Table of contents

The 12th European Workshop on Advanced Control and Diagnosis (ACD 2015) took place at the Research Centre NTIS - New Technologies for the Information Society, Faculty of Applied Sciences, University of West Bohemia, Pilsen, Czech Republic, on November 19 - 20, 2015.

The annual European Workshop on Advanced Control and Diagnosis has been organized since 2003 by Control Engineering departments of several European universities in Germany, France, the UK, Poland, Italy, Hungary, and Denmark to bring together senior and junior academics and engineers from diverse fields of automatic control, fault detection, and signal processing. The workshop provides an opportunity for researchers and developers to present their recent theoretical developments, practical applications, or even open problems. It also offers a great opportunity for industrial partners to express their needs and priorities and to review the current activities in the fields.

A total of 74 papers have been submitted for ACD 2015. Based on the peer reviews 48 papers were accepted for the oral presentation and 10 papers for the poster presentation. The accepted papers covered areas of control theory and applications, identification, estimation, signal processing, and fault detection. In addition, four excellent plenary lectures were delivered by Prof. Fredrik Gustafsson (Automotive Sensor Mining for Tire Pressure Monitoring), Prof. Vladimír Havlena (Advanced Process Control for Energy Efficiency), Prof. Silvio Simani (Advanced Issues on Wind Turbine Modelling and Control), and Prof. Robert Babuška (Learning Control in Robotics). The ACD 2015 was for the first time in the workshop history co-sponsored by the International Federation of Automatic Control (IFAC).

On behalf of the ACD 2015 organising committee, we would like to thank all those who prepared and submitted papers, participated in the peer review process, supported, and attended the workshop.

All papers published in this volume of Journal of Physics: Conference Series have been peer reviewed through processes administered by the proceedings Editors. Reviews were conducted by expert referees to the professional and scientific standards expected of a proceedings journal published by IOP Publishing.

The motivation for this paper comes from a real need to have an overview about the challenges of modelling and control for very demanding systems, such as wind turbine systems, which require reliability, availability, maintainability, and safety over power conversion efficiency. These issues have begun to stimulate research and development in the wide control community particularly for these installations that need a high degree of "sustainability". Note that this topic represents a key point mainly for offshore wind turbines with very large rotors, since they are characterised by challenging modelling and control problems, as well as expensive and safety critical maintenance works. In this case, a clear conflict exists between ensuring a high degree of availability and reducing maintenance times, which affect the final energy cost. On the other hand, wind turbines have highly nonlinear dynamics, with a stochastic and uncontrollable driving force as input in the form of wind speed, thus representing an interesting challenge also from the modelling point of view. Suitable control methods can provide a sustainable optimisation of the energy conversion efficiency over wider than normally expected working conditions. Moreover, a proper mathematical description of the wind turbine system should be able to capture the complete behaviour of the process under monitoring, thus providing an important impact on the control design itself. In this way, the control scheme could guarantee prescribed performance, whilst also giving a degree of "tolerance" to possible deviation of characteristic properties or system parameters from standard conditions, if properly included in the wind turbine model itself. The most important developments in advanced controllers for wind turbines are addressed, and open problems in the areas of modelling of wind turbines are also outlined.

Design conditions for existence of the H_∞ linear state feedback control for discretetime stochastic systems with state-multiplicative noise and polytopic uncertainties are presented in the paper. Using an enhanced form of the bounded real lemma for discrete-time stochastic systems with state-multiplicative noise, the LMI-based procedure is provided for computation of the gains of linear, as well as nonlinear, state control law. The approach is illustrated on an example demonstrating the validity of the proposed method.

This paper introduces an adaptation of the classical linear control theory representation of zeros, poles and gain into a bilinear approach. The placement of poles at the complex plane is a complete description of plants dynamics; hence it is a convenient form from which calculation of various properties, e.g. rise time, settling time, is plausible. Such technique can be adjusted into the bilinear structure if poles of a quasi-linear representation (linear with respect to input) are concerned. The research outcomes with conclusion on the equivalent poles displacement and generalized rules for a 2nd order bilinear system equivalent poles input dependent loci. The proposed approach seems to be promising, as simplification of design and identification of a bilinear system increases transparency during modelling and control in practical applications and hence it may be followed by applicability of such structure in common industrial problems.

This paper deals with the problem of designing an iterative learning control algorithm for discrete linear systems using repetitive process stability theory. The resulting design produces a stabilizing output feedback controller in the time domain and a feedforward controller that guarantees monotonic convergence in the trial-to-trial domain. The results are also extended to limited frequency range design specification. New design procedure is introduced in terms of linear matrix inequality (LMI) representations, which guarantee the prescribed performances of ILC scheme. A simulation example is given to illustrate the theoretical developments.

In this paper a fault tolerant control strategy that combines the backstepping procedure and the quasi-continuous high-order sliding mode controller is proposed. The fault tolerance principle is based on a hierarchical application of the backstepping methodology ensuring the finite time convergence of the desired system states, in spite of the considered fault situations. The additive effect of the faults and disturbances is canceled out by the hierarchical application of the quasi-continuous controller ensuring fault-tolerance. The effect of Lebesgue measurable noise over the precision of the proposed controller is studied. Simulation results based on a nonlinear model of the F16 jet fighter show the efficiency of the proposed techniques.

The paper deals with dual adaptive control problem, where the functional uncertainties in the system description are modelled by a non-parametric Gaussian process regression model. Current approaches to adaptive control based on Gaussian process models are severely limited in their practical applicability, because the model is re-adjusted using all the currently available data, which keeps growing with every time step. We propose the use of recursive Gaussian process regression algorithm for significant reduction in computational requirements, thus bringing the Gaussian process-based adaptive controllers closer to their practical applicability. In this work, we design a bi-criterial dual controller based on recursive Gaussian process model for discrete-time stochastic dynamic systems given in an affine-in-control form. Using Monte Carlo simulations, we show that the proposed controller achieves comparable performance with the full Gaussian process-based controller in terms of control quality while keeping the computational demands bounded.

The paper presents new conditions, adequate in design of proportional-integral virtual actuators and utilizable in fault-tolerant control structures which are stabilizable by dynamic output controllers. Taking into account disturbance conditions and changes of variables after the virtual actuator activation, the design conditions are outlined in terms of the linear matrix inequalities within the bounded real lemma forms. Using tuning parameters in design, and with suitable choice of the order of dynamic output controller, the approach provides a way to obtain acceptable dynamics of the closed loop system after activation of the virtual actuator.

The aim of this paper is to develop and design a State Space Neural Network toolbox for a non-linear system identification with an artificial state-space neural networks, which can be used in a model-based robust fault diagnosis and control. Such toolbox is implemented in the MATLAB environment and it uses some of its predefined functions. It is designed in the way that any non-linear multi-input multi-output system is identified and represented in the classical state-space form. The novelty of the proposed approach is that the final result of the identification process is the state, input and output matrices, not only the neural network parameters. Moreover, the toolbox is equipped with the graphical user interface, which makes it useful for the users not familiar with the neural networks theory.

It is often not feasible or even impossible to identify a plant in open loop. This might be because the plant contains unstable poles, or it is simply too expensive to remove the plant from its intended operation, among other possibilities. There are several methods for identifying a plant in closed loop [4], and one such method is the Hansen scheme [1]. Standard identification using Hansen scheme demands generating the identification signals indirectly. In this paper it is instead proposed to use the relationship between the Youla factorization of a plant and its stabilizing controller to directly measure the signals used for identification. A simulation example and identification of a gas bearing is given to show the method in action. Rotors supported by controllable gas bearings are open loop stable systems. However as the rotational speed is increased feedback control is necessary in order to keep the system stable. Furthermore because the dynamics of such a system depends on the rotational speed it is needed to conduct an identification while the system is part of a closed loop scheme. The authors believe the paper able to contribute towards a simpler and more direct way of identifying closed loop plants using Hansen scheme.

Control of the plasma field in the tokamak requires reliable estimation of the plasma boundary. The plasma boundary is given by a complex mathematical model and the only available measurements are responses of induction coils around the plasma. For the purpose of boundary estimation the model can be reduced to simple linear regression with potentially infinitely many elements. The number of elements must be selected manually and this choice significantly influences the resulting shape. In this paper, we investigate the use of formal model structure estimation techniques for the problem. Specifically, we formulate a sparse least squares estimator using the automatic relevance principle. The resulting algorithm is a repetitive evaluation of the least squares problem which could be computed in real time. Performance of the resulting algorithm is illustrated on simulated data and evaluated with respect to a more detailed and computationally costly model FREEBIE.

This work is aimed at a comprehensive discussion of algorithms for the kinematic parameters identification of robotic manipulators. We deal with an open-loop geometric calibration task, when a full 6D robot's end-effector pose is measured. Effective solutions of such a task is of high interest in many practical applications, because it can dramatically improve key robot characteristics. On the first step, we select optimal calibration configurations. A comparative analysis of three different algorithms and two observability indexes used for numerical optimization is provided. Afterwards, using the acquired and pre-processed experimental data we identify modified Denavit-Hartenberg parameters of the manipulator. Estimates are obtained resolving original nonlinear forward kinematics relations. Finally, we compare nominal and calibrated geometric parameters and show how much deviations in these parameters affect robot positioning accuracy. To the best of our knowledge, such integrated efforts are new for the KUKA LWR4+ robot and Nikon K610 optical coordinate measuring machine (CMM), which were used in the study. Discussion of practical issues on how to organise the experiment is an additional contribution of this work. The proposed procedure is highly automated and can be implemented to improve manipulator's performance on a periodic basis.

Since wind turbines became one of the most often source of renewable energy, appropriate health and condition monitoring systems are required. Especially proper monitoring of offshore plants is very significant because the accessibility is difficult and inspections are very costly. In comparison with conventional rotating machine vibration monitoring, where steady conditions and stationary signal are usually assumed, the wind turbines are characterized by unsteady conditions due to variable rotational speed. Hence the vibration signal is non-stationary and interpretation of signal signatures may be more complex. The common approach to analyze such non-stationary signals is the use of a time-frequency method, usually Short-Time Fourier Transform, which is the most popular one due to its simplicity. Nevertheless, there are other methods which can give a different view at the analyzed data and provide new information. This article investigates the potential use of some other time-frequency methods, namely Wavelet Transform, Wigner-Ville distribution and Hilbert-Huang transform in wind plants monitoring systems and apply these methods to real measured data with additional simulated bearing fault signal. Finally, the mentioned methods are compared based on computational complexity, readability and interpretability. Though the last two criteria are very subjective, Short-Time Fourier Transform was finally chosen as the most effective method followed by Wavelet Transform.

This paper presents the basic principle of Prony's method, which is used for determining electrical power system parameters such as phasors, angular speeds as well as damping factors in a large frequency range. For fixed sampling frequencies, the approach to finding an optimal model order and to extracting transient parameters for further identification of different system effects is shown. Furthermore, it is shown that the method can be applied to the diverse plausibility checks for a confirmation of the result obtained using a conventional Fourier based approach. It will be presented that the significant advantage of the Prony algorithm in comparison to Fourier based methods is the possibility of computing an accurate frequency and in determining a damping factor.

In this paper an Advanced Process Control system aimed at controlling and optimizing a pusher type reheating furnace located in an Italian steel plant is proposed. The designed controller replaced the previous control system, based on PID controllers manually conducted by process operators. A two-layer Model Predictive Control architecture has been adopted that, exploiting a chemical, physical and economic modelling of the process, overcomes the limitations of plant operators' mental model and knowledge. In addition, an ad hoc decoupling strategy has been implemented, allowing the selection of the manipulated variables to be used for the control of each single process variable. Finally, in order to improve the system flexibility and resilience, the controller has been equipped with a supervision module. A profitable trade-off between conflicting specifications, e.g. safety, quality and production constraints, energy saving and pollution impact, has been guaranteed. Simulation tests and real plant results demonstrated the soundness and the reliability of the proposed system.

This paper concerns a control approach of a fleet of Unmanned Aerial Vehicles (UAV) based on virtual leader. Among others, optimization methods are used to develop the virtual leader control approach, particularly the particle swarm optimization method (PSO). The goal is to find optimal positions at each instant of each UAV to guarantee the best performance of a given task by minimizing a predefined objective function. The UAVs are able to organize themselves on a 2D plane in a predefined architecture, following a mission led by a virtual leader and simultaneously avoiding collisions between various vehicles of the group. The global proposed method is independent from the model or the control of a particular UAV. The method is tested in simulation on a group of UAVs whose model is treated as a double integrator. Test results for the different cases are presented.

The paper deals with development of software framework for rapid generation of remote virtual laboratories. Client-server architecture is chosen in order to employ real-time simulation core which is running on a dedicated server. Ordinary web browser is used as a final renderer to achieve hardware independent solution which can be run on different target platforms including laptops, tablets or mobile phones. The provided toolchain allows automatic generation of the virtual laboratory source code from the configuration file created in the open- source Inkscape graphic editor. Three virtual laboratories presenting advanced motion control algorithms have been developed showing the applicability of the proposed approach.

Gas bearings are popular for their high speed capabilities, low friction and clean operation, but require low clearances and suffer from poor damping properties. The poor damping properties cause high disturbance amplification near the natural frequencies. These become critical when the rotation speed coincides with a natural frequency. In these regions, even low mass unbalances can cause rub and damage the machine. To prevent rubbing, the variation of the rotation speed of machines supported by gas bearings has to be carefully conducted during run-ups and run-downs, by acceleration and deceleration patterns and avoidance of operation near the critical speeds, which is a limiting factor during operation, specially during run-downs. An approach for reducing the vibrations is by feedback controlled lubrication. This paper addresses the challenge of reducing vibrations in rotating machines supported by gas bearings to extend their operating range. Using H_∞-design methods, active lubrication techniques are proposed to enhance the damping, which in turn reduces the vibrations to a desired safe level. The control design is validated experimentally on a laboratory test rig, and shown to allow safe shaft rotation speeds up to, in and above the two first critical speeds, which significantly extends the operating range.

Energy is an important factor in today's industrial environment. Pump systems account for about 20% of the total industrial electrical energy consumption. Several studies show that with proper monitoring, control and maintenance, the efficiency of pump systems can be increased. Controlling pump systems with intelligent systems can help to reduce a pump's energy consumption by up to one third of its original consumption. The research in this paper was carried out in the scope of a research project which involves modelling and simulation of pump systems. This paper focuses on the future implementation of modelling capabilities in PLCs (programmable logic controllers). The whole project aims to use a pump itself as the sensor rather than introducing external sensors into the system, which would increase the cost considerably. One promising approach for an economic and robust industrial implementation of this intelligence is the use of PLCs. PLCs can be simulated in multiple ways; in this project, Codesys was chosen for several reasons which are explained in this paper. The first part of this paper explains the modelling of a pump itself, the process load of the asynchronous motor with a control system, and the simulation possibilities of the motor in Codesys. The second part describes the simulation and testing of a system realized. The third part elaborates the Codesys system structure and interfacing of the system with external files. The final part consists of comparing the result with an earlier Matlab/SIMULINK model and original test data.

Embedded control units for transportation systems make use of advanced nonlinear control methods. In this research article a new nonlinear control method is applied to spark ignited (SI) engines. The proposed SI engine's control scheme is based on differential flatness theory The considered method succeeds the efficient control of the SI engine parameters such as intake pressure and turn speed. The method makes use of a state-space model of the SI-engine in the so-called triangular form. The controller design proceeds by showing that each row of the state-space model of the SI engine stands for a differentially flat system, where the flat output is chosen to be the associated state variable. Next, for each subsystem which is linked with a row of the state-space model, a virtual control input is computed, that can invert the subsystem's dynamics and can eliminate the subsystem's tracking error. From the last row of the state-space description, the control input that is actually applied to the SI engine is found. This control input contains recursively all virtual control inputs which were computed for the individual subsystems associated with the previous rows of the state-space equation. Thus, by tracing the rows of the state-space model backwards, at each iteration of the control algorithm, one can finally obtain the control input that should be applied to the SI-engine so as to assure that all its state vector elements will converge to the desirable setpoints.

This paper addresses the estimation problem for discrete-time systems where both measurements and control commands are sent to a central station through a lossy network without delivery acknowledgment. The central unit implements the estimation and control algorithms. We propose a jump observer that uses the expected value of the unknown control input at the actuator to run an open loop estimation. Then, the absence of acknowledgment in the control input transmission is dealt with through the introduction of an unknown disturbance. The observer update is performed by means of jumping gains when there are available measurements. We employ an statistic of the control error (new disturbance), i.e., the difference between the control inputs at the plant and at the observer, to schedule the observer gains in real time. The observer is designed to minimize the H_∞ norm from disturbances, measurement noises and control input errors, to estimation error. The infinite-dimensional design problem is turn into an optimization problem over polynomials using sum-of-squares decomposition techniques. Benefits of the proposal are shown in a simulation example.

This paper introduces a thermosensing embedded system with a sensor bus that uses wavelets for the purposes of noise location and denoising. From the principle of the filter bank the measured signal is separated in two bands, low and high frequency. The proposed algorithm identifies the defined noise in these two bands. With the Wavelet Packet Transform as a method of Discrete Wavelet Transform, it is able to decompose and reconstruct bus input signals of a sensor network. Using a seminorm, the noise of a sequence can be detected and located, so that the wavelet basis can be rearranged. This particularly allows for elimination of any incoherent parts that make up unavoidable measuring noise of bus signals. The proposed method was built based on wavelet algorithms from the WaveLab 850 library of the Stanford University (USA). This work gives an insight to the workings of Wavelet Transformation.

The paper deals with state estimation of nonlinear stochastic dynamic systems. In particular, the iterated extended Kalman filter is studied. Three recently proposed iterated extended Kalman filter algorithms are analyzed in terms of their performance and specification of a user design parameter, more specifically the step-length size. The performance is compared using the root mean square error evaluating the state estimate and the noncredibility index assessing covariance matrix of the estimate error. The performance and influence of the design parameter, are analyzed in a numerical simulation.

In this paper, the problem of interval observer design for LTI uncertain systems is addressed. The aim is to estimate simultaneously the upper and lower bounds of unmeasurable states and unknown inputs. Based on a set membership approach, the proposed method interest consists in reducing the estimate conservatism assuming that all known and unknown inputs, disturbances, uncertainties and noises are bounded with a priori known bounds. For that, decoupling methods are used to reduce the latter propagation. The estimation problem is led back to a singular model representation. An example is given to illustrate the proposed design method.

This paper deals with state estimation on a pressurized water pipe modeled by nonlinear coupled distributed hyperbolic equations for non-conservative laws with three known boundary measures. Our objective is to estimate the fourth boundary variable, which will be useful for leakage detection. Two approaches are studied. Firstly, the distributed hyperbolic equations are discretized through a finite-difference scheme. By using the Lipschitz property of the nonlinear term and a Lyapunov function, the exponential stability of the estimation error is proven by solving Linear Matrix Inequalities (LMIs). Secondly, the distributed hyperbolic system is preserved for state estimation. After state transformations, a Luenberger-like PDE boundary observer based on backstepping mathematical tools is proposed. An exponential Lyapunov function is used to prove the stability of the resulted estimation error. The performance of the two observers are shown on a water pipe prototype simulated example.

The paper focuses on robust state estimation and control for nonlinear system with uncertain parameters. In particular, the problem is oriented towards a practical application for a laboratory three-tank system. The proposed approach starts with a general description of the system and assumptions regarding uncertain parameters and a nonlinear function. The subsequent part of the paper is concerned with the design of the robust observer and controller for nonlinear systems. To confirm the performance of the proposed approaches, the final part presents results obtained for the laboratory three-tank system.

This paper presents a LPV (Linear Parameter Varying) solution for a mixed passive-active architecture used to mitigate the microvibrations generated by reaction wheels in satellites. In particular, H_∞/LPV theory is used to mitigate low frequency disturbances, current baseline for high frequency microvibration mitigation being based on elastomer materials. The issue of multiple harmonic microvibrations is also investigated. Simulation results from a test benchmark provided by Airbus Defence and Space demonstrate the potential of the proposed method.

This paper analyses a functional control of the Robotino. The proposed control strategy considers a functional decoupling control strategy realized using a geometric approach and the invertibility property of the DC-drives with which the Robotino is equipped. For a given control structure the functional controllability is proven for motion trajectories of class C³, continuous functions with third derivative also being continuous. Horizontal, Vertical and Angular motions are considered and the decoupling between these motions is obtained. Control simulation results using real data of the Robotino are shown. The used control which enables to produce the presented results is a standard Linear Model Predictive Control (LMPC), even though for sake of brevity the standard algorithm is not shown.

Process control is a challenging research topic for both academia and industry for a long time. Controllers evolved from the classical SISO approach to modern fuzzy or neuro-fuzzy embedded devices with networking capabilities, however PID algorithms are still used in the most industrial control loops. In this paper, we focus on the implementation of a PID controller using mbed NXP LPC1768 development board. This board integrates a powerful ARM Cortex- M3 core and has networking capabilities. The implemented controller can be remotely operated by using an Internet connection and a standard Web browser. The main advantages of the proposed embedded system are customizability, easy operation and very low power consumption. The experimental results obtained by using a simulated process are analysed and shows that the implementation can be done with success in industrial applications.

This paper proposes a controller for motion control of the Robotino. The proposed controller considers a functional decoupling control strategy realized using a geometric approach and the invertibility property of the DC-drives with which the Robotino is equipped. Horizontal, Vertical and Angular motions are considered and once the decoupling between these motions is obtained, a Model Predictive Control (MPC) strategy is used in combination with the inverse DC-drive model. Simulation results using real data of Robotino are shown.

This paper presents conceptual control solution for reliable and energy efficient operation of heating, ventilation and air conditioning (HVAC) systems used in large volume building applications, e.g. warehouse facilities or exhibition centres. Advanced two-level scalable control solution, designed to extend capabilities of the existing low-level control strategies via remote internet connection, is presented. The high-level, supervisory controller is based on Model Predictive Control (MPC) architecture, which is the state-of-the-art for indoor climate control systems. The innovative approach benefits from using passive heating and cooling control strategies for reducing the HVAC system operational costs, while ensuring that required environmental conditions are met.

Distribution Static Compensator (DSTATCOM) has been used as a custom power device for voltage regulation and load compensation in the distribution system. Controlling the switching angle has been the biggest challenge in DSTATCOM. Till date, Proportional Integral (PI) controller is widely used in practice for load compensation due to its simplicity and ability. However, PI Controller fails to perform satisfactorily under parameters variations, nonlinearities, etc. making it very challenging to arrive at best/optimal tuning values for different operating conditions. Fuzzy logic and neural network based controllers require extensive training and perform better under limited perturbations. Model predictive control (MPC) is a powerful control strategy, used in the petrochemical industry and its application has been spread to different fields. MPC can handle various constraints, incorporate system nonlinearities and utilizes the multivariate/univariate model information to provide an optimal control strategy. Though it finds its application extensively in chemical engineering, its utility in power systems is limited due to the high computational effort which is incompatible with the high sampling frequency in these systems. In this paper, we propose a DSTATCOM based on Finite Control Set Model Predictive Control (FCS-MPC) with Instantaneous Symmetrical Component Theory (ISCT) based reference current extraction is proposed for load compensation and Unity Power Factor (UPF) action in current control mode. The proposed controller performance is evaluated for a 3 phase, 3 wire, 415 V, 50 Hz distribution system in MATLAB Simulink which demonstrates its applicability in real life situations.

In terms of system diagnosis, several studies are generally performed. The diagnosis is composed of three different parts: detecting, isolating and estimating the value of the faults. If many results have been obtained for linear systems with a known model, the situation is quite different in the case of nonlinear systems behavior, especially when the model is not known a priori. This paper proposes to discuss the latter case using a study of the dissimilarities between data. The dissimilarities are evaluated by a nonlinear function of the Euclidean distances. To this end, a radial basis function is used, and a directional study is introduced for fault diagnosis. The relevance of the proposed technique is illustrated on simulated data.

In this work, we present a fault detection strategy applicable to the blade and pitch system in offshore wind turbines. First, we model the system and possible faults and propose a PI observer to identify the faults. Then, the observer is designed accounting the sensors measurement noise, and addressing a trade off between the needs of false alarm rate, minimum detectable fault and detection time. By means of a well known benchmark, several simulations show the goodness of the approach and its flexibility to explicitly fix the fault detector performance.

This paper deals with the application of state space neural network model to design a Fault Detection and Isolation diagnostic system. The work describes approach based on multimodel solution where the SIMO process is decomposed into simple models (SISO and MISO). With such models it is possible to generate different residual signals which later can be evaluated with simple thresholding method into diagnostic signals. Further, such diagnostic signals with the application of Binary Diagnostic Table (BDT) can be used to fault isolation. All data used in experiments is obtain from the simulator of the real-time laboratory stand of Modular Servo under Matlab/Simulink environment.

Data-driven approaches are widely applied for fault detection in industrial process. Recently, a new method for fault detection called principal component pursuit(PCP) is introduced. PCP is not only robust to outliers, but also can accomplish the objectives of model building, fault detection, fault isolation and process reconstruction simultaneously. PCP divides the data matrix into two parts: a fault-free low rank matrix and a sparse matrix with sensor noise and process fault. The statistics presented in this paper fully utilize the information in data matrix. Since the low rank matrix in PCP is similar to principal components matrix in PCA, a T² statistic is proposed for fault detection in low rank matrix. And this statistic can illustrate that PCP is more sensitive to small variations in variables than PCA. In addition, in sparse matrix, a new monitored statistic performing the online fault detection with PCP-based method is introduced. This statistic uses the mean and the correlation coefficient of variables. Monte Carlo simulation and Tennessee Eastman (TE) benchmark process are provided to illustrate the effectiveness of monitored statistics.

An adaptation of unitary system principle in fault detection filter design for continuous-time linear MIMO systems is presented in the paper. The conformation is based on an enhanced fault residual transfer function matrix with unitary construction and offers the key advantages on providing high residual sensitivity with respect to faults. Reflecting the emplacement of singular values in unitary construction, an associated structure of linear matrix inequalities with built-in structured constraints is outlined to verify the filter stability. The proposed design conditions are verified by the numerical illustrative example.

Fault detection for automotive semi-active shock absorbers is a challenge due to the non-linear dynamics and the strong influence of the disturbances such as the road profile. First obstacle for this task, is the modeling of the fault, which has been shown to be of multiplicative nature. Many of the most widespread fault detection schemes consider additive faults. Two model-based fault algorithms for semiactive shock absorber are compared: an observer-based approach and a parameter identification approach. The performance of these schemes is validated and compared using a commercial vehicle model that was experimentally validated. Early results shows that a parameter identification approach is more accurate, whereas an observer-based approach is less sensible to parametric uncertainty.

An optimization based method is proposed in this paper for the computation of Lyapunov functions and regions of attractions for nonlinear systems containing polynomial and rational terms. The Lyapunov function is given in a special quadratic form, and the negativity of its derivative is ensured using appropriate LMI conditions. The conservatism of the solution is reduced by utilizing Finsler's lemma. The number of monomial and rational terms in the computational problem is kept as low as possible using linear fractional transformation (LFT) and automatic model simplification steps. The operation of the method is illustrated on two examples taken from the literature.

In this paper, the new method of the positive minimal realisation two-dimensional linear system with delays described by general model using multi-dimensional digraphs theory D⁽ⁿ⁾ has been presented. For the proposed method, a digraph-based algorithm was constructed. The algorithm is based on a parallel computing method to gain needed speed and computational power for such a solution. The proposed solution allows minimal digraphs construction for any positive two-dimensional linear system with delays. The proposed method was discussed and illustrated with some numerical examples.

The article proves the separation theorem for optimal control of stochastic systems in the case when an observed continuous-time process possess memory of arbitrary ration relating to a state vector.

This paper presents a method of the determination of a positive stable realisation of the fractional continuous-time positive system consisting of n subsystems with one fractional order and with different fractional orders. For the proposed method, a digraph-based algorithm was constructed. In this paper, we have shown how we can realise the transfer matrix using electrical circuits consisting of resistances, inductances, capacitances and source voltages. The proposed method was discussed and illustrated with some numerical examples.

The article proves the separation theorem for optimal control of stochastic systems in the case when an observed continuous-discrete-time process possess memory of arbitrary ration relating to a state vector.

Autocorrelation and non-normality of process characteristic variables are two main difficulties that industrial engineers must face when they should implement control charting techniques. This paper presents new issues regarding the probability distribution of wavelets coefficients. Firstly, we highlight that wavelets coefficients have capacities to strongly decrease autocorrelation degree of original data and are normally-like distributed, especially in the case of Haar wavelet. We used AR(1) model with positive autoregressive parameters to simulate autocorrelated data. Illustrative examples are presented to show wavelets coefficients properties. Secondly, the distributional parameters of wavelets coefficients are derived, it shows that wavelets coefficients reflect an interesting statistical properties for SPC purposes.

One of the issues of steam turbines diagnostics is monitoring of rotor thermal stress that arises from nonuniform temperature field. The effort of steam turbine operator is to operate steam turbine in such conditions, that rotor thermal stress doesn't exceed the specified limits. If rotor thermal stress limits are exceeded for a long time during machine operation, the rotor fatigue life is shortened and this may lead to unexpected machine failure. Thermal stress plays important role during turbine cold startup, when occur the most significant differences of temperatures through rotor cross section. The temperature field can't be measured directly in the entire rotor cross section and standardly the temperature is measured by thermocouple mounted in stator part. From this reason method for numerical solution of partial differential equation of heat propagation through rotor cross section must be combined with method for calculation of temperature on rotor surface. In the first part of this article, the application of finite volume method for calculation of rotor thermal stress is described. The second part of article deals with optimal trend generation of thermal flux, that could be used for optimal machine loading.

The increasing amount of data in industrial automation systems overburdens the user in process control and diagnosis tasks. One possibility to cope with these challenges consists of using smart assistance systems that automatically monitor and optimize processes. This article deals with aspects of data-driven assistance systems such as assistance functions, process models and data acquisition. The paper describes novel approaches for self-diagnosis and self-optimization, and shows how these assistance functions can be integrated in different industrial environments. The considered assistance functions are based on process models that are automatically learned from process data. Fault detection and isolation is based on the comparison of observations of the real system with predictions obtained by application of the process models. The process models are further employed for energy efficiency optimization of industrial processes. Experimental results are presented for fault detection and energy efficiency optimization of a drive system.

The paper deals with probabilistic methods for designing the active fault detectors that improve the quality of detection using an auxiliary input signal. Two probabilistic methods that assume a similar stochastic model of a monitored system are considered and compared with a special attention to various difficulties associated with active fault detector designs. The active fault detector design based on a general detection cost function is compared with the model sequence selection error minimization design in terms of assumptions and theoretical properties. Practical aspects of both methods are also considered and demonstrated through a numerical example.

Traditional fault diagnosis methods based on hidden Markov model (HMM) use a unified method for feature extraction, such as principal component analysis (PCA), kernel principal component analysis (KPCA) and independent component analysis (ICA). However, every method has its own limitations. For example, PCA cannot extract nonlinear relationships among process variables. So it is inappropriate to extract all features of variables by only one method, especially when data characteristics are very complex. This article proposes a switched feature extraction procedure using PCA and KPCA based on nonlinearity measure. By the proposed method, we are able to choose the most suitable feature extraction method, which could improve the accuracy of fault diagnosis. A simulation from the Tennessee Eastman (TE) process demonstrates that the proposed approach is superior to the traditional one based on HMM and could achieve more accurate classification of various process faults.

In this paper a pseudo-sliding mode control is proposed by introducing a continuous and smooth input signal in order to guarantee both chattering elimination and boundedness of sliding variable derivative in the presence of non-zero external disturbance. For this purpose, having fixed a suitable sliding manifold, a homogeneous differential equation describing the sliding variable evolution is considered. It is discussed later in this paper that the input signal formed on the basis of this equation provides asymptotic convergence of the sliding variable and its derivative to zero as well as the asymptotic stability of the non-linear system in the absence of external disturbance. The dynamics of the system affected by non-zero external disturbance make the state vector converge to domains in a vicinity of the origin at the exponential rate, as the norm of arbitrary trajectory is limited to decreasing exponential function. In order to expand the variety of controllers based on a reaching law and providing the above-mentioned properties, a certain class of functions is presented.

The research topic of autonomous cars and the communication among them has attained much attention in the last years and is developing quickly. Among others, this research area spans fields such as image recognition, mathematical control theory, communication networks, and sensor fusion. We consider an intersection scenario where we divide the shared road space in different cells. These cells form a grid. The cars are modelled as an autonomous multi-agent system based on the Distributed Model Predictive Control algorithm (DMPC). We prove that the overall system reaches stability using Optimal Control for each multi-agent and demonstrate that by numerical results.

Due to the aging population more and more people require mobility assistance in form of a wheelchair. Generally it would be desirable that such wheelchairs would be easy to use and would allow their users the possibility to move in any direction at any time. Concepts which allow such movements are existing since many years but have for several reasons not found their way to the market. Additionally for semi-autonomous (assisted) operation and fully autonomous operation (e. g. an empty wheelchair driving to its charging station) the control task is much less challenging for such drive system, because no complex manoeuvres needs to be considered and planned. In an ongoing research a drive system for a wheelchair was developed which offers such possibilities employing a relatively simple mechanical design. This drive system is based on a certain steering principle which is based on torque differences between different wheels. This allows a relatively simple mechanical design but poses challenges on the control of the vehicle. This paper describes two possible approaches to address this challenge - the use of an event based control and the application of multiple software agents. Both approaches can solve the control problem individually but can also complement each other for better system performance. The paper stars with a description of the wheelchair drive system. Then the asynchronous event based control software is described as well the multi agent based approach. The next sections report the results of the experiments and discuss the further improvements.

This paper deals with advanced methods for the evaluation of a bladed disc behavior in terms of the wheel vibration and blade service time consumption. These methods are developed as parts of the noncontact vibration monitoring system of the steam turbine shrouded blades. The proposed methods utilize the time-frequency processing (cross spectra) and the method using least squares to analyse the data from the optical and magnetoresistive sensors, which are mounted in the stator radially above the rotor blades. Fundamentally, the blade vibrations are detected during the blade passages under the sensors and the following signal processing, which covers also the proposed methods, leads to the estimation of the blade residual service life. The prototype system implementing above mentioned techniques was installed into the last stage of the new steam turbine (LP part). The methods for bladed disc mode shape evaluation were successfully verified on the signals, which were obtained during the commission operation of the turbine.

This paper deals with the state feedback control problem for linear uncertain systems subject to both matched and unmatched perturbations. The proposed control law is based on an the Integral Sliding Mode Control (ISMC) approach to tackle matched perturbations as well as the H_∞ paradigm for robustness against unmatched perturbations. The proposed method also parallels the work presented in [1] which addressed the same problem and proposed a solution involving an Algebraic Riccati Equation (ARE)-based formulation. The contribution of this paper is concerned by the establishment of a Linear Matrix Inequality (LMI)-based solution which offers the possibility to consider other types of constraints such as 𝓓-stability constraints (pole assignment-like constraints). The proposed methodology is applied to a pilot three-tank system and experiment results illustrate the feasibility. Note that only a few real experiments have been rarely considered using SMC in the past. This is due to the high energetic behaviour of the control signal.

It is important to outline that the paper does not aim at proposing a LMI formulation of an ARE. This is done since 1971 [2] and further discussed in [3] where the link between AREs and ARIs (algebraic Riccati inequality) is established for the H_∞ control problem. The main contribution of this paper is to establish the adequate LMI-based methodology (changes of matrix variables) so that the ARE that corresponds to the particular structure of the mixed ISMC/H_∞ structure proposed by [1] can be re-formulated within the LMI paradigm.

The paper deals with development of a unified framework for generation of 3D visualizations of complex mechatronic systems. It provides a high-fidelity representation of executed motion by allowing direct employment of a machine geometry model acquired from a CAD system. Open-architecture multi-platform solution based on latest web standards is achieved by utilizing a web browser as a final 3D renderer. The results are applicable both for simulations and development of real-time human machine interfaces. Case study of autonomous underwater vehicle control is provided to demonstrate the applicability of the proposed approach.

In model-based diagnosis (MBD), structural models can provide useful information for fault diagnosis and fault-tolerant control design. In particular, they are known for supporting the design of analytical redundancy relations (ARRs) which are widely used to generate residuals for diagnosis. On the other hand, systems are increasingly complex whereby it is necessary to develop decentralized architectures to perform the diagnosis task. Decentralized diagnosis is of interest for on-board systems as a way to reduce computational costs or for large geographically distributed systems that require to minimizing data transfer. Decentralized solutions allow proper separation of industrial knowledge, provided that inputs and outputs are clearly defined. This paper builds on the results of [1] and proposes an optimized approach for decentralized fault-focused residual generation. It also introduce the concept of Fault-Driven Minimal Structurally-Overdetermined set (FMSO) ensuring minimal redundancy. The method decreases communication cost involved in decentralization with respect to the algorithm proposed in [1] while still maintaining the same isolation properties as the centralized approach as well as the isolation on request capability.

The high complexity of manufacturing processes and the continuously growing amount of data lead to excessive demands on the users with respect to process monitoring, data analysis and fault detection. For these reasons, problems and faults are often detected too late, maintenance intervals are chosen too short and optimization potential for higher output and increased energy efficiency is not sufficiently used. A possibility to cope with these challenges is the development of self-learning assistance systems, which identify relevant relationships by observation of complex manufacturing processes so that failures, anomalies and need for optimization are automatically detected. The assistance system developed in the present work accomplishes data acquisition, process monitoring and anomaly detection in industrial and agricultural processes. The assistance system is evaluated in three application cases: Large distillation columns, agricultural harvesting processes and large-scale sorting plants. In this paper, the developed infrastructures for data acquisition in these application cases are described as well as the developed algorithms and initial evaluation results.

This paper is concerned with the integration of control and diagnosis functionalities into the development of complete systems which include mechanical, electrical and electronic subsystems. For the development of such systems the strategies, methods and tools of integrated product development have attracted significant attention during the last decades. Today, it is generally observed that product development processes of complex systems can only be successful if the activities in the different domains are well connected and synchronised and if an ongoing communication is present - an ongoing communication spanning the technical domains and also including functions such as production planning, marketing/distribution, quality assurance, service and project planning. Obviously, numerous approaches to tackle this challenge are present in scientific literature and in industrial practice, as well. Today, the functionality and safety of most products is to a large degree dependent on control and diagnosis functionalities. Still, there is comparatively little research concentrating on the integration of the development of these functionalities into the overall product development processes. The main source of insight of the presented research is the product development process of an Automated Guided Vehicle (AGV) which is intended to be used on rough terrain. The paper starts with a background describing Integrated Product Development. The second section deals with the product development of the sample product. The third part summarizes some insights and formulates first hypotheses concerning control and diagnosis in Integrated Product Development.

The paper deals with identification of the noise covariance matrices affecting the linear system described by the state-space model. In particular, the stress is laid on the autocovariance least-squares method which belongs into to the class of the correlation methods. The autocovariance least-squares method is revised for a general linear stochastic dynamic system and is implemented within the publicly available MATLAB toolbox Nonlinear Estimation Framework. The toolbox then offers except of a large set of state estimation algorithms for prediction, filtering, and smoothing, the integrated easy-to-use method for the identification of the noise covariance matrices. The implemented method is tested by a thorough Monte-Carlo simulation for various user-defined options of the implemented method.