Papers by Keyword: Nonlinear Modeling

Abstract: This study describes a practical and systematic procedure for identifying and modeling nonlinear systems based on the input-output analysis. Frequency domain data is used to obtain a reduced linear models of nonlinear systems. A coherence function is introduced to specify the identification accuracy. The procedure is applied to a small vertical take-off and landing air vehicle. Control compensators are then designed based on the identified models and autonomous hovering is successfully achieved. Simulation results demonstrate the effectiveness and superiority of this method in comparison with other classical approaches.

Abstract: An objective reliability analysis of structural members made of advanced cementitious composites must be based on good knowledge of stochastic properties of individual mechanical fracture parameters of utilized material models. The article presents a comprehensive approach to the design and assessment of precast structural elements including: The series of fracture tests of the two concrete mixtures with various ages in two configurations (three point bending and wedge splitting test, subsequent identification of material parameters using effective crack model, work of fracture method and artificial neural networks, execution of destructive tests of scaled structural members and creation of deterministic models of these tests using collected data. In subsequent phases of the project reliability analysis of tested beams will be carried out in order to obtain stochastic parameters of structural response of prestressed elements to shear load. The obtained data will be used to calibrate the analytical equation describing the response of element exposed to both normal and shear forces. The entire process will be concluded by reliability-based optimization of manufactured components.

Abstract: To improve the modeling performance of Recurrent Wavelet Neural Network (RWNN), a training algorithm based on Immune Evolving Algorithm (IEA) is proposed. In the process of RWNN training, IEA is mainly used to optimize the connection weight, translating and scaling parameter. The experiment result on Duffing chaotic time series shows that the proposed RWNN training algorithm has a good prediction capability in the field of nonlinear modeling.

Abstract: In this paper a new seismic base isolator, called High Damping Hybrid Seismic Isolator (HDHSI), is proposed. It is obtained by the assembly in series of a Lead Rubber Bearing (LRB) and a Friction Slider (FS) with a high friction coefficient. The HDHSI device is in contrast with the Resilient-Friction Base Isolator (R-FBI) with the aim of optimizing the Electricité De France (EDF) system. The mathematical model of a structure base isolated by a HDHSI system is analyzed with a two Degree of Freedom System (2-DOF) in which the superstructure is assimilated to a rigid body. Nonlinear finite elements are adopted for modeling the HDHSI device. A dynamic nonlinear analysis is performed and the hysteretic cycles are derived and evaluated for the single components and for the innovative HDHSI device.

Abstract: In this paper we have analyzed the influence of the strain hardening behavior of High Damping Rubber Bearings (HDRBs) adopted for a base isolation system of a Reinforced Concrete (RC) isolated structure. For the modeling of the rubber isolators we have adopted an evolution of the Bouc-Wen’s hysteretic model taking into account the incremental hardening effect which appears when the shear strain of the HDRB exceeds the limit value around 100% usually adopted in design practice. The incremental hardening effect is sometimes neglected in the design but it is an important aspect because it ensures a seismic protection of the base isolated structure also in presence of exceptional seismic events for intensity or frequency content. In this paper we have highlighted the significant influence of this phenomenon in the seismic response of the isolated structure by reporting the cyclic behavior of a HDRB respectively neglecting and considering this aspect.

Abstract: This paper investigates the coupled effect of the supporting soil flexibility and pounding between neighbouring, insufficiently separated buildings under earthquake excitation. Two adjacent three-storey structures, modelled as inelastic lumped mass systems with different structural characteristics, have been considered in the study. The models have been excited using the time history of the Kobe earthquake of 1995. A nonlinear viscoelastic pounding force model has been employed in order to effectively capture the impact forces during collisions. A discrete element model has been incorporated to simulate the horizontal and rotational movements of the supporting soil. Numerical simulations have been performed using developed software based on the Matlab code. The variation in storeys peak displacements, peak accelerations and peak impact forces for various gap sizes is presented in the paper and comparisons are made with the results obtained for colliding buildings with fixed-base supports. The results of the study indicate that the incorporation of the soil-structure interaction decreases both storey peak displacements and peak impact forces during collisions, whereas increase the peak accelerations at each floor level.

Abstract: The aim was to develop a nonlinearity model of quantitative analysis of vitamin C content by infrared spectroscopy and provide theoretical basis for nondestructive testing of vitamin C content in drugs. The infrared spectrum of vitamin C was smoothed and noise-eliminated by wavelet transform, which may obtained smooth curve and enhanced spectrum resolution. Absorption peak about 3000cm^-1 was selected as characteristic absorption peak to set up linearity or nonlinearity model. The result showed that nonlinearity model, correlation coefficient 0.9967, had more nicety than that of linearity model. The result of traditional titration of iodine as reference system, relative deviation of nonlinearity model was 2.75%, but that of linearity model was 28.03%. It was feasible and effective to determine vitamin C content by infrared spectroscopy based on nonlinear modeling.