Loading...
50 results
Search Results
Now showing 1 - 10 of 50
- An efficient parallel implementation of a least squares problemPublication . Ruano, Antonio; Fleming, P. J.; Jones, D. I.Least squares solutions are a very important problem, which appear in a broad range of disciplines (for instance, control systems, statistics, signal processing). Our interest in this kind of problems lies in their use of training neural network controllers.
- Applications of neural networks to control systemsPublication . Ruano, AntonioThis work investigates the applicability of artificial neural networks to control systems. The following properties of neural networks are identified as of major interest to this field: their ability to implement nonlinear mappings, their massively parallel structure and their capacity to adapt. Exploiting the first feature, a new method is proposed for PID autotuning. Based on integral measures of the open or closed loop step response, multilayer perceptrons (MLPs) are used to supply PID parameter values to a standard PID controller. Before being used on-line, the MLPs are trained offline, to provide PID parameter values based on integral performance criteria. Off-line simulations, where a plant with time-varying parameters and time varying transfer function is considered, show that well damped responses are obtained. The neural PID autotuner is subsequently implemented in real-time. Extensive experimentation confirms the good results obtained in the off-line simulations. To reduce the training time incurred when using the error back-propagation algorithm, three possibilities are investigated. A comparative study of higherorder methods of optimization identifies the Levenberg-Marquardt (LM)algorithm as the best method. When used for function approximation purposes, the neurons in the output layer of the MLPs have a linear activation function. Exploiting this linearity, the standard training criterion can be replaced by a new, yet equivalent, criterion. Using the LM algorithm to minimize this new criterion, together with an alternative form of Jacobian matrix, a new learning algorithm is obtained. This algorithm is subsequently parallelized. Its main blocks of computation are identified, separately parallelized, and finally connected together. The training time of MLPs is reduced by a factor greater than 70 executing the new learning algorithm on 7 Inmos transputers.
- A neural network PID autotunerPublication . Ruano, Antonio; Lima, João; Mamat, R.; Fleming, P. J.Proportional, Integral and Derivative (PID) regulators are standard building blocks for industrial automation. Their popularity comes from their rebust performance and also from their functional simplicity. Whether because the plant is time-varying, or because of components ageing, these controllers need to be regularly retuned.
- Comparison of alternative approaches to neural network PID autotuningPublication . Ruano, Antonio; Lima, João; Mamat, R.; Fleming, P. J.In this paper, a scheme for the automatic tuning of PID controllers on-line, with the assistance of trained neural networks, is proposed. The alternative approaches are presented and compared.
- Recent developments in neural network PID autotuningPublication . Ruano, AntonioPID controllers are widely used in industrial applications. Because the plant can be time variant, methods of autotuning of this type of controllers, are of great economical importance, see (Astrom, 1996). Since 1942, with the work of Ziegler and Nichols (Ziegler and Nichols, 1942), several methods have been proposed in the literature. Recently, a new technique using neural networks was proposed (Ruano et al., 1992). This technique has been shown to produce good tunings as long as certain limitations are met.
- B-splines neural networks assisted PID autotuningPublication . Azevedo, Ana Beatriz da Piedade de; Ruano, AntonioThis paper describes previous works (1), (2), on neural network pid autotuning. Basically, neural networks are employed to supply PID parameters, according to the ITAE criterion, to a standard PID controller.
- Parallel implementation of a learning algorithm for multilayer perceptrons using transputersPublication . Ruano, Antonio; Jones, D. I.; Fleming, P. J.In this paper the parallelization of a new learning algorithm for multilayer perceptrons, specifically targeted for nonlinear function approximation purposes, is discussed. Each major step of the algorithm is parallelized, a special emphasis being put in the most computationally intensive task, a least-squares solution of linear systems of equations.
- A computational study of a parallel Branch and Bound algorithm for the quadratic 0-1 programming problem on transputersPublication . Schutz, G.; Pires, F. M.; Ruano, AntonioDiscrete optimization problems are very difficult to solve, even if the dimantion is small. For most of them the problem of finding an ε-approximate solution is already NP-hard.
- Dynamic temperature models of a soiless greenhousePublication . Cunha, J. B.; Ruano, Antonio; Faria, E. A.; Couto, C.In this paper climate discrete-time dynamic models for the inside air temperature of a soilless greenhouse are identified, using data acquired during two different periods of the year. These models employ data from air temperature and relative humidity.
- Estudo computacional de um algoritmo genético para o problema de optimização de rotas de veículosPublication . Schutz, G.; Pires, F. M.; Ruano, AntonioO problema básico da distribuição e/ou recolha de produtos é um problema de Optimização Combinatória que consiste em determinar o conjunto de rotas que partem de um depósito central, cuja localização é conhecida, servem um conjunto de clientes com procuras e localizações pré - definidas, minimizando a distância total percorrida. Este é um problema NP-difícil, para o qual poucos métodos exactos foram desenvolvidos, sendo demasiado demorados e não sendo sequer exequíveis para a generalidade dos problemas de dimensão média. Assim, as abordagens mais comuns e eficientes baseiam-se em métodos heurísticos. Numa classificação superficial podem-se considerar duas classes de métodos heurísticos: os “clássicos” e os “modernos”. Os primeiros incluem, entre outros, os métodos construtivos de rota, os métodos de duas fases e os de melhoramento de rotas. Estes métodos foram basicamente desenvolvidos nos anos 60 e 70, embora continuem a ser utilizados e melhorados. Por outro lado, os métodos “modernos” têm como característica comum o recurso à pesquisa local utilizando técnicas de intensificação e diversificação dessa pesquisa. Entre outras, têm aparecido recentemente heurísticas baseadas em: Simulação de Têmpera; algoritmos Genéticos; Pesquisa Tabu; Algoritmos Difusos e Redes Neuronais. Combinações destas técnicas têm conduzido a algoritmos híbridos e a meta-heurísticas. Neste trabalho é feita a apresentação de um Algoritmo Genético para o Problema de Optimização de Rotas de Veículos. Apresenta-se a sua implementação e um estudo computacional comparativo com duas heurísticas “clássicas” num conjunto de problemas - teste conhecidos da literatura.