System oriented neural networks - Problem formulation, methodology and application

A novel methodology is proposed for the development of neural network models for complex engineering systems exhibiting nonlinearity. This method performs neural network modeling by first establishing some fundamental nonlinear functions from a priori engineering knowledge, which are then constructed and coded into appropriate chromosome representations. Given a suitable fitness function, using evolutionary approaches such as genetic algorithms, a population of chromosomes evolves for a certain number of generations to finally produce a neural network model best fitting the system data. The objective is to improve the transparency of the neural networks, i.e. to produce physically meaningful

Comparative analysis of the DC performance of DG MOSFETs on highly-doped and near-intrinsic silicon layers

A comparison of dc characteristics of fully depleted double-gate (DG) MOSFETs with respect to low-power circuit applications and device scaling has been performed by two-dimensional device simulation. Three different DG MOSFET structures including a conventional N+ polysilicon gate device with highly doped Si layer, an asymmetrical P+/N+ polysilicon gate device with low doped Si layer and a midgap metal gate device with low doped Si layer have been analysed. It was found that DG MOSFET with mid-gap metal, gates yields the best dc parameters for given off-state drain leakage current and highest immunity to the variation of technology parameters (gate length, gate oxide thickness and Si layer thickness). It is also found that an asymmetrical P+/N+ polysilicon gate DG MOSFET design offers comparable dc characteristics, but better parameter immunity to technology tolerances than a conventional DG MOSFET. (C) 2004 Elsevier Ltd. All rights reserved.

Improved fault diagnosis in multivariate systems using regression-based reconstruction

Subspace monitoring has recently been proposed as a condition monitoring tool that requires considerably fewer variables to be analysed compared to dynamic principal component analysis (PCA). This paper analyses subspace monitoring in identifying and isolating fault conditions, which reveals that the existing work suffers from inherent limitations if complex fault senarios arise. Based on the assumption that the fault signature is deterministic while the monitored variables are stochastic, the paper introduces a regression-based reconstruction technique to overcome these limitations. The utility of the proposed fault identification and isolation method is shown using a simulation example and the analysis of experimental data from an industrial reactive distillation unit.

PLS modelling and fault detection on the Tennessee Eastman benchmark

This paper describes the application of multivariate regression techniques to the Tennessee Eastman benchmark process for modelling and fault detection. Two methods are applied : linear partial least squares, and a nonlinear variant of this procedure using a radial basis function inner relation. The performance of the RBF networks is enhanced through the use of a recently developed training algorithm which uses quasi-Newton optimization to ensure an efficient and parsimonious network; details of this algorithm can be found in this paper. The PLS and PLS/RBF methods are then used to create on-line inferential models of delayed process measurements. As these measurements relate to the final product composition, these models suggest that on-line statistical quality control analysis should be possible for this plant. The generation of `soft sensors' for these measurements has the further effect of introducing a redundant element into the system, redundancy which can then be used to generate a fault detection and isolation scheme for these sensors. This is achieved by arranging the sensors and models in a manner comparable to the dedicated estimator scheme of Clarke et al. 1975, IEEE Trans. Pero. Elect. Sys., AES-14R, 465-473. The effectiveness of this scheme is demonstrated on a series of simulated sensor and process faults, with full detection and isolation shown to be possible for sensor malfunctions, and detection feasible in the case of process faults. Suggestions for enhancing the diagnostic capacity in the latter case are covered towards the end of the paper.

Plant-wide predictive control for a thermal power plant based on a physical plant model

A constrained non-linear, physical model-based, predictive control (NPMPC) strategy is developed for improved plant-wide control of a thermal power plant. The strategy makes use of successive linearisation and recursive state estimation using extended Kalman filtering to obtain a linear state-space model. The linear model and a quadratic programming routine are used to design a constrained long-range predictive controller One special feature is the careful selection of a specific set of plant model parameters for online estimation, to account for time-varying system characteristics resulting from major system disturbances and ageing. These parameters act as nonstationary stochastic states and help to provide sufficient degrees-of-freedom to obtain unbiased estimates of controlled outputs. A 14th order non-linear plant model, simulating the dominant characteristics of a 200 MW oil-fired pou er plant has been used to test the NPMPC algorithm. The control strategy gives impressive simulation results, during large system disturbances and extremely high rate of load changes, right across the operating range. These results compare favourably to those obtained with the state-space GPC method designed under similar conditions.

Disturbance attenuation in linear systems via dynamical compensators: A parametric eigenstructure assignment approach

A simple approach is proposed for disturbance attenuation in multivariable linear systems via dynamical output compensators based on complete parametric eigenstructure assignment. The basic idea is to minimise the H-2 norm of the disturbance-output transfer function using the design freedom provided by eigenstructure assignment. For robustness, the closed-loop system is restricted to be nondefective. Besides the design parameters, the closed-loop eigenvalues are also optimised within desired regions on the left-half complex plane to ensure both closed-loop stability and dynamical performance. With the proposed approach, additional closed-loop specifications can be easily achieved. As a demonstration, robust pole assignment, in the sense that the closed-loop eigenvalues are as insensitive as possible to open-loop system parameter perturbations, is treated. Application of the proposed approach to robust control of a magnetic bearing with a pair of opposing electromagnets and a rigid rotor is discussed.

Simulation of coupled heat and moisture transfer in air-conditioned buildings

The simultaneous heat and moisture transfer in the building envelope has an important influence on the indoor environment and the overall performance of buildings. In this paper, a model for predicting whole building heat and moisture transfer was presented. Both heat and moisture transfer in the building envelope and indoor air were simultaneously considered; their interactions were modeled. The coupled model takes into account most of the main hygrothermal effects in buildings. The coupled system model was implemented in MATLAB-Simulink, and validated by using a series of published testing tools. The new program was applied to investigate the moisture transfer effect on indoor air humidity and building energy consumption under different climates. The results show that the use of more detailed simulation routines can result in improvements to the building's design for energy optimisation through the choice of proper hygroscopic materials, which would not be indicated by simpler calculation techniques.

Convergence and tracking analysis of a variable normalised LMF (XE-NLMF) algorithm

The least-mean-fourth (LMF) algorithm is known for its fast convergence and lower steady state error, especially in sub-Gaussian noise environments. Recent work on normalised versions of the LMF algorithm has further enhanced its stability and performance in both Gaussian and sub-Gaussian noise environments. For example, the recently developed normalised LMF (XE-NLMF) algorithm is normalised by the mixed signal and error powers, and weighted by a fixed mixed-power parameter. Unfortunately, this algorithm depends on the selection of this mixing parameter. In this work, a time-varying mixed-power parameter technique is introduced to overcome this dependency. A convergence analysis, transient analysis, and steady-state behaviour of the proposed algorithm are derived and verified through simulations. An enhancement in performance is obtained through the use of this technique in two different scenarios. Moreover, the tracking analysis of the proposed algorithm is carried out in the presence of two sources of nonstationarities: (1) carrier frequency offset between transmitter and receiver and (2) random variations in the environment. Close agreement between analysis and simulation results is obtained. The results show that, unlike in the stationary case, the steady-state excess mean-square error is not a monotonically increasing function of the step size. (c) 2007 Elsevier B.V. All rights reserved.

Kinematic Design of a 6-DOF Parallel Manipulator with Decoupled Translation and Rotation

A new three-limb, six-degree-of-freedom (DOF) parallel manipulator (PM), termed a selectively actuated PM (SA-PM), is proposed. The end-effector of the manipulator can produce 3-DOF spherical motion, 3-DOF translation, 3-DOF hybrid motion, or complete 6-DOF spatial motion, depending on the types of the actuation (rotary or linear) chosen for the actuators. The manipulator architecture completely decouples translation and rotation of the end-effector for individual control. The structure synthesis of SA-PM is achieved using the line geometry. Singularity analysis shows that the SA-PM is an isotropic translation PM when all the actuators are in linear mode. Because of the decoupled motion structure, a decomposition method is applied for both the displacement analysis and dimension optimization. With the index of maximal workspace satisfying given global conditioning requirements, the geometrical parameters are optimized. As a result, the translational workspace is a cube, and the orientation workspace is nearly unlimited.

Predictive Current Control of Doubly Fed Induction Generators

This paper presents a predictive current control strategy for doubly-fed induction generators (DFIG). The method predicts the DFIG’s rotor current variations in the synchronous reference frame fixed to the stator flux within a fixed sampling period. This is then used to directly calculate the required rotor voltage to eliminate the current errors at the end of the following sampling period. Space vector modulation is used to generate the required switching pulses within the fixed sampling period. The impact of sampling delay on the accuracy of the sampled rotor current is analyzed and detailed compensation methods are proposed to improve the current control accuracy and system stability. Experimental results for a 1.5 kW DFIG system illustrate the effectiveness and robustness of the proposed control strategy during rotor current steps and rotating speed variation. Tests during negative sequence current injection further demonstrate the excellent dynamic performance of the proposed PCC method.

The Influence of Friction Stir Welding Process Idealization on Residual Stress and Distortion Predictions for Future Airframe Assembly Simulations

The ability to accurately predict residual stresses and resultant distortions is a key product from process assembly simulations. Assembly processes necessarily consider large structural components potentially making simulations computationally expensive. The objective herein is to develop greater understanding of the influence of friction stir welding process idealization on the prediction of residual stress and distortion and thus determine the minimum required modeling fidelity for future airframe assembly simulations. The combined computational and experimental results highlight the importance of accurately representing the welding forging force and process speed. In addition, the results emphasize that increased CPU simulation times are associated with representing the tool torque, while there is potentially only local increase in prediction fidelity.