the enhanced model not only provides reduced runoff coefficients but also achieves a better spatial balance of the retention water volume. (2) Compared to the conditions before optimization, medium- and high -density impervious surfaces significantly change after optimization, and the connectivity between patches is reduced. The optimized layout alters the original gradient connection method, preventing the aggregation of runoff from high -density impervious surfaces and thereby reducing the risk of urban waterlogging during heavy rain. This study can provide a reference for optimizing the spatial layout of impervious surfaces aimed at urban waterlogging prevention in urban renewal planning.

关键词： Urban renewal Urban waterlogging prevention Impervious surface optimization multiobjective genetic algorithm Hydrological model Multiple linear programming

来源：评论

学校读者我要写书评

暂无评论

Process Simulation and Optimization of Ammonia-Containing Gas Separation and Ammonia Recovery with Ionic Liquids

引用

ACS SUSTAINABLE CHEMISTRY & ENGINEERING 2021年第1期9卷 312-325页

作者： Zhan, Guoxiong Cao, Fei Bai, Lu Chang, Fei Zhou, Beiya Duan, Yuanmeng Zeng, Shaojuan Dong, Haifeng Li, Zengxi Zhang, Xiangping Univ Chinese Acad Sci Sino Danish Coll Beijing 100049 Peoples R China Chinese Acad Sci Key Lab Green Proc & Engn State Key Lab Multiphase Complex Syst Beijing Key Lab Ion Liquids Clean ProcInst Proc Beijing 100190 Peoples R China Univ Chinese Acad Sci Coll Chem & Engn Beijing 100049 Peoples R China Bohai Univ Coll Math & Phys Jinzhou 121013 Liaoning Peoples R China

Ionic liquids (ILs) have been experimenta ly proved to be effective for ammonia-containing gas separation and recovery. A systematic strategy including thermodynamic models, process simulation, multiobjective genetic algorithm, and assessment for novel IL-based separation of ammonia-containing tail gas and ammonia recovery process was proposed. The conventional IL ([C(4)Mim][NTf2]) and the functional IL ([C(4)im][NTf2]) were selected, and their NH3 removal performance was investigated. Physical properties of models of IL systems were established with temperature-dependent equations, and gas-liquid phase equilibria of the NH3-IL system were molded with the nonrandom two liquid model equation. Total purification cost (TPC), total process CO2 emission (TPCOE), and thermodynamic efficiency (eta(eff)) were selected as the objective functions to be optimized. Process simulation results indicated that under same operational parameters, using functional ILs results in lower NH3 concentration in purified gas and higher removal efficiency than that of conventional ILs. After optimization, a series of solutions satisfying the constraints was provided by the Pareto front. The lowest objective functions can achieve 0.0211 $/N m(3) (TPC), 265.67 kg CO2/h (TPCOE), and 48.05% (eta(eff)). Moreover, using functional ILs could greatly decrease purification cost and energy consumption and avoid wastewater discharge, which is an inevitable environmental problem in the water scrubbing process.

关键词： NH3-containing tail gas separation process simulation multiobjective genetic algorithm ionic liquid

来源：评论

学校读者我要写书评

暂无评论

The green and economical supply of coated board with intermodal distribution: an implementation of novel evolutionary algorithm

引用

OPERATIONAL RESEARCH 2024年第2期24卷 19-19页

作者： Shoukat, Rizwan Zhang, Xiaoqiang Southwest Jiaotong Univ Sch Transportat & Logist Chengdu Peoples R China Natl Engn Lab Integrated Transportat Big Data Appl Chengdu Peoples R China Lab Natl United Engn Lab Integrated & Intelligent 111 North Second Ring Rd Chengdu Peoples R China

In practice, factories emit significant amounts of GHGs, which have been linked to health and environmental issues. These gases are produced during industrial processes and, in many cases, during transportation. Second, historically, generic costs such as production fixed and variable costs, transportation costs, and inventory holding costs have dominated the overall cost of the supply chain (SC). The objective of this research is to model a SC problem to minimize costs while also accounting for GHG emissions from facility to facility (Shanghai, China to Lahore, Pakistan). Initially, the total expenses and GHG emissions of the SC are modeled as a bi-objective mixed-integer linear programme (BOMILP). In comparison to other optimization models, the developed BOMILP model simultaneously optimizes transportation costs and GHG emissions while taking weight and distance constraints into account in the intermodal network. The real-world data comes from one of Asia's largest paper and board industries. To solve the BOMILP problem, a metaheuristic technique such as a multiobjective genetic algorithm (MOGA) and multiobjective bat algorithm (MOBA) is used. To improve the performance of the MOGA, we used multiple crossover operators. Random crossover operator selection and collision crossover are two strategies that have been implemented. The collision crossover is based on elastic collision, and the random crossover operator-selection strategy randomly selects the best operator. The Pareto optimum solution of the MOGA assists decision-makers in making the best trade-off between cost and emissions. Our findings show that the cost of producing (reeling, coating, rewinding) and storing white bleach board (WBB) accounts for 72% of the total cost of the SC, while GHG emissions are 83,484 kg CO2e in the delivery of 1 tonne of material from Shanghai, China to Lahore, Pakistan via intermodal transportation.

关键词： Transportation costs GHG emissions Metaheuristic technique multiobjective genetic algorithm Pareto optimum solution

来源：评论

学校读者我要写书评

暂无评论

multiobjective GA Optimization for Energy Efficient Electric Vehicle Drivetrains 16

Multiobjective GA Optimization for Energy Efficient Electric...

引用

16th International Conference on Ecological Vehicles and Renewable Energies (EVER)

作者： Chakraborty, Sajib Padmaji, Vinayak Tran, Dai-Duong Araujo, Javier A. Corea Geury, Thomas El Baghdadi, Mohamed Hegazy, Omar Vrije Univ Brussel VUB ETEC Dept Pl Laan 2 B-1050 Brussels Belgium Vrije Univ Brussel VUB MOBI Res Grp Pl Laan 2 B-1050 Brussels Belgium Flanders Make B-3001 Heverlee Belgium Applus IDIADA Grp Santa Oliva 43710 Spain

ISBN: (纸本)9781665449021

This paper investigates the impact of using wide bandgap (WBG) technology-based bidirectional interleaved HV DC/DC converters on the performance of battery electric vehicles (BEVs). An existing electric vehicle is upgraded using off-the-shelf components. There are a variety of batteries, high voltage (HV) DC/DCs, inverters, electric motors, transmissions, etc., available off-the-shelf;hence, numerous possible combinations can be formed, which make the optimal component selection process more complicated through analytical methods. In this paper, a multiobjective genetic algorithm (MOGA) is adopted to minimize the electric energy consumption by improving drivetrain efficiency based on the optimal variant selection of the components. It is found from the virtual simulation framework in MATLAB/Simulink (R) that overall, there is a 9.2% reduction in the energy consumption over a given driving cycle, i.e., Worldwide Harmonized Light Vehicles Test Procedure-3a (WLTP3a). To this end, the drivetrain performance in terms of acceleration time from 0-90 km/h is also improved by 10.2%, while the efficiency is improved by 1.5% compared to the conventional e-drivetrain.

关键词： multiobjective genetic algorithm HV DC/DC converter WBG Electric Vehicle accleration time and energy consumption

来源：评论

学校读者我要写书评

暂无评论

Multipath Error Modeling Methodology for GNSS Integrity Monitoring Using a Global Optimization Strategy

引用

REMOTE SENSING 2022年第9期14卷 2130-2130页

作者： Fang, Xinqi Song, Dan Shi, Chuang Fan, Lei Hu, Ziye Beihang Univ Sch Elect Informat Engn Xueyuan Rd 37 Beijing 100191 Peoples R China Minist Ind & Informat Technol Key Lab Nav & Commun Fus Technol Xueyuan Rd 37 Beijing 100191 Peoples R China Beihang Univ Res Inst Frontier Sci Xueyuan Rd 37 Beijing 100191 Peoples R China China Elect Technol Grp Corp Inst 54 Zhongshanxi Rd 589 Shijiazhuang 050011 Hebei Peoples R China

Valid multipath error model is the prerequisite for high-performance GNSS integrity monitoring. It is indispensable to civil aviation and other Safety-of-Life (SoL) users. The model must perfectly bound multipath error while preventing the constructed model from being too conservative. Nevertheless, no sound methodologies to meet both the requirements have been introduced in previous literatures, and subsequently, practices always require iterative manual trade-offs. To improve the efficiency of multipath modeling, we propose a new automatic multipath error modeling methodology. It quantifies the above requirements in the objective function of multiobjective genetic algorithm (GA) so that multipath modeling can be managed automatically. Moreover, through introducing a new model that is based on two inflation factors, conservatism of modeling results can be significantly reduced. Experiments based on a 4-month dataset of BDS-3 Medium Earth Orbit (MEO) satellites show that constructed multipath models effectively bound actual error in each elevation bin. In addition, the new model form with two inflation factors brings average CDF difference reduction of 67.4% at B1I and 50.6% at B3I, which means significantly mitigation in terms of conservatism.

关键词： multipath error modeling Cumulative Distribution Function (CDF) overbounding BDS-3 multiobjective genetic algorithm

来源：评论

学校读者我要写书评

暂无评论

One instruction set computer with optimised polarity-tunable model of double gate CNTFETs

引用

IET CIRCUITS DEVICES & SYSTEMS 2020年第6期14卷 770-779页

作者： Ponnayan, Reena Monica Thazhathu, Sreedevi Vellithiruthi Vellore Inst Technol Sch Elect Engn Chennai Tamil Nadu India

Emerging devices such as double gate carbon nanotube field effect transistors (DG CNTFETs) have opened up manifold possibilities for reconfigurable logic design. The thickness of gate oxide and the employment of inhomogeneous dielectrics over and under the carbon nanotubes (CNTs) impact the operation of DG CNTFETs. In this work, the dielectric constant and the thickness of the gate oxide are optimised to suppress the ambipolar conduction in CNTFETs. A multi-objective genetic algorithm-based approach is proposed to optimise these parameters. The contributions in this study are two-fold: firstly, a DG CNTFET is fabricated with the optimised parameters. By exploiting the ability to select the conduction behaviour using a second gate in the fabricated DG CNTFETs, the device is electrostatically programmed to behave as an n or p-type CNTFET. Secondly, a one instruction set computer is simulated with the optimised model of DG CNTFET. A universal static logic cell which implements 16 logic functions, a D-Latch and a D-FF were built with DG CNTFETs. The logic circuits with polarity-tunable DG CNTFETs outperform other logic structures. There is a 40% betterment in performance primarily due to the reduced number of logic levels and by 25% due to the reduced delay.

关键词： genetic algorithms logic circuits carbon nanotube field effect transistors permittivity semiconductor device models flip-flops one instruction set computer optimised polarity-tunable model double gate carbon nanotube field effect transistors gate oxide p-type CNTFET polarity-tunable DG CNTFET fabricated DG CNTFET double gate CNTFET reconfigurable logic design inhomogeneous dielectrics dielectric constant ambipolar conduction multiobjective genetic algorithm universal static logic cell logic circuits logic functions D-latch D-flip flops C

来源：评论

学校读者我要写书评

暂无评论

Design optimisation of an axial-flux reluctance magnetic coupling based on a two-dimensional semi-analytical model

引用

IET ELECTRIC POWER APPLICATIONS 2020年第5期14卷 901-910页

作者： Lubin, Thierry Vahaj, Amir Abbas Rahideh, Akbar Univ Lorraine Fac Sci & Technol Grp Rech Energie Elect Nancy F-54506 Vandoeuvre Les Nancy France Shiraz Univ Technol Dept Elect & Elect Engn Shiraz *** Iran

In this study, design optimisation of an axial-flux reluctance magnetic coupling is presented. The optimal design procedure is based on a two-dimensional (2D) semi-analytical model defined at the mean radius combined with a multi-objective genetic algorithm (NSGA-II). In order to take into account the end-effects in the radial direction, a correction factor is defined to improve the torque and the axial-force determination. The obtained results are compared with those of 3D non-linear finite element simulations and experimental results. It is shown that the proposed semi-analytical model is very accurate and requires very little computing time.

关键词： genetic algorithms finite element analysis magnetic flux couplings design engineering optimisation design axial-flux reluctance magnetic coupling two-dimensional semianalytical model multiobjective genetic algorithm axial-force determination 2D semianalytical model NSGA-II 3D nonlinear finite element simulations

来源：评论

学校读者我要写书评

暂无评论

Simultaneous filter tuning and calibration of the camera and inertial measurement unit camera for a vision inertial navigation system

引用

IET IMAGE PROCESSING 2020年第12期14卷 2756-2767页

作者： Arbabmir, Mohammadvali Ebrahimi, Masoud Tarbiat Modares Univ Dept Mech Engn Jalal AleAhmad Tehran Iran

In this study, a novel approach based on multi-objective genetic algorithm (MO-GA) is used for simultaneous tuning the unscented Kalman Filter (UKF) parameters and camera and inertial measurement unit camera calibration in a vision inertial navigation system (VINS). This system consists of visual odometry and inertial navigation system (INS) which integrates with a UKF. In order to obtain simultaneous tuning and calibration of the parameters and variables, the MO-GA minimises the root mean square error of the position and velocity of the vehicle on a selected trajectory of the benchmark data set. Then, the tuned parameters and calibrated variables are placed in the VINS and an adjusted VINS (AVINS) is obtained. For investigating the AVINS, the mentioned system is compared with INS only, VINS based on calibration data of the benchmark data set, and GPS/INS as Real Data on the identical trajectory. Furthermore, in order to evaluate the results of the proposed approach, the AVINS is examined in the second trajectory. The results indicate the proper performance of the presented approach in the simultaneous tuning the filter parameters and calibrating the variables of sensors that are used in the uncalibrated VINS.

关键词： cameras Global Positioning System genetic algorithms calibration Kalman filters inertial navigation nonlinear filters units (measurement) image sensors mean square error methods minimisation image filtering root mean square error minimisation adjusted vision inertial navigation system visual odometry unscented Kalman filter parameters MO-GA AVINS INS VINS inertial measurement unit camera calibration UKF multiobjective genetic algorithm

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：