Multilayer volterra filter and its applications

The paper presents the application of the successive linearisation to the neural network implementation of the Volterra filter. Applying the signal flow graph approach, the new learning rules for adaptation of weights of the multilayer network structure obtained are given. The multilayer structure presented is applied to signal processing, including identification of the parameters of the plant, noise cancelling and signal prediction. The results of the simulation of the filter in these three basic application modes are presented and discussed.     

Frequency-multiple generation technique using Chebyshev polynomials

A frequency-multiple generation technique based on Cheby?shev polynomials is presented in this letter. The circuit contains a minimum number of multipliers and operational-amplifier summers.     

Computer simulation of SC circuits using NAP

The adaptation of the NAP program to the frequency analysis of SC circuits is presented in the letter. According to this approach the equivalent circuit contains resistances and inductances or capacitances as the representations of the switched or unswitched capacitors in the frequency domain.     

Modified Method for Dynamic Stress‐Strain Curve Determination of Closed‐Cell Foams

The paper presents a modified method for the non‐linear dynamic stress‐strain curve determination with the use of cushion curves provided by the manufacturers of closed‐cell foams. This method consists of two main stages: stage 1 – the cushion curves transformation to the coordinate system of dynamic stress‐energy density, combined with the cubic b‐spline curve approximation, and stage 2 – development of the dynamic stress‐strain curve considering that the area under the curve is equal to the energy density of a collision. The stress‐strain curves are one of the basic parameters of the foam material models used to describe the collision process, in the modelling environments of finite element method (FEM), e.g. LS‐DYNA. Performed verification consisting in the reconstruction of cushion curves using LS‐DYNA showed that the proposed method of the dynamic stress‐strain curves determination is efficient and effective. Taking into account that access to the stress‐strain curves is hindered (cushioning material manufacturers do not publish them), and the access to charts of cushion curves is easier, the proposed method for determining the stress‐strain curves can relevantly shorten the process of preparing the input data, which are required during modelling of spatial structures using FEM. Copyright © 2016 John Wiley & Sons, Ltd.