Neophytos Chiras’s research while affiliated with University of Wales and other places

This paper examines the output properties of static power-series nonlinearities driven by periodic multiharmonic signals with emphasis given to their effect on linear frequency response function (FRF) measurements. The analysis is based on the classification of nonlinear distortions into harmonic and interharmonic contributions. The properties of harmonic contributions are examined in detail and explicit formulae are derived, by which the number of harmonic contributions generated at the test frequencies can be calculated for odd-order nonlinearities up to, and including, the ninth order. Although an analytic solution for any odd-order nonlinearity is still under investigation, a heuristic methodology is developed that solves this problem. It is shown that the derived formulae provide a useful tool in the examination of the behavior of FRF measurements in the presence of nonlinear distortions. Based on these formulae, different approaches in classifying nonlinear distortions are then compared with respect to their suitability in assessing the influence of system nonlinearities on linear FRF measurements.

The aim of the study presented in this paper is to provide some initial results over the potential use of best linear approximation models in the design of model-based linear controllers for systems suffering nonlinear distortions. This is illustrated on a nonlinear mechanical resonating system. Parametric best linear approximation models are estimated from frequency response function measurements using random phase multisines. A full nonlinear model for the system is also estimated. This provides the basis for the design of simple optimal linear controllers, and the performance obtained from the controllers based on the nonlinear model is set as the benchmark. Parametric linear models are also estimated from measurements taken using Schroeder phase multisine signals and controllers based on these models are designed. It is shown that in the presence of nonlinear distortions the choice of input excitation in estimating models that describe the best linear approximation to the nonlinear system is crucial. It is also shown that this choice will have an effect on the performance of the model-based controllers.

This paper presents PID controller designs based on NARMAX and feedforward neural network models of a Spey gas turbine engine. Both models represent the dynamic relationship between the fuel flow and shaft speed. Due to the engine non-linearity, a single set of PID controller parameters is not sufficient to control the gas turbine throughout the operating range. Gain-scheduling PID controllers are therefore used in order to obtain optimum control. A comparison between the controller designs based on the two model representations is also made.

This paper examines the estimation of a global nonlinear gas turbine model using NARMAX techniques. Linear models estimated on small-signal data are first examined and the need for a global nonlinear model is established. A nonparametric analysis of the engine nonlinearity is then performed in the time and frequency domains. The information obtained from the linear modelling and nonlinear analysis is used to restrict the search space for nonlinear modelling. The nonlinear model is then validated using large-signal data and its superior performance illustrated by comparison with a linear model. This paper illustrates how periodic test signals, frequency domain analysis and identification techniques, and time-domain NARMAX modelling can be effectively combined to enhance the modelling of an aircraft gas turbine.

A new method for estimating the highest significant order of nonlinearity of Volterra type systems is presented. The method is based on the use of multisine signals and the possibility of testing the system at three different amplitudes. The performance of the proposed method is demonstrated in simulation and it is shown that it is possible to estimate the highest order of nonlinearity of Volterra type systems very accurately. The method can be used to provide essential prior knowledge about the nonlinearity and thus aid the accurate representation of the system under test.

New insights on the frequency-domain output properties of nonlinear systems driven by multiharmonic signals will be discussed in this paper. These are based on a detailed analysis of the mechanisms at work on the generation of a class of nonlinear distortions, termed harmonic components. This leads to the derivation of formulae to calculate the number of harmonic components generated by any order of nonlinearity at the excitation frequencies during frequency response function measurements. Based on these formulae the classification of nonlinear distortions in harmonic and interharmonic components is compared against other methodologies on the classification of nonlinear distortions.

In this paper a feedforward neural network is used to model the fuel flow to shaft speed relationship of a Spey gas turbine engine. The performance of the estimated model is validated against a range of small and large signal engine tests. It is shown that the performance of the estimated models is superior to that of the estimated linear models.

This paper presents controllers based on linear and nonlinear models of an aircraft gas turbine engine. These models along with the performance indexes, ITAE, IAE and IES are used to estimate the parameters of a PID controller. The parameters and hence the quality of the controllers are very much dependent on the accuracy of the models. It is shown that the Nonlinear AutoRegressive Moving Average with eXogenous (NARMAX) representation provides a comprehensive benchmark for controller design, since it models the global dynamics of the engine. However, the linear models provide an accurate representation for small signal inputs and give a reliable method for estimating the controller parameters for a limited operating range. A comparison between controllers using linear and NARMAX models is made.

In this paper the orthogonal estimation algorithm is used to estimate a NARMAX model for an aircraft gas turbine. The performance of the model is validated using a range of small and large signal tests and by examining the model's static and dynamic characteristics. The NARMAX representation allows the direct mapping to the frequency domain by computing the Higher Order Frequency Response Functions (HOFRF) thus allowing physical interpretation to be made.

The estimation of a nonlinear autoregressive moving average with exogenous inputs (NARMAX) model of an aircraft gas turbine is presented. A method is proposed whereby periodic signals with certain harmonic content are used to qualify the nature of the nonlinearity of the engine in the frequency domain. The static behavior of the engine is investigated in the time domain to approximate the order of nonlinearity and this information is used a priori to restrict the search space of the potential NARMAX models. A forward-regression orthogonal estimation algorithm is then employed to select the model terms using the error reduction ratio. The performance of the estimated NARMAX model is illustrated against a range of small- and large-signal engine tests