Optimal design of double folded stub microstrip filter by neural network modelling and particle swarm optimization

Abstract

Abstract-Optimization of design parameters based on electromagnetic simulation of microwave circuits is a timeconsuming and iterative procedure. To provide a fast and accurate frequency response for a given case study, this paper employs a neural network modelling approach. First, one of the case study's outputs, i.e., scattering parameter (|S 21 |) in dB, is predicted using a neural network model. Then the particle swarm optimization is employed to optimize the design parameters. The proposed method in designing the filter compares with two others methods for a case study. The simulation results show the capability of the proposed method in designing an optimized filter in a proper time.Index Terms-neural networks, Microwave structures, particle swarm optimization, modelling.I. INTRODUCTION Neural networks are information processing systems inspired by the human brain's ability to learn from observations [1]. Neural networks are efficient alternatives to conventional methods, such as numerical modelling methods which could be computationally expensive, or analytical methods which could be difficult to obtain, or empirical models whose range and accuracy could be limited.Due to their ability and adaptability to learn, generalizability, fast real-time operation, and ease of implementation, they have been used in microwave design problems, RF, and microwave computeraided design problems [2].In microwave applications, a neural network can train to mimic the electrical behavior of the circuit. Then, it can be used in high-level simulation and design, providing a fast response to the learned task [3], [4]. Moreover, neural networks can be employed along with evolutionary algorithms such as genetic algorithm [5] and particle swarm optimization [6] to optimize the system under study.In this application, a neuarl network can use with an electromagnetic packages, such as HFSS, CST, ADS, to optimize the design.In this paper, to accelerate the frequency sweep required in the full-wave simulators, for calculating the S-parameters of a double folded stub microstrip filter, an artificial neural network is employed.The neural network is trained, by achieved patterns from an electromagnetic package, to approximate the S-parameters in the region of interest. The proposed model has capability to perform the frequency Sistan and Baluchestan, Iran Email: amirbanookh@ece.usb.ac.ir, smbaraka@ece.usb.ac.ir Journal of Microwaves, Optoelectronics and Electromagnetic Applications, Vol. 11, No. 1, June 2012 Brazilian The neural network structure is explained in the next section. Section 3 is dedicated to describe the of particle swarm optimization. The case study is introduced in section 4. To verify the proposed model, simulation results are illustrated in section 5. The last section is conclusions.II. THE STRUCTURE OF NEURAL NETWORK In this section, the structure of neural networks and its ability to model the behavior of RF and microwave components are described.The MLP is a commonly used neural network structure, in which th...

Nonlinear Autoregressive Neural Network Models for Prediction of Transformer Oil-Dissolved Gas Concentrations

“…The superior performance presented was independent of the delay parameter, d, used in the model. This result is coherent with the literature that states that nonlinear autoregressive network models are less sensitive to long-term time dependencies, besides presenting better generalization and learning capacities [22]. Moreover, an additional increase in performance is due to the application of the discrete wavelet transform to create simplified and sparse versions of the original data.…”

“…The Artificial Neural Network (ANN) is a mathematical method that aims to simulate the human brain in the knowledge acquisition process, with successful applications in nonlinear mapping between input and output variables, pattern recognition and classification, optimization, just to name a few [20][21][22].…”

Metodologia para a redução da vibração e da ondulação de torque de um motor de relutância chaveado especial a partir do uso de simulações multifásicas e de algoritmos genéticos.

“…Inspirada na biologia de cérebro humano, a rede neural tem alguns elementos básicos como neurônios artificiais, sinapses, pesos neurais e funções de transferência [78], [79]. Redes neurais artificiais são extensamente utilizadas como aproximadores universais de função, reconhecimento e classificação de padrões, sistemas de otimização, entre outras aplicações [80], [81]. Neste caso, foram consideradas como variáveis de entrada os ângulos de comutação (ϴ on e ϴ c ) e, como variável de saída, o nível de vibração (aceleração) medido na frequência de 400 Hz (componente fundamental da ondulação de torque) [1].…”