检索结果-内蒙古大学图书馆

Modelling hybrid and backpropagation adaptive neuro-fuzzy inference systems for flood forecasting

NATURAL HAZARDS 2021年第1期108卷 519-566页

作者： Tabbussum, Ruhhee Dar, Abdul Qayoom Natl Inst Technol Srinagar Dept Civil Engn Srinagar 190006 Jammu & Kashmir India

The ability of the adaptive neuro-fuzzy inference algorithm architecture to simulate floods is explored in this research. The development of models for flood forecasting has been centered on two adaptive neuro-fuzzy inference (ANFIS) algorithms. The Takagi-Sugeno fuzzy inference systems (FIS) generated through subtracted clustering were trained using hybrid and backpropagation training algorithms. Multiple statistical performance evaluators were used to assess the performability of the established models. The validity and predictive power of the models are evaluated by estimating a flood occurrence in the study area. In designing the models, a total of 12 inputs were employed. The best performability was found for the ANFIS model created utilizing a hybrid training algorithm with mean square error (MSE) of 0.00034, co-efficient of correlation (R-2) of 97.066%, root mean square error (RMSE) of 0.018, Nash-Sutcliffe model efficiency (NSE) of 0.968, mean absolute error (MAE) of 0.0073 and combined accuracy (CA) of 0.018, indicating the possible usage of exploiting the established model for prediction of floods.

关键词： Adaptive neuro-fuzzy inference system backpropagation learning algorithm Gaussian membership function Hybrid learning algorithm Subtractive clustering Takagi– Sugeno fuzzy inference system

来源：评论

学校读者我要写书评

暂无评论

Distributed Neural Observer-Based Formation Strategy of Non-Affine Nonlinear Multi-Agent Systems with Unknown Dynamics

引用

JOURNAL OF CIRCUITS SYSTEMS AND COMPUTERS 2021年第5期30卷

作者： Manouchehri, Pejman Ghasemi, Reza Toloei, Alireza Mohammadi, Fazel Islamic Azad Univ Elect Engn Dept Damavand Branch Tehran Iran Univ Qom Elect Engn Dept Qom Iran Shahid Beheshti Univ Fac Engn Mech Engn Dept Tehran Iran Univ Windsor Elect & Comp Engn Dept Windsor ON N9B 1K3 Canada

The state estimation in Multi-Agent Systems (MASs) is a challenging problem. This is due to the fact that (1) controlling nonaffine nonlinear MASs is a difficult task and also (2) the agents in MASs have direct impacts on each other. This paper presents a new distributed Neural Networks (NN) observer for the nonlinear dynamical model of MASs with nonaffine unknown dynamical agents. The proposed scheme uses the backpropagation learning algorithm to estimate the unknown nonlinear functions of the agents. Compared with the previous studies, which primarily concentrated on the observer design for Multiple Input Multiple Output (MIMO) systems, the proposed method is applied to nonaffine nonlinear MASs. The advantages of this method are the overall stability, the fast convergence of the observer error to zero and the robustness against both uncertainties and disturbances. Nonlinear flexible-joint robots and nonlinear dynamic duffing chaotic systems are simulated to demonstrate the effectiveness and robustness of the proposed method. The proposed method is also compared with the Luenberger observer. The guaranteed stability, better performance in the presence of agents' uncertainties, robustness against disturbances are the main advantages of the proposed method compared with the traditional observer.

关键词： backpropagation learning algorithm chaotic system distributed neural observer nonaffine nonlinear systems Multi-Agent System (MAS)

来源：评论

学校读者我要写书评

暂无评论

SOLAR RADIATION PREDICTION BASED ON RECURRENT NEURAL NETWORKS TRAINED BY LEVENBERG-MARQUARDT backpropagation learning algorithm

SOLAR RADIATION PREDICTION BASED ON RECURRENT NEURAL NETWORK...

引用

IEEE PES Innovative Smart Grid Technologies

作者： Nian Zhang Pradeep K. Behera Department of Electrical and Computer Engineering University of the District of Columbia Department of Civil and Mechanical Engineering University of the District of Columbia

ISBN: (纸本)9781457721588

In response to the growing concern over the use of fossil fuels, renewable energy industries have been significant economic drivers in many parts of the United States. In the recent years there is a strong growth in solar power generation industries that requires prediction of solar energy to develop highly efficient stand-alone photovoltaic systems as well as hybrid power systems. In order to accomplish the goal, we propose a predictive model that is based on recurrent neural networks trained with the Levenberg-Marquardt backpropagation learning algorithm to forecast the solar radiation using the past solar radiation and solar energy. This computational intelligence modeling tool explored the impact of solar radiation and solar energy in forecasting reliable long-run solar energy. Based on the excellent experimental results including the mean squared error analysis, error autocorrelation function analysis, regression analysis, and time series response, it demonstrated that the proposed neural network structure and the learning algorithm could be very useful in training the recurrent neural network for the solar radiation prediction.

关键词： Solar radiation prediction Time series prediction Neural networks backpropagation learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Artificial neural network approach for a class of fractional ordinary differential equation

引用

NEURAL COMPUTING & APPLICATIONS 2017年第4期28卷 765-773页

作者： Jafarian, Ahmad Mokhtarpour, Masoumeh Baleanu, Dumitru Islamic Azad Univ Urmia Branch Dept Math Orumiyeh Iran Cankaya Univ Dept Math Eskisehir Yolu 29 Km TR-06810 Ankara Turkey Inst Space Sci Bucharest Romania

The essential characteristic of artificial neural networks which against the logistic traditional systems is a data-based approach and has led a number of higher education scholars to investigate its efficacy, during the past few decades. The aim of this paper was concerned with the application of neural networks to approximate series solutions of a class of initial value ordinary differential equations of fractional orders, over a bounded domain. The proposed technique uses a suitable truncated power series of the solution function and transforms the original differential equation in a minimization problem. Then, the minimization problem is solved using an accurate neural network model to compute the parameters with high accuracy. Numerical results are given to validate the iterative method.

关键词： Feed-forward neural network Fractional differential equation Approximate solution backpropagation learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Application of machine learning techniques for solving real world business problems: the case study - target marketing of insurance policies

Application of machine learning techniques for solving real ...

引用

作者： Juozenaite, Ineta Universidade Nova de Lisboa

学位级别：硕士

The concept of machine learning has been around for decades, but now it is becoming more and more popular not only in the business, but everywhere else as well. It is because of increased amount of data, cheaper data storage, more powerful and affordable computational processing. The complexity of business environment leads companies to use data-driven decision making to work more efficiently. The most common machine learning methods, like Logistic Regression, Decision Tree, Artificial Neural Network and Support Vector Machine, with their applications are reviewed in this work. Insurance industry has one of the most competitive business environment and as a result, the use of machine learning techniques is growing in this industry. In this work, above mentioned machine learning methods are used to build predictive model for target marketing campaign of caravan insurance policies to achieve greater profitability. Information Gain and Chi-squared metrics, Regression Stepwise, R package "Boruta", Spearman correlation analysis, distribution graphs by target variable, as well as basic statistics of all variables are used for feature selection. To solve this real-world business problem, the best final chosen predictive model is Multilayer Perceptron with backpropagation learning algorithm with 1 hidden layer and 12 hidden neurons.

关键词： Machine learning Logistic Regression Decision Tree CART Artificial Neural Network Multilayer Perceptron backpropagation learning algorithm Support Vector Machine Kernel Gaussian Radial Basis Function Target Marketing Insurance Policies Information Gain Chi-squared metric Regression Stepwise Stratified Data Partitioning Spearman Correlation

来源：评论

学校读者我要写书评

暂无评论

Time Series Forecasting With Orthogonal Endocrine Neural Network Based on Postsynaptic Potentials

引用

JOURNAL OF DYNAMIC SYSTEMS MEASUREMENT AND CONTROL-TRANSACTIONS OF THE ASME 2017年第4期139卷 041006-041006页

作者： Milovanovic, Miroslav Antic, Dragan Milojkovic, Marko Nikolic, Sasa S. Spasic, Miodrag Peric, Stanisa Univ Nis Fac Elect Engn Dept Control Syst Aleksandra Medvedeva 14 Nish 18000 Serbia

This paper presents a new type of endocrine neural network (ENN). ENN utilizes artificial glands which enable the network to be adaptive to external disturbances. Sensitivity is controlled by the hormone decay rate and the value of the sensitivity parameter. The network presented in this paper is improved by making the sensitivity parameter selftuning and implementing orthogonal activation functions inside the network structure. Automatic tuning is performed on the basis of the biological principle of postsynaptic potentials by implementing inhibitory and excitatory glands inside the standard backpropagation learning algorithm of developed orthogonal ENN. These additional network functionalities enable extra sensitivity to external conditions and an additional network feature of activation sharpening. The network was tested on real-time series of experimental data with a purpose to forecast exchange rate of the three widely used international currencies.

关键词： endocrine neural network time series forecasting orthogonal functions postsynaptic potential sensitivity parameter backpropagation learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Applications of Ground-Penetrating Radar (GPR) to Detect Hidden Beam Positions

引用

JOURNAL OF TESTING AND EVALUATION 2017年第3期45卷 911-921页

作者： Kilic, Gokhan Izmir Univ Econ Dept Civil Engn Izmir Turkey

Ground-penetrating radar (GPR) uses electromagnetic waves to investigate the structures. In this investigation method, an electromagnetic wave is transmitted using an antenna and the received signal is recorded. Detection of beam positions in this GPR data requires the skills of a trained human operator. This study utilized a multi-layer neural network to detect beam positions in the GPR data. The visual description and definition of GPR data has major disadvantages and a neural network has been studied to overcome these shortcomings. A set of 32,740 training vectors with a length of 64 data was implemented to train the neural network. A new set of 16,370 testing vectors with a length of 64 data was then prepared to test the performance. Testing results suggest that the neural network is promising methods for the detection of beam positions in the GPR data.

关键词： backpropagation learning algorithm Bayes optimal decision rule Gram-Charlier series GPR and data processing neural network

来源：评论

学校读者我要写书评

暂无评论

Polynomial Based Functional Link Artificial Recurrent Neural Network adaptive System for predicting Indian Stocks

引用

INTERNATIONAL JOURNAL OF COMPUTATIONAL INTELLIGENCE SYSTEMS 2015年第6期8卷 1004-1016页

作者： Bebarta, D. K. Biswal, Birendra Dash, P. K. GMR Inst Technol Rajam AP India Siksha O Anusandhan Univ Bhubaneswar Odisha India

A low complexity Polynomial Functional link Artificial Recurrent Neural Network (PFLARNN) has been proposed for the prediction of financial time series data. Although different types of polynomial functions have been used for low complexity neural network architectures earlier for stock market prediction, a comparative study is needed to choose the optimal combinations of the nonlinear functions for a reasonably accurate forecast. Further a recurrent version of the Functional link neural network is used to model more accurately a chaotic time series like stock market indices with a lesser number of nonlinear basis functions. The proposed PFLARNN model when trained with the well known gradient descent algorithm produces reasonable accuracy with a choice of range of weight parameters of the network. However, to improve the accuracy of the forecast further, the weight parameters of the recurrent functional neural network are optimized using an evolutionary learning algorithm like the differential evolution (DE). A comparison with other well known neural architectures shows that the proposed low complexity neural model can provide significant prediction accuracy for one day advance and speed of convergence using the International Business Machines Corp. (IBM) stock market indices.

关键词： PFLARNN Polynomial functions backpropagation learning algorithm differential evolution IBM stock indices MAPE AMAPE

来源：评论

学校读者我要写书评

暂无评论

Intelligent controlled three-phase squirrel-cage induction generator system using wavelet fuzzy neural network for wind power

引用

IET RENEWABLE POWER GENERATION 2013年第5期7卷 552-564页

作者： Lin, Faa-Jeng Tan, Kuang-Hsiung Fang, Dun-Yi Lee, Yih-Der Natl Cent Univ Dept Elect Engn Chungli 320 Taiwan Natl Def Univ Chung Cheng Inst Technol Sch Def Sci Tao Yuan 335 Taiwan Inst Nucl Energy Res Engn Technol & Facil Operat Div Tao Yuan 335 Taiwan

An intelligent controlled three-phase squirrel-cage induction generator (SCIG) system for grid-connected wind power application using wavelet fuzzy neural network (WFNN) is proposed in this study. First, the indirect field-oriented mechanism is implemented for the control of the SCIG system. Then, an AC/DC power converter and a DC/AC power inverter are developed to convert the electric power generated by a three-phase SCIG from variable-voltage and variable-frequency to constant-voltage and constant-frequency. Moreover, the intelligent WFNN controller is proposed for both the AC/DC power converter and DC/AC power inverter to improve the transient and steady-state responses of the SCIG system at different operating conditions. Three online trained WFNNs using backpropagation learning algorithm are implemented as the tracking controllers for the DC-link voltage of the AC/DC power converter and the active power and reactive power outputs of the DC/AC power inverter. Furthermore, the network structure and the online learning algorithm of the WFNN are introduced in detail. Finally, some experimental results are provided to demonstrate the effectiveness of the proposed SCIG system for wind power.

关键词： AC-DC power convertors asynchronous generators backpropagation DC-AC power convertors electric current control fuzzy neural nets machine vector control neurocontrollers phase control power generation control power grids reactive power control wavelet transforms wind power plants Intelligent controlled three-phase squirrel-cage induction generator system wavelet fuzzy neural network SCIG grid-connected wind power application indirect field-oriented mechanism AC-DC power converter DC-AC power inverter electric power generation variable-voltage variable-frequency constant-voltage constant-frequency intelligent WFNN controller WFNN training backpropagation learning algorithm tracking controller DC-link voltage controller reactive power output active power output

来源：评论

学校读者我要写书评

暂无评论

A PSO based integrated functional link net and interval type-2 fuzzy logic system for predicting stock market indices

引用

APPLIED SOFT COMPUTING 2012年第2期12卷 931-941页

作者： Chakravarty, S. Dash, P. K. Siksha OAnusandhan Univ Bhubaneswar Orissa India Reg Coll Management Autonomous Bhubaneswar Orissa India

This paper presents an integrated functional link interval type-2 fuzzy neural system (FLIT2FNS) for predicting the stock market indices. The hybrid model uses a TSK (Takagi-Sugano-Kang) type fuzzy rule base that employs type-2 fuzzy sets in the antecedent parts and the outputs from the Functional Link Artificial Neural Network (FLANN) in the consequent parts. Two other approaches, namely the integrated FLANN and type-1 fuzzy logic system and Local Linear Wavelet Neural Network (LLWNN) are also presented for a comparative study. backpropagation and particle swarm optimization (PSO) learning algorithms have been used independently to optimize the parameters of all the forecasting models. To test the model performance, three well known stock market indices like the Standard's & Poor's 500 (S&P 500), Bombay stock exchange (BSE), and Dow Jones industrial average (DJIA) are used. The mean absolute percentage error (MAPE) and root mean square error (RMSE) are used to find out the performance of all the three models. Finally, it is observed that out of three methods, FLIT2FNS performs the best irrespective of the time horizons spanning from 1 day to 1 month. (C) 2011 Elsevier B. V. All rights reserved.

关键词： Functional Link Artificial Neural Network Fuzzy logic system Interval Type-2FLS backpropagation learning algorithm Particle swarm optimization

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

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

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：