Transactions of the Institute of Measurement and Control recent issues

PI control of first-order lag plus time-delay plants: root locus design for optimal stability

Cogan, B., de Paor, A. M., Quinn, A. — Thu, 24 Sep 2009 06:20:38 PDT

We present a procedure for the design of a PI controller for a general first-order lag plus time-delay plant. We derive two equations that allow a designer to calculate the PI controller parameters using only the plant parameters. The central idea of this design procedure is to use the root locus method to select controller parameters that put the system’s rightmost eigenvalue as far to the left as possible in the (, ) plane. When the system is operating at this point, it is said to be operating at a point of optimum stability in the root locus sense.

Robust performance design of single-input fuzzy control system for plant with time delay

Yordanova, S. — Thu, 24 Sep 2009 06:20:38 PDT

Fuzzy logic has recently gained popularity in the development of robust non-linear control convenient for real-time applications, satisfying the high-performance requirements of the control systems by dealing with data imprecision and noise. There are, however, problems in designing a unique controller that will provide system stability and desirable performance in the presence of plant model uncertainties and time delays. The aim of this investigation is to develop a simple approach for the design of an incremental PI-like single-input fuzzy controller (SI FC) that ensures a robust performance in the closed-loop system for the control of industrial plants with time delays and model uncertainties. The main results are: 1) derivation of a robust performance criterion of the SI FC system in the frequency domain, employing the stabilization of the system linear part, the linearization of the fuzzy unit and the Popov stability approach; 2) development of an SI FC tuning algorithm; 3) application of the design approach for the control of air temperature in a furnace using MATLAB^TM, and assessment of the advantages over an ordinary PI controller. The designed SI FC preserves system stability and performance for a high range of plant model uncertainties.

Composite disturbance-observer-based control and variable structure control for non-linear systems with disturbances

Wei, X., Zhang, H., Guo, L. — Thu, 24 Sep 2009 06:20:38 PDT

A novel type of control scheme combining the disturbance-observer-based control (DOBC) with variable structure control (VSC) is proposed for a class of continuous non-linear systems subject to disturbances. The disturbances are supposed to include two parts. One in the input channel is generated by an exogenous system with uncertainty, which can represent the harmonic signals with modelling perturbations. The other is supposed to have the bounded H₂ norm. The disturbance observers based on regional pole placement and D-stability theory are presented, which can be designed separately from the controller design. By integrating DOBC with VSC laws, the disturbances can be rejected and the desired dynamic performances can be guaranteed for continuous systems in finite time with known and unknown non-linear dynamics, respectively. Simulations for a flight control system are given to demonstrate the effectiveness of the results and compare the proposed results with the previous schemes.

Time scaling in PID controller tuning

Atherton, D. P., Boz, A. F. — Thu, 24 Sep 2009 06:20:38 PDT

The paper examines the problem of tuning the parameters of a PID controller for the frequently assumed first-order plus dead time plant. The original contribution is to look at the problem from a time-scaling viewpoint so that the controller parameters can be given as a function of the time delay to time constant ratio. This enables the controller parameter results of different rules to be easily compared and reveals those that are not consistent under time scaling.

Sliding mode controlled bioreactor using a time-varying sliding surface

Tokat, S. — Thu, 24 Sep 2009 06:20:39 PDT

Bioreactor processes based on substrate concentration have both highly non-linear behaviour and parameter uncertainties mostly bounded with physical and biological constraints, which make them a subject area of sliding mode control. This work deals with sliding mode control of a fermentation process in a continuous stirred bioreactor. A new sliding surface is defined using angular information as a function of time that provides adjustable continuous movement of the sliding surface. Computer simulations are presented to show the effectiveness and efficiency of the proposed method and to make a quantitative comparison with the conventional sliding mode controller having a constant sliding surface. The excess amount of substrate concentration in the mixture causes undesired by-product information. Therefore, the substrate concentration should not stay above its desired value for a long time. It is shown in the simulations that the performance of the proposed method, especially in regard to the settling time of the substrate concentration and reaching time, is improved. It is also shown that different system state trajectories could be obtained with the proposed method by adjusting the defined parameters of the angular information.

A particle-filtering approach for on-line fault diagnosis and failure prognosis

Orchard, M. E., Vachtsevanos, G. J. — Tue, 30 Jun 2009 02:24:18 PDT

This paper introduces an on-line particle-filtering (PF)-based framework for fault diagnosis and failure prognosis in non-linear, non-Gaussian systems. This framework considers the implementation of two autonomous modules. A fault detection and identification (FDI) module uses a hybrid state-space model of the plant and a PF algorithm to estimate the state probability density function (pdf) of the system and calculates the probability of a fault condition in real-time. Once the anomalous condition is detected, the available state pdf estimates are used as initial conditions in prognostic routines. The failure prognostic module, on the other hand, predicts the evolution in time of the fault indicator and computes the pdf of the remaining useful life (RUL) of the faulty subsystem, using a non-linear state-space model (with unknown time-varying parameters) and a PF algorithm that updates the current state estimate. The outcome of the prognosis module provides information about the precision and accuracy of long-term predictions, RUL expectations and 95% confidence intervals for the condition under study. Data from a seeded fault test for a UH-60 planetary gear plate are used to validate the proposed approach.

Fabrication and evaluation of hybrid silica/polymer optical fibre sensors for large strain measurement

Haiying Huang, , Majumdar, A., Cho, J.-S. — Tue, 30 Jun 2009 02:24:18 PDT

Silica-based optical fibre sensors are widely used in structural health monitoring systems for strain and deflection measurement. One drawback of silica-based optical fibre sensors is their low strain toughness. In general, silica-based optical fibre sensors can only reliably measure strains up to 2%. Recently, polymer optical fibre sensors have been investigated for large strain measurement. Because of their high optical losses, the length of the polymer optical fibres is limited to 100 m. In this paper, we present a novel economical technique to fabricate hybrid silica/polymer optical fibre sensors that are capable of measuring large strain. First, stress analysis of a surface-mounted optical fibre sensor is performed to understand the effect of mechanical properties on the load distribution between the optical fibre and the host structure. Next, the procedure for fabricating a tapered polymer sensing element between two silica optical fibres is explained. The experimental set-up and the components used in the fabrication process are described in detail. Test results of the fabricated silica/polymer optical fibre strain sensor are presented.

Extrinsic curvature-based fault isolation for multi-input--multi-output systems with non-linear fault models

Subbarao, K., Vemuri, A. T. — Tue, 30 Jun 2009 02:24:18 PDT

This paper presents an online fault isolation methodology for identifying faulty signals in a multiinput—multi-output dynamical system. It is hypothesized that faults in a dynamical system can be suitably represented via non-linear functions. The isolation scheme, which is implemented online, relies on adaptive non-linear estimates of these non-linear fault functions based on the system input—output data. The non-linear fault estimation is achieved using radial basis function neural network (RBFNN) architecture while the fault isolation is accomplished using extrinsic curvature of the learned RBFNN model. The proposed approach is implemented on an F-5A Freedom Fighter aircraft's lateral-directional model and the results are presented to illustrate the concept.

On improving performance of aircraft engine gas path fault diagnosis

Yan, W.Z., Li, J.C., Goebel, K.F. — Tue, 30 Jun 2009 02:24:18 PDT

Aircraft engine fault diagnosis plays a crucial role in cost-effective operations of aircraft engines. However, designing an engine fault diagnostic system with the desired performance is a challenging task, because of several characteristics associated with aircraft engines. Geared towards achieving the highest possible performance of fault diagnosis, this paper explores strategies on improving diagnosis performance. Specifically, we introduce flight regime mapping and a two-level multiple classifier system as means to improve classification performance. By designing a real-world aircraft fault diagnostic system, we demonstrate that the strategies adopted in this study are effective in improving the performance of aircraft engine fault diagnostic systems.

Comparison of prognostic algorithms for estimating remaining useful life of batteries

Saha, B., Goebel, K., Christophersen, J. — Tue, 30 Jun 2009 02:24:18 PDT

The estimation of remaining useful life (RUL) of a faulty component is at the centre of system prognostics and health management. It gives operators a potent tool in decision making by quantifying how much time is left until functionality is lost. RUL prediction needs to contend with multiple sources of errors, like modelling inconsistencies, system noise and degraded sensor fidelity, which leads to unsatisfactory performance from classical techniques like autoregressive integrated moving average (ARIMA) and extended Kalman filtering (EKF). The Bayesian theory of uncertainty management provides a way to contain these problems. The relevance vector machine (RVM), the Bayesian treatment of the well known support vector machine (SVM), a kernel-based regression/classification technique, is used for model development. This model is incorporated into a particle filter (PF) framework, where statistical estimates of noise and anticipated operational conditions are used to provide estimates of RUL in the form of a probability density function (pdf). We present here a comparative study of the above-mentioned approaches on experimental data collected from Li-ion batteries. Batteries were chosen as an example of a complex system whose internal state variables are either inaccessible to sensors or hard to measure under operational conditions. In addition, battery performance is strongly influenced by ambient environmental and load conditions.

Physics-of-failure-based prognostics for electronic products

Pecht, M., Jie Gu, — Tue, 30 Jun 2009 02:24:18 PDT

This paper presents a physics-of-failure (PoF)-based prognostics and health management approach for effective reliability prediction. PoF is an approach that utilizes knowledge of a product's life cycle loading and failure mechanisms to perform reliability design and assessment. PoF-based prognostics permit the assessment of product reliability under its actual application conditions. It integrates sensor data with models that enable in situ assessment of the deviation or degradation of a product from an expected normal operating condition (ie, the product's `health') and the prediction of the future state of reliability. A formal implementation procedure, which includes failure modes, mechanisms, and effects analysis, data reduction and feature extraction from the life cycle loads, damage accumulation, and assessment of uncertainty, is presented. Then, applications of PoF-based prognostics are discussed.

Condition-based maintenance using dynamic decisions by Petri nets and Dempster--Shafer theory: a matrix-based approach

Ballal, P., Lewis, F. — Tue, 30 Jun 2009 02:24:18 PDT

This paper provides a novel method for job and resource dispatching for condition-based maintenance using free-choice Petri nets (PNs) and Dempster—Shafer (DS) evidence theory. In free-choice PNs (also called free-choice multi re-entrant flow-lines — FMRF), a choice exists for the tasks to be performed. Hence decisions have to be made in real time. In case of unreliable sensors in the system, there is uncertainty in decision making. This uncertainty is modelled in this paper using the DS theory of combination, which is used to decide which path to take in the event of choice decisions in the Petri net. The computations required in DS theory are difficult to perform because of the many required summations over set inclusions and intersections. Therefore, a new matrix formulation is given herein for efficiently combining evidence and finding belief, plausibility and other quantities in DS theory. This, coupled with a previously developed matrix formulation for PNs, makes it straightforward to implement the decision framework as a discrete event controller using computer software. Examples are given showing how to apply the matrix formulation for decision making in intelligent diagnostics.

Engine performance trending: practical insights

Mylaraswamy, D., Olson, L., Nwadiogbu, E. — Tue, 30 Jun 2009 02:24:18 PDT

A vehicle health usage monitoring system (HUMS) makes it possible to create an accurate picture of the current and expected condition of a propulsion engine. Gas turbine engines, the focus of this paper, convert fuel energy into useful mechanical work by following a well-defined thermodynamic cycle. In this paper, the authors present the principles of engine performance trending, describing them with examples. In engine performance trending, the underlying thermodynamic cycle is analysed using three key steps: cycle characterization using noisy sensor data, interpreting deviations with respect to engine faults and trending the deviations to assess future behaviour. In practical terms, engine performance trending exposes the correct analytic choice based on data and customer requirements. The first half of this paper describes a hierarchy of techniques for characterizing the underlying thermodynamic cycle, from simple energy balance to more complex aerodynamic analysis. Here, we discuss pattern recognition algorithms (eg, fuzzy logic and principal component analysis) for interpreting deviations and the principles of predictive trending. In the second half of the paper, we give examples to describe the practical customization needed to meet the requirements for HUMS. In our final remarks, we discuss open research issues, primarily in the area of constrained parameter estimation, as well as practical validation of engine performance trending. This paper is intended for HUMS practitioners in academia and industry.

Towards the re-verification of process tank calibrations

Howell, J. — Wed, 18 Mar 2009 04:22:11 PDT

Re-verification is needed to ensure that the calibration (the relationship between measured level and measured volume) that is obtained during commissioning has not changed over time. This can be achieved, in part, by metering in solution and correlating with marks identified a priori. Mark identification and correlation are discussed and possible error sources are outlined.

A novel adaptive control algorithm based on non-linear Laguerre--Volterra observer

Zhang, H.-T., Tischenko, L., Yu, P.-Z. — Wed, 18 Mar 2009 04:22:11 PDT

By expanding each kernel using an orthonormal Laguerre series, a Volterra functional series is used to represent the input—output relation of a non-linear dynamic system. When Volterra series and Laguerre series truncations are allowed, an appropriate choice of the Laguerre filter pole permits a description of the process dynamics with a small number of parameters. Feeding back the error of the outputs of the plant and the model, we design a novel non-linear state observer, based on which a stable output feedback control law is derived for both regulator and tracking problems. To support this algorithm, we present the theoretical analyses of its nominal stability, which allows us to obtain the state feedback gain and the observer gain solely by solving two linear matrix inequalities (LMIs). In addition, another theorem is also given to show its capability of minimizing the steady-state tracking errors. To handle more complex dynamics, we improve the standard recursive least square estimation (RLSE) identification method to a normalized one with guaranteed convergence. Finally, control simulations on a benchmark problem — a continuous stirring tank reactor (CSTR) process — and experiments on a chemical reactor temperature control system are performed. This method, especially its essential idea of a Volterra non-linear observer, has shown great potential for the control of a large class of non-linear dynamic systems.

Hot wire probe calibration using artificial neural network

Erdil, A., Arcaklioglu, E. — Wed, 18 Mar 2009 04:22:11 PDT

Hot wire anemometers (HWAs) are used as research tools in fluid mechanics. In HWA measurements, a hot wire probe calibration stage must be carried out before the experiments, in order to determine the relation between the probe voltage and fluid velocity at the nozzle exit, because of variation in the ambient conditions. In this study, a hot wire probe calibration system is used to calibrate a one-dimensional hot wire probe by means of an AN-1005 HWA, which operates as a constant temperature anemometer (CTA). Experimental results have been used to train an artificial neural network (ANN) in order to produce a new calibration curve for new conditions, because calibration is time consuming and difficult. The network has yielded an R² value of 0.999, and very small root-mean-squared values, which indicate that predicted values are close to the experimental ones, and the ANN model gives a good approximation for the calibration curve under different conditions. In addition, formulations have been prepared according to a hidden number of neurons studied. Since the necessary formulae have been given, anyone could use these variables to obtain predictions from the ANN as if the hot wire probe has been operated.

Motion conditions justification of an omni-directional mobile platform by electro-mechanical transduction matrix

Kuo, Y.L., Lin, J.Y., Lee, Y.N. — Wed, 18 Mar 2009 04:22:11 PDT

The paper presents a new approach to electro-mechanical transduction applied to motion conditions justification of an omni-directional mobile platform without implementing conventional mechanical sensors. Since the experimental platform is driven by four DC motors, the electro-mechanical transduction of a DC motor is identified by theoretical derivation. Also, the motion modelling of the mobile platform is presented. Based on the evaluation of mechanical impedance, an experimental example demonstrates that the platform has the function of road surface sensing, which can be used as a base for controlling the motion of the platform. The verification of the rotating speed of a DC motor is also provided, and the results show that the proposed approach can accurately respond to the motion of the platform.

Optimal LuGre friction model identification based on genetic algorithm and sliding mode control of a piezoelectric-actuating table

Huang, S.-J., Chiu, C.-M. — Wed, 18 Mar 2009 04:22:11 PDT

A piezoelectric friction actuating mechanism is employed to construct a long travelling range sub-micro X—Y positioning table. The piezoelectric material is used to generate high-frequency oscillation for actuating a fingertip, which in is contact with a slide to induce back-and-forth motion. The LuGre friction model is chosen to simulate the friction dynamics of this positioning mechanism. The genetic algorithm (GA) is employed to search for the optimal friction model parameters. However, this piezoelectric actuating system has an obvious non-linear and time-varying dead-zone offset control voltage related to the static friction and preload. The GA-estimated LuGre dynamic model is still not accurate enough for model-based precision control design. Hence, sliding mode control (SMC) with robust behaviour is employed to design a non-linear controller for this piezoelectric friction actuating mechanism. The Laypunov-like design strategy is adopted to satisfy the system stability criterion. Tracking control of different trajectories is planned to investigate the motion control performances and the steady-state errors of the SMC non-linear controller based on the GA-estimated model. The dynamic experimental results of the proposed non-linear controllers are also compared with that of a model-based PID controller.

Exponential stability of stochastic switched systems

Filipovic, V. — Wed, 18 Mar 2009 04:22:11 PDT

This paper proposes a method for exponential m-stability analysis of stochastic switched systems. The models, in a finite set of models, are non-linear stochastic models. It is assumed that there is no jump in the state at switching instants and there is no Zeno behaviour, ie, there is finite number of switches on every bounded interval. The stochastic hybrid systems have wide applications: transmission control protocol flows with congestion avoidance and slow-start modes; estimation in distributed networked systems; air traffic control; process control and communication networks for control systems. For analysis of stochastic switched system the multiple Lyapunov functions are used and exponential m-stability is proved. From the main result of the paper: 1) the exponential m₁ -stability of stochastic switched systems whereby m₁ (0, m); 2) the stability in probability. The exponentially stable equilibrium of system is relevant for practice because such systems are robust to perturbations.

Editorial: Recent Advances in Control Theory and Applications in Turkey

Efe, M. O. — Wed, 14 Jan 2009 08:23:31 PST

Review of fuzzy system models with an emphasis on fuzzy functions

Turksen, I. B. — Wed, 14 Jan 2009 08:23:31 PST

Fuzzy system modelling (FSM) is one of the most prominent tools that can be used to identify the behaviour of highly non-linear systems with uncertainty. In the past, FSM techniques utilized Type 1 fuzzy sets in order to capture the uncertainty in the system. However, since Type 1 fuzzy sets express the belongingness of a crisp value x^' of an input variable x in a fuzzy set A by a crisp membership value µ_A(x^'), they cannot fully capture the uncertainties associated with higher-order imprecisions in identifying membership functions. In the future, we are likely to observe higher types of fuzzy sets, such as Type 2 fuzzy sets. The use of Type 2 fuzzy sets and linguistic logical connectives has drawn a considerable amount of attention in the realm of FSM in the last two decades. In this paper, we first review Type 1 fuzzy system models known as Zadeh, Takagi— Sugeno and Turksen models; then we review potentially future realizations of Type 2 fuzzy systems again under the headings of Zadeh, Takagi—Sugeno and Turksen fuzzy system models, in contrast to Type 1 fuzzy system models. Zadeh's and Takagi—Sugeno's models are essentially fuzzy rule base (FRB) models, whereas Turksen's models are essentially fuzzy function (FF) models. Type 2 fuzzy system models have a higher predictive power. One of the essential problems of Type 2 fuzzy system models is computational complexity. In data-driven FSM methods discussed here, a fuzzy C-means (FCM) clustering algorithm is used in order to identify the system structure, ie, either the number of fuzzy rules or alternately the number of FFs.

An adaptive grey PID-type fuzzy controller design for a non-linear liquid level system

Kayacan, E., Kaynak, O. — Wed, 14 Jan 2009 08:23:31 PST

Product-sum-type fuzzy controllers are known to have similar characteristics to PD-type controllers. In the case of type-0 control systems, PID-type fuzzy controllers have been proposed in the literature in order to eliminate the steady-state error. However, these control methods, essentially based on conventional PID theory, have no predictive capabilities. The concept of grey system theory, which has a certain prediction capability, offers an alternative approach for various kinds of conventional control methods, such as PID control and fuzzy control. This paper proposes a grey prediction-based fuzzy PID controller that can overcome the stated shortcomings. In order to obtain a better controller performance, another fuzzy controller is designed to change the step size of the grey predictor. A non-linear liquid level system is taken as a test bed. The grey model developed is examined under several different conditions and it is shown that the proposed grey fuzzy PID controller can predict the future output value of the system. It is clear that the proposed adaptive PID-type fuzzy controller is effective in controlling such a non-linear system accurately by changing the step size adaptively for real-time working.

Linear time-varying sliding surface design based on co-ordinate transformation for high-order systems

Tokat, S., Eksin, I., Guzelkaya, M. — Wed, 14 Jan 2009 08:23:31 PST

In this study, a sliding mode controller with a linear time-varying sliding surface is proposed for high-order systems by generalizing the co-ordinate transformed sliding surface design algorithm devised by the authors for second-order systems. The sliding mode control law is formulated for the sliding surface that has been defined by using a time-varying function. The equivalent control term of the proposed controller is expressed as a sum of the equivalent control term of the conventional sliding mode controller and an additive signal, which is a linear function of system tracking error vector and a time-dependent monotonous function. Simulations are performed on a third-order non-linear model with external disturbances and parameter variations. The performance of the sliding mode controller with the proposed design methodology is compared both with a conventional sliding mode controller and with another sliding mode controller that also uses an additive term in the control law to minimize the reaching time. The simulation results have shown that the proposed method has improved the robustness and the transient response with respect to related sliding mode controllers.

LPV gain-scheduling controller design for a non-linear quarter-vehicle active suspension system

Onat, C., Kucukdemiral, I.B., Sivrioglu, S., Yuksek, I., Cansever, G. — Wed, 14 Jan 2009 08:23:31 PST

There always exists a conflict between ride comfort and suspension deflection performances during the vibration control of suspension systems. Active suspension control systems, which are designed by linear methods, can only serve as a trade-off between these conflicting performance criteria. Both performance objectives can only be accomplished at the same time by using a nonlinear controller. This paper addresses the non-linear induced L₂ control of an active suspension system, which contains non-linear spring and damper elements. The design method is based on the linear parameter varying (LPV) model of the system. The proposed method utilizes the bilinear damping characteristic, stiffening spring characteristic when the suspension deflection approaches the structural limits, mass variations and parameter-dependent weighting filters. Simulation studies both in time and frequency domain demonstrate that the active suspension system controlled by the proposed method always guarantees an agreement between acceleration (comfort) and suspension deflection magnitudes together with a high ride performance.

Seven tuning schemes for an ADALINE model to predict floor pressures in a subsonic cavity flow

Efe, M. O., Debiasi, M., Peng Yan, , Ozbay, H., Samimy, M. — Wed, 14 Jan 2009 08:23:31 PST

This paper presents a simple yet effective one-step-ahead predictor based on an adaptive linear element (ADALINE). Several tuning schemes are studied to see whether the obtained model is consistent. The process under investigation is a subsonic cavity flow system. The experimental data obtained from the system is post-processed to obtain a useful predictor. The contribution of the paper is to demonstrate that despite the spectral richness of the observed data, a simple model with various tuning schemes can help to a satisfactory extent. Seven algorithms are studied, including the least mean squares (LMS), recursive least squares (RLS), modified Kaczmarz's algorithm (MK), stochastic approximation algorithm (SA), gradient descent (GD), Levenberg—Marquardt optimization technique (LM) and sliding mode-based tuning (SM). The model and its properties are discussed comparatively.