Search

During the course of evolution, colonies of ants, bees, wasps, bacteria and termites have developed sophisticated behavior, intricate communication capabilities, decentralized colony control, group foraging strategies and a high degree of worker cooperation when tackling tasks. Utilizing these capabilities, any bio-inspired optimization techniques using analogy of swarming principles and social behavior in nature — swarm intelligence — have been adopted to solve a variety of engineering and robotics problems. In this paper, new approaches of bacterial colony optimization method with variable speed based on uniform, Gaussian, and Cauchy probability distribution were tested. Bacterial colony, a swarm intelligence methodology, is evaluated for a problem of optimization of a PID (proportional-integral-derivative) multivariable controller. The simulation results are compared with classical bacterial colony approach with constant velocity for a case study of control of a robotic manipulator system with two degree of freedom.

A new design for a model-free learning adaptive control (MFLAC), based on pseudo-gradient concepts with compensation using neural network, is presented in this paper. A radial basis function neural network using differential evolution optimization technique is applied to the control design. Motivation for developing a new approach is to overcome the limitation of the conventional MFLAC design, which cannot guarantee satisfactory control performance when the plant has different gains for the operational range. Robustness of the MFLAC with neural compensation scheme is compared to the MFLAC without compensation. Simulation results for a nonlinear chemical reactor are given to show the advantages of the proposed compensation approach.

Foraging theory originated in attempts to address puzzling findings that arose in ethological studies of food seeking and prey selection among animals. The potential utilization of biomimicry of social foraging strategies to develop advanced controllers and cooperative control strategies for autonomous vehicles is an emergent research topic. The activity of foraging can be focused as an optimization process. In this paper, a bacterial foraging approach for path planning of mobile robots is presented. Two cases study of static environment with obstacles are presented and evaluated. Simulation results show the performance of the bacterial foraging in different environments in the planned trajectories.

This paper approaches a computational solving of the industrial layout problem considering a set of hard-constraints not handled in previous works. The intention here has been to provide a new consistent benchmark problem that may help researchers to test their proposed algorithms. In order to handle the experiments, the computational optimization tool AVOLI (Visual Environment for Industrial Layout Optimization) is utilized. By using this computational tool, the provided hypothetic new facility layout problem mimicking a real one is solved in two steps: first a constructiveheuristic-based initial solution is generated and then Tabu Search (TS) heuristic is used to improve it.

The combination of wavelets with neural networks can hopefully remedy each others weaknesses, resulting in wavelet based neural network capable of handling system identification problems of a moderately large dimension. A wavelet based neural network is a nonlinear regression structure that represents nonlinear mappings as the superposition of dilated and translated versions of a function, which is found both in the space and frequency domains. In this paper, a wavelet-based neural network is introduced for the nonlinear identification of dynamic systems with chaotic behavior (chaotic time series). The structure of the wavelet based neural network is similar to that of radial basis function neural networks, except that here the activation function of the hidden nodes is replaced by wavelet functions. The proposed wavelet-based neural network is evaluated on two case studies: (i) the Henon map, and (ii) the Rossler system. Simulation results demonstrate the accuracy and the reliability of the proposed identification methodology based on a wavelet based neural network.

Model based predictive control has become an important form of advanced control. Generalised predictive controllers (GPCs) have been successfully applied to control complex processes in the industry during the last decade. In this paper, we present a novel method named modified particle swarm optimization, which is based on Gaussian and Cauchy distributions for the optimization of an adaptive GPC using an ARMA (Auto Regressive Moving Average) mathematical model. The simulation results demonstrated the successful GPC tuning and the suitability of the modified particle swarm optimization for the control of a non-linear process of Hammerstein type.

A prevenção de fraude em cartão de crédito é uma importante aplicação comercial para aplicação de abordagens de métodos de previsão e inteligência computacional. A inteligência computacional é uma associação de metodologias bio-inspiradas que incluem, como principais membros, as redes neurais, sistemas nebulosos, computação evolutiva, inteligência coletiva e computação probabilística. Recentemente, a aplicação de técnicas da inteligência computacional no suporte de tarefas de serviço, tais como detecção e identificação de anomalia, classificação de padrão, diagnóstico, prognóstico, estimação e controle, tem emergido nos ambientes industrial e comercial. Este artigo apresenta um sistema computacional eficiente baseado em metodologias da inteligência computacional para detecção de fraude em operações reais de dados de cartão de crédito em transações de comércio eletrônico. O novo método proposto é denominado FControl® de detecção de fraude e classificação integra conceitos de sistemas inteligentes híbridos baseado em abordagens de redes neurais, sistemas nebulosos e computação evolutiva. O método proposto gera soluções de qualidade em termos de eficiência e sucesso de previsão. O programa computacional do FControl® está atualmente instalado em um Intel Pentium IV 2.4 MHz (bi-processador e RAM de 4 Gbytes) na companhia Ciashop E-Commerce e está correntemente em uso para detecção de fraude para 250 companhias com serviços de comércio eletrônico. The prevention of credit card fraud is an important commercial application for prediction methods and computational intelligence of pattern recognition approaches. Computational intelligence is an association of bio-inspired computational methodologies which are founded principally on neural networks, fuzzy systems, evolutionary computation, swarm intelligence, and probabilistic computation. The application of computational intelligence techniques in the support of service tasks such as anomaly detection and identification, pattern classification, diagnostics, prognostics, estimation and control has recently emerged in industrial and commercial environments. This paper presents an efficient system based on computational intelligence methodologies for the detection of fraud in the operation of real credit card data in E-commerce transactions. This new proposed method, called FControl® for fraud detection and classification, integrates concepts of hybrid intelligent systems based on neural networks, fuzzy systems and evolutionary computation approaches. The proposed method generates quality solutions in terms of prediction efficiency and success. The computational program of the FControl® system has been installed on an Intel Pentium IV 2.4 MHz (bi-processor and 4 Gbytes RAM) at the company Ciashop E-Commerce and is currently in use for fraud detection by 250 companies that offer E-commerce services.

Este artigo apresenta uma nova abordagem para o projeto de um controlador PID (proporcional, integral e derivativo) multivariável baseado em uma rede neural e um algoritmo genético. O projeto do controlador PID multivariável é dividido em três etapas distintas. Na primeira etapa, uma rede neural de funções radiais de base é utilizada para identificação do processo multivariável. Na segunda etapa, o projeto do controlador é realizado, de forma off-line, baseado na sintonia de ganhos do controlador PID. Esta etapa é realizada através de uma otimização via algoritmos genéticos visando-se controlar o modelo do processo obtido pela rede neural de funções radiais de base. Na terceira etapa, os ganhos do controlador PID obtidos através de otimização genética e modelo neural são validados no processo prático. O desempenho desta proposta de projeto de controle multivariável é apresentado e discutido. O projeto do algoritmo de controle é validado de forma experimental em um processo multivariável denominado bola e chapa (ball-and-plate). Este processo consiste de uma chapa articulada em seu centro tal que a inclinação da chapa pode ser manipulada em duas direções perpendiculares. O objetivo básico é controlar a posição de uma bola que rola livremente sobre a chapa, aplicando-se tensões aos motores, estas baseadas no conhecimento da posição da bola medida pelo sistema de visão. This paper presents a new approach for the PID (proportional, integral, and derivative) multivariable controller design based on a neural network and a genetic algorithm. The multivariable PID controller design is divided in three steps. In the first step, a radial basis neural network is utilized for multivariable process identification. In the second step, the controller design is realized by an off-line procedure. This procedure is based on tuning of the PID controller gains by optimization with genetic algorithms aiming to control the model obtained by radial basis neural network. In the third step, the PID controller gains obtained with genetic optimization and neural model are validated in a practical process. Performance of this approach to multivariable control design is presented and discussed. The control algorithm design is validated of experimental way in a multivariable process called ball-and-plate. This process consists of a plate pivoted at its center such that the slope of the plate can be manipulated in two perpendicular directions. The basic control objective is to control the position of a free rolling ball on a plate, by applying tensions to two DC motors, according to the ball postion, as measured by vision system.

The differential evolution (DE) is an improved version of evolution strategies and Nelder-Mead simplex methods. DE has been successfully applied in various fields, such as optimization nonlinear functions, multi sensor fusion, control system, and system identification. The potentialities of DE are its simple structure, easy use, convergency speed and robustness. This paper proposes a new hybrid methodology for solving the economic load dispatch problem with valvepoint effect. The proposed hybrid method combinates the DE and sequential quadratic programming (SQP) technique. The DE is used to produce good potential solutions, and the SQP is used to fine-tune the DE run. The hybrid methodology and its variants are validated for a test system consisting of 40 thermal units whose incremental fuel cost function takes into account the valvepoint loading effects. The proposed hybrid method outperforms other state-of-theart algorithms in solving load dispatch problems with the valve-point effect.

Particle swarm optimization (PSO) is a population-based swarm intelligence algorithm that shares many similarities with evolutionary computation techniques. However, the PSO is driven by the simulation of a social psychological metaphor motivated by collective behaviours of bird and other social organisms instead of the survival of the fittest individual. Inspired by the classical PSO method and quantum mechanics theories, this work presents new Quantum-behaved PSO (QPSO) approaches using mutation operator with exponential probability distribution. The simulation results demonstrate good performance of the QPSO in solving a well-studied continuous optimization problem of mechanical engineering design.Copyright © 2006 by ASME

Two nonlinear identification methods are employed in this paper in an experimental comparative approach to generate dynamical models for a thermal-vacuum system. Used for space environment emulation and satellite qualification, a thermal-vacuum chamber presents highly nonlinear and time-delay characteristics. While, in the first nonlinear identification approach, Particle Swarm Optimization (PSO) derive a Takagi-Sugeno fuzzy model, the second one was based on NARMAX polynomial identification technique. PSO is a stochastic global optimization technique that uses a population of particles, where the position and velocity of each particle represent a solution to the problem. It is employed as an auxiliary mechanism for finding out a T-S fuzzy model. The NARMAX polynomial identification technique uses a criterion called Error Reduction Ratio (ERR) computed by employing an orthogonal least squares method whose terms are selected in a forward-regression manner. Results indicate that both methods are feasible solutions for eliciting models from the available data.Copyright © 2006 by ASME

Advanced conceptions to design industrial control systems are, in general, dependent of mathematical models of the controlled process. Also, the task of the controllers is to achieve an optimum performance when facing with various types of disturbances that are, sometimes, unknown in practical applications. An advanced methodology for control design is the variable structure control (VSC) with sliding mode. This paper considers the non-linear control of industrial processes based on self-tuning discrete VSC technique. A novel and systematic VSC design methodology is proposed, which integrates an estimator based on recursive least square algorithm, a discrete quasi-sliding surface and a optimization method. Contrary to the trial and error selection of the variable structure feedback gains reported in the literature, the selection in present work is done using differential evolution (DE) optimization, an evolutionary computation technique, with a new operator based on normative knowledge of cultural algorithms. The proposed design has been applied to a control valve described by a non-linear Wiener model. Numerical simulation results reveal that the new DE approach is applicable and promising for the VSC design.Copyright © 2006 by ASME

In spite of many efforts the PID (Proportional-Integral-Derivative) controller continues to be the main component in industrial control systems, included in the following forms: embedded and programmable logic controllers, and distributed control systems. From the viewpoint of simplicity and effectiveness, the PID controller represents a good solution for controlling many practical applications. HoweveR, the control literature still showing different tuning structures for the PID control to overcome complex dynamics and operational limitations [1]. Some difficult problems in the industry are presented in order to be answered or understood, not from the enthusiastic form neither from the absence of knowledge of plant operators or engineers, as the tuning aspects of a PID controller [2].

This paper presents a new discrete-time sliding-mode control design for multiple-input multi-output (MIMO) systems with tuning parameters by particle swarm optimization (PSO). PSO is a kind of evolutionary algorithm based on a population of individuals and motivated by the simulation of social behavior instead of the survival of the fittest individual. Several control algorithms are presented, two decoupling design and six new approaches of the coupling design of sliding-mode control without the necessity of calculate the process interactor matrix. SMC needs a design tool for parameter configuration and efficient practically to deal with multivariable processes. Simulations are carried out using both decoupling and coupling discrete-time SMC designs. Results shown that the new proposed approach for designing the discrete-time coupling SMC is a powerful tool and it performs better than the decoupling design, usually utilized in MIMO process. The simulations are assessed on a robotic manipulator of two degree-of-freedom (2-DOF), that constitute a MIMO nonlinear coupling dynamic system, with treatment of payload mass and link length variations. Simulation results show that the application of this control strategy effectively improve the trajectory tracking precision of position and velocity variables.

Variable structure control (VSC) technique is an important approach to deal with industrial control problems and is an area of very great interest. This paper presents a self-tuning discrete-time VSC design where the optimization task of controller parameters is realized by a new Lamarckian evolution approach based on evolutionary programming combined with a direct search method. Simulation results on a continuous stirred tank reactor, that presents open-loop unstable dynamic and nonlinear behaviors, are shown.

This paper presents a predictive control method based on switching among a set of mathematical models for a nonlinear experimental process. An adaptive predictive control configuration with multi-model identification scheme is appropriate to deal with systems subjected to sudden parameter changes or running at several operating points with different characteristics. The methodology uses a set of discrete-time mathematical models (multiple models) of the process obtained from an off-line technique by fast evolutionary programming with mutation operator based on Cauchy distribution. Experimental tests, identification and control tasks are carried out in a laboratory scale fan-and-plate plant. The good performance shown by the predictive controller based on multiple models confirms the usefulness and robustness of the proposed control method.

This paper presents the configuration of a fuzzy system of Takagi-Sugeno-Kang type based on optimization through a memetic algorithm. The memetic algorithm is composed by a fast evolutionary programming combined with simulated annealing algorithm. The fuzzy-memetic system is evaluated for two case studies: (i) prediction of chaotic system with maps involving non-differentiable functions called Lozi map, and (ii) identification of a system composed of a nonlinear continuous stirred tank reactor. Simulations deal the estimation and validation procedures of dynamic model based on fuzzy-memetic system. The performance of the fuzzy-memetic system for case studies in time series prediction and nonlinear identification are presented and discussed.

In this paper a new method for autotuning thePID controller gains based on fuzzy rules is proposed and a new rule base for gain and phase margins is derived. The proposed control scheme offers advantages over the conventional fuzzy PID controller (FPID) such as: i) it is necessary only one rule base; ii) it may be completely autotuned, requiring only one relay feedback experiment; and iii) it shows stability and the robustness characteristic is conceptually simple. The fuzzy PID control algorithm has been successfully applied to simulation and practical examples. The application of the describing function for the stability and design of FPID control system is investigated.

Generalized predictive controller (GPC) has robust performance when faced to control non-minimum phase and open-loop unstable processes. GPC has been originally developed with linear predictor models which leads to a formulation that can be solved analytically. If a nonlinear model is used then nonlinear optimization and nonlinear modeling techniques are necessary. This paper presents a new method of multivariable GPC design for nonlinear systems. The proposed GPC design incorporates the neural network model based on radial basis function neural network and optimization methodology of cost function by Lamarckian hybrid method using evolutionary programming and Nelder-Mead’s simplex algorithm. Steps of the control law implementation and intelligent procedure are discussed. Simulation results on a nonlinear MIMO (multi-input multi-output) system are presented to show the effectiveness of the neuro-evolutionary GPC method. The new GPC design is applicable for improve the approximation of nonlinear systems with unknown interactor matrix using neural network model and an efficient control law optimization task by simplex-evolutionary hybrid methodology.

Structures and design issues of fuzzy PID (proportional-integral-derivative) controllers (FLC-PID’s) are presented and evaluated in this paper. Configuration and basic characteristic of several structures of FLC-PID based on models proposed in the literature — (PD + I), (PI + D conventional), incremental (PD + I), (PD + PI) — are here reviewed and implemented. FLC-PID’s are assessed on a horizontal balance process, consisting of two propellers driven by two DC motors. Such process offers control complexities and can become unstable by using classical controllers. Experimental results, robustness and performance of FLC-PID’s are illustrated and discussed.

This paper evaluates methodologies of the computational intelligence applied to a nonlinear process identification. The different intelligent methodologies are: hybrid genetic algorithm with simulated annealing and modified Elman neural network. Experimental tests of identification are carried out using a laboratory scale wind tunnel plant. Experiments, for the estimation of the mathematical model of the process, are carried out using step function and pseudo-random binary sequence signals as process input. A white noise signal is used in the validation phase. The performance of each technique, in the estimation and validation phases of the process mathematical model, is presented and discussed.

This paper describes the application of Evolutionary Algorithms (EAs) to the optimisation of a simplified supply chain in an integrated production-inventory-distribution system. The performance of four EAs (Genetic Algorithm (GA), Evolutionary Programming (EP), Evolution Strategies (ES) and Differential Evolution (DE)) was evaluated with numerical simulations. Results were also compared with other similar approaches in the literature. DE was the algorithm that led to better results, outperforming previously published solutions. The robustness of EAs in general, and the efficiency of DE, in particular, suggest their great utility for the supply chain optimisation problem, as well as for other logistics-related problems.

Estudar modelagens através de dados geodésicos temporais com a possibilidade de predizer a posição de linha de costa é uma tarefa importante e pode auxiliar significativamente na gestão costeira. A área de estudo neste trabalho se refere ao município de Matinhos no estado do Paraná, Brasil. As linhas de costa temporais utilizadas para testar a modelagem preditiva são provenientes respectivamente da fotogrametria analógica para anos 1954, 1963, 1980, 1991 e 1997 e de levantamentos geodésicos utilizando GPS (Global Position System) para 2001, 2002, 2005 e 2008 (como controle). Dois testes com as redes neurais artificiais foram organizados mudando alguns parâmetros como: arquitetura, número de neurônios nas camadas ocultas e algoritmos de treinamentos. Quando comparados o valor dos resíduos entre a predição e a linha de costa de controle, os melhores resultados estatísticos indicam que o MAPE (mean absolute percentage error) são 0,28% utilizando a rede neural parcialmente recorrente de Elman com o algoritmo de treinamento quase-Newton e 0,46% para o caso da rede neural perceptron multicamadas com o algoritmo de treinamento utilizando o método Bayesiano com regularização. The study of models using geodetic temporal data which can possibly predict the shoreline position is an important task and can significantly contribute to coastal management. The studied area is located at municipality of Matinhos in the Paraná State, Brazil. The temporal shoreline used to test the prediction model is respectively from analog photogrammetric data, related to the years 1954, 1963, 1980, 1991 and 1997, and GPS (Global Position System) geodetic surveys for 2001, 2002, 2005 and 2008 (as control). Two different tests with artificial neural network were organized setting the parameters like: architecture, number of neuron in hidden layers and the training algorithms. Comparing the residuals between the prediction to the shoreline of control, the best statistical results show the MAPE (Mean Absolute Percentage Error) is 0,28% using the Elman partially recurrent network with quasi-Newton training function and 0,46% using the neural network multilayer perceptron with Bayesian regulation training function.

Nowadays, a variety of controllers used in process industries are still of the proportional–integral–derivative (PID) types. PID controllers have the advantage of simple structure, good stability, and high reliability. A relevant issue for PID controllers design is the accurate and efficient tuning of parameters. In this context, several approaches have been reported in the literature for tuning the parameters of PID controllers using evolutionary algorithms, mainly for single-input single-output systems. The systematic design of multi-loop (or decentralized) PID control for multivariable processes to meet certain objectives simultaneously is still a challenging task. This paper proposes a new chaotic firefly algorithm approach based on Tinkerbell map (CFA) to tune multi-loop PID multivariable controllers. The firefly algorithm is a metaheuristic algorithm based on the idealized behavior of the flashing characteristics of fireflies. To validate the performance of the proposed PID control design, a multi-loop multivariable PID structure for a binary distillation column plant (Wood and Berry column model) and an industrial-scale polymerization reactor are taken. Simulation results indicate that a suitable set of PID parameters can be calculated by the proposed CFA. Besides, some comparison results of a genetic algorithm, a particle swarm optimization approach, traditional firefly algorithm, modified firefly algorithm, and the proposed CFA to tune multi-loop PID controllers are presented and discussed.

Mathematical models are often used to simplify landform representation. Its importance is due to the possibility of describing phenomena by means of mathematical models from a data sample. High processing power is needed to represent large areas with a satisfactory level of details. In order to accelerate the solution of complex problems, it is necessary to combine two basic components in heterogeneous systems formed by a multicore with one or more GPUs. In this paper, we present a methodology to represent landform attributes on multicore and multi-GPU systems using high performance computing techniques for efficient solution of two-dimensional polynomial regression model that allow to address large problem instances.

CNPq - Conselho Nacional de Desenvolvimento Científico e Tecnológico CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior Building empirical models from data is of fundamental importance in engineering, and pushing the boundaries of current engineering technology requires us to model and understand nonlinear systems. In this thesis, nonlinear differentiable models and its applications are studied. This class of models has gained traction in machine learning tasks with the introduction of deep learning. Indeed, deep models of stacked differentiable components have recently achieved super-human performance on several tasks, including computer games, image classification, and medical diagnosis. The application of nonlinear differentiable models is studied for modeling signals and systems both for engineering and machine learning applications. One central question is the role of recurrence and the pros and cons of recurrent and feedforward models. The question is approached from more than one angle: 1) by studying the effect of recurrence in neural networks in terms of robustness to noise, computational cost, and convergence; 2) by analyzing the smoothness of the cost function in nonlinear system identification problems and its relation to the model internal dynamics – and proposing the use of a technique called multiple shooting for improving the cost-function smoothness; and, 3) by investigating the interplay between the internal dynamics, the attractors and the expressiveness of the model in deep recurrent neural networks. The more applied part of the thesis consists of the use of deep neural networks to solve complex tasks and to model nonlinear behavior from real data. Data from the Telehealth Center of Minas Gerais is used to train a deep neural network capable of identifying abnormalities in the electrocardiogram exam with performance superior to the medical residents in the studied scenario. Also, a deep neural network is used for modeling an electronic oscillator and an F-16 aircraft using data from ground vibration experiments, obtaining competitive results in both cases. Construir modelos empíricos a partir de dados é de fundamental importância em engenharia e, além disso, o entendimento e a capacidade de modelar sistemas não lineares são necessários para o desenvolvimento de tecnologias de fronteira. Nesse trabalho, modelos diferenciáveis não lineares e suas aplicações são estudados. Esta classe de modelos tem ganhado força na área de aprendizado de máquina com a introdução do aprendizado profundo. De fato, modelos profundos de componentes diferenciáveis alcançaram, recentemente, desempenho superior ao humano em diversas tarefas, incluindo a competição em jogos digitais, classificação de imagens e diagnóstico de exames médicos. A aplicação de modelos não lineares diferenciáveis é estudada para modelar sinais e sistemas, tanto no contexto de aplicações em engenharia quanto no contexto de aprendizado de máquina. Uma questão central é o papel da recorrência, e os prós e os contras de modelos recorrentes. A questão é abordada de mais de um ângulo: 1) estudando o efeito da recorrência em redes neurais em termos da robustez a ruído, custo computacional e convergência; 2) analisando a suavidade da função de custo na identificação de sistemas não lineares e a relação com a dinâmica interna do modelo – e propondo o uso da técnica de múltiplos tiros para melhorar a suavidade da função custo; e, 3) investigando a relação entre dinâmica interna, atractores e expressividade do modelo em redes neurais recorrentes. A parte mais aplicada desta tese consiste no uso de redes neurais profundas para resolver tarefas complexas e modelar comportamento não linear a partir de dados reais. Dados do Centro de Telessaúde do estado de Minas Gerais são usados para treinar uma rede neural capaz de identificar abnormalidades no eletrocardiograma com desempenho superior ao de residentes de medicina no cenário estudado. Além disso, uma rede neural profunda é usada para modelar um oscilador eletrônico e uma aeronave F-16 usando dados de um ensaio de vibrações, obtendo resultados competitivos nos dois casos.