Keywords: Fault tolerant control / fault recovery
Abstract: This paper proposes a fault-tolerant control (FTC) strategy based on virtual actuators for discrete-time descriptor systems subject to multiplicative and additive actuator faults. The goal of this FTC strategy is to reconfigure the closed-loop system by means of a virtual actuator that allows maintaining the use of the nominal controller without having to retune it. In order to achieve this goal, we define the structure of a dynamical virtual actuator for descriptor systems including multiplicative and additive actuator faults. In order to use a state-feedback controller for implementing the FTC strategy, a state observer for descriptor systems is also defined. Under the proposed FTC scheme, the separation principle of the virtual actuator, the nominal closed-loop system and the state observer is established. Therefore, all the gains can be designed separately. Finally, a numerical example is provided to show the effectiveness of the proposed FTC strategy.

Keywords: Fault tolerant control / fault recovery, Autonomous vehicles, Fault detection and isolation
Abstract: In this study, the parameter identification based fault tolerant control of an autonomous underwater vehicle (AUV) dynamics is presented. The Kalman filter (KF) is used in order to identify the control distribution matrix elements, that correspond to the faulty actuator; thus, the control reconfiguration is carried out using this identified control distribution matrix. A linear quadratic regulator (LQR) is designed for the modeled AUV as a controller. A fault in the actuator is identified by the KF and LQR controller is reconfigured for the identified model. In simulations, the linearized model of dynamics of steering subsystem of AUV is considered, and the performance of the proposed actuator failure identification and reconfigurable control techniques are examined.

Keywords: Fault tolerant control / fault recovery, Supervisory control, Autonomous vehicles
Abstract: Real-time tolerance to novel disturbances is often needed to avoid catastrophic outcomes. However, no solutions currently exist: e.g., isolating the underlying fault and implementing an appropriate control may be too time-consuming or require a priori fault information. For such circumstances, we propose a novel approach that uses fast randomized control switching to stabilize a system, without any state estimation. We first show theoretically that we can synthesize a properly tuned controller that uses a randomized approach to guarantee mean-square stability and is guaranteed to converge to an optimal solution. We then empirically demonstrate the efficacy of this approach on trajectory following in a quadcopter UAV.

Keywords: Fault tolerant control / fault recovery, Model-based methods, Discrete event and hybrid systems
Abstract: This paper reports an application of an max-plus algebra fault-tolerant control to a system of cubes' packing. The system constitutes the main part of of the laboratory "Process Automation and Visualization" that is located in Institute of Control and Computation Engineering, University of Zielona Gora. Such systems belong to the class of discrete event systems for which the evolution of the states depends on the occurrence of events. The essential problem concerns a control of independent, parallel conveyor belts transporting cubes to processing units. The contribution of this paper is a mathematical description of a set of multiple conveyor belts. The proposed solution guarantees cube handling by conveyor belts according to a reference sequence. This sequence is provided by the Manufacturing Execution System (MES). In addition, the proposed approach has to deal with both synchronization (in particular the one based on mutual exclusion mode), concurrency issues as well as faults. Subsequently, all necessary constraints are developed and incorporated within the transportation strategy. As a result, the predictive fault-tolerant control strategy is developed. The final part of the paper exhibits an illustrative example, which shows the performance of the proposed approach.

Keywords: Fault tolerant control / fault recovery, Fault detection and isolation
Abstract: This paper considers the active fault tolerant control (FTC) for hexacopter to handle disturbance torque along Z_{E}-axis under a total fault in one rotor. The proposed approach is based on the control system reconfiguration of remaining rotors. As the system is no longer controllable in all six degrees of freedom (DOF), one dimension should be abandoned. The existing approaches usually ignore the dynamic of yaw and consider only the stabilization of position. When the hexacopter is controlled without the consideration of yaw in faulty case, a transient disturbance torque occurs along Z_{E}-axis in the transient phase and causes a large yaw rate. The FTC approach proposed in this paper is able to reduce the yaw rate effectively, and the tracking error of position is bounded. The proposed approach is demonstrated by simulation examples.

Keywords: Fault tolerant control / fault recovery, Aeronautics / aerospace
Abstract: This article presents the design and validation of a fault-tolerant Y* flight controller for the Japanese Aerospace Exploration Agency's (JAXA) Multi-Purpose Aviation Laboratory-alpha (MuPAL-alpha) aircraft. A structured H-infinity controller is synthesized by using non-smooth optimization to address performance and robustness constraints (specifically, uncertain actuator delays and varying airspeed) as well as loss-of-efficiency faults of a specific range with reduced online numerical complexity. The performance of the closed-loop system is validated through hardware-in-the-loop (HIL) simulations using the actual MuPAL-alpha aircraft, and they show that an improved fault tolerant capability is achieved when compared to Y* design robust only against the actuator delays.

Keywords: Fault detection and isolation, Model-based methods
Abstract: This work investigates the issue of interval fault estimation based on a zonotopic technique for discrete-time LPV systems with peak bounded uncertainties. First, taking the faults as an auxiliary state vector, an augmented descriptor system is generated. Then, a robust observer considering L∞ performance is designed for it. The design conditions of the observer are given it terms of LMIs. Next, the bounds of the augmented states estimation are generated based on zonotopes that rigorously consider the unknown disturbance bounds. Finally, simulation results illustrate the performance of the proposed interval fault estimation method.

Keywords: Model-based methods
Abstract: The aim of this paper is to provide a robust observer design using a Linear Matrix Inequality (LMI) technique along with the analysis and comparison of the interval observer and set-membership approaches within the state estimation framework. The propagation of the uncertainty is taken into account considering a zonotopic-set representation. Both approaches are compared and related using Linear Quadratic Regulator (LQR). Finally, a simulation example based on a three-tank system is employed to both illustrate and discuss the effectiveness of the proposed approach.

Keywords: Fault detection and isolation, Model-based methods
Abstract: A new on-line active fault detection method is proposed based on state set-membership ellipsoid in this paper. First, the system states are estimated based on H∞ observer method. To implement on-line fault detection, the moving-window approach is used in this paper Then, an optimal auxiliary input signal is designed to guarantee the separation of state ellipsoids of normal model and state ellipsoids of fault model. Finally, compared with the method of output set-membership ellipsoid, simulation results demonstrate the advantages of the proposed method.

Keywords: Fault detection and isolation, Model-based methods, Automobile
Abstract: The main contribution of this paper is a robust H2 observer design to estimate the actuator degradation under the influence of disturbances and parametric uncertainty. In this design, the observer conception, based on the H2 frequencyshaping filter and theorem of Parseval, allows the estimation of degradation in polynomial form while coping with system uncertainty. Moreover, the usage of the proposed method also relaxes the conditions for unknown input observer (UIO) or the min-max optimization in linear matrix inequality (LMI) for the disturbance attenuation, which facilitates the robust observer design by engineers. Finally, an application to suspension system is presented to illustrate the performance of the proposed method.

Keywords: Fault tolerant control / fault recovery, Mechatronic and robotics, Model-based methods
Abstract: In this paper, the control problem of a service robot under exogenous forces that might be exerted on the system is addressed, considering them as faults. Proposed approach is based on using a fault estimation scheme together with a Robust Unknown-Input Observer (RUIO) that allows to estimate the fault as well as the robot state, along with a state feedback control strategy. After the fault occurrence, from the fault estimation, a feedforward control action is added to the feedback control action to compensate the fault effect. To cope with the robot non-linearity, its model is transformed into a quasi-LPV model. Both the state-feedback and RUIO are designed using an LMI-based approach considering a gain-schedu-ling alike scheme, where global conditions have been considered. To illustrate the proposed fault-tolerant approach, a mobile service robot TIAGo, developed by PAL robotics, is used.

Keywords: Fault tolerant control / fault recovery
Abstract: In this paper, we design a new robust Observer-based Discrete-Time Sliding Mode Control for an uncertain linear discrete-time delay system. The considered system is subject to time-varying states delay, unmatched Unknown and norm-Bounded UBB parametric uncertainties, external disturbances and nonlinearities which are assumed to satisfy a Hölder condition. A new delay-dependent sufficient conditions guaranteeing the robust asymptotic stability of the closed-loop system is derived in terms of Linear Matrix Inequalities. With respect to these conditions, a new discrete-time sliding mode control law is designed based on a novel observer structure and a new sliding surface. A numerical example is presented to illustrate the efficiency of our proposed approach.

Keywords: Fault tolerant control / fault recovery, Building supervision, Model-based methods
Abstract: In this paper a fault diagnosis and fault tolerant control based on model predictive control approach is applied for achieving efficiency in Nearly zero energy buildings. The system, for which the MPC-FDI/FTC controller has been developed, includes a heat pump for temperature regulation and a micro-grid energy system composed of PV panels, micro wind turbine, a battery pack. The objective is to meet the balance between energy production and consumption while minimizing the total cost of the energy drawn from the main grid and maintaining the building temperature within given bounds. Constraints concerning the system operation are also considered. Potential faults that may affect the efficiency of the system have been considered and the necessary actions to deal with them have been proposed.

Keywords: Maintenance policies, Building supervision, Design for reliability and safety
Abstract: Non-residential buildings consume a significant amount of energy which represents a wide part of energy resources. Buildings are equipped with several systems like Lightning, HVAC (Heating, Ventilation and Air conditioning) ... Nevertheless, occupants’ comfort is highly influenced by HVAC reliability. Moreover, HVAC energy consumption forms a significant among of total building consumption. HVAC system often has hidden performance problems leading to energy losses. In order to guarantee an acceptable comfort, energy savings and reduce maintenance costs, a proper maintenance of HVAC system is needed. Nowadays, traditional maintenance strategies are unable to provide an effective solution. HVAC performance problems can be avoided with the help of Predictive Maintenance. This later consist of diagnostic, prognostic and decision-making processes Predictive Maintenance has the ability to anticipate HVAC problems through prognostic. Then, maintenance decision is taken in order to perform the right maintenance and the right time.

Keywords: Fault tolerant control / fault recovery, Building supervision
Abstract: This paper presents an active Fault-Tolerant Control (FTC) scheme for the temperature regulation of buildings subject to actuator faults. An optimal control strategy based on a regularized predictive economic criterion is proposed to determine the remedial actions to be taken when actuator faults are diagnosed. These corrective actions can be used either to generate a less « aggressive » control scenario by minimizing the number of on/off cycles of the faulty actuator or to reduce the number of actuators used at any time. One of the interests of this strategy is to prevent simple faults from developing into serious failures for actuators particularly vulnerable to repeated on/off cycles (in particular heat pump compressors). To account for the « aggressiveness » of a control scenario, the penalty term associated with the regularized economic criterion is defined using the parsimony property of ell_{epsilon}-norm (epsilon small).

Keywords: Fault tolerant control / fault recovery, Fault detection and isolation, Building supervision
Abstract: This paper presents a novel integrated decentralized fault-diagnosis and fault-tolerant approach based on the decentralized economic model predictive control strategy for an HVAC system in a multi-zone building. The Interconnected zones of the building are coupled primarily due to the walls they share, whose direct temperature measurement is either difficult to obtain or is not available. To overcome this problem and estimate the wall temperatures of zones, a full order nonlinear state-observer is designed. Further, using these estimates, we decompose the system into control-oriented thermal models of individual zones. Subsequently, an integrated decentralized fault-diagnosis and fault-tolerant approach based on the decentralized economic model-predictive control is designed which can be easily scaled to buildings with a large number of zones, thereby, making it suitable for real-time implementation. A one-storey building with three zones is used for demonstrating the efficacy of control and monitoring approach.

Keywords: Fault tolerant control / fault recovery, Health monitoring, Power plants / energy transport
Abstract: In this paper, an innovative health-aware control approach is presented for autonomous racing vehicles. Based on a dynamical model of the vehicle, a controller is computed by means of the Linear Matrix Inequality (LMI), where the general objective is to maximize progress on the track subject to win racing and saving energy. The main contribution of the paper consists in preserving the state of charge (SoC) and optimizing a lap time to achieve the best path of a racing vehicle. The control design is divided into two parts, path planner and controller. The first optimization problem is related to the path planner where the objective is to optimize the lap time and maximize the SoC to obtain the best trajectory under the constraints of the circuit. The proposed approach is solved by LMI based Model Predictive Control (MPC) driven from Lyapunov stability. The second part is focused on a controller gain synthesis solved by LMI formulation with integral action for tracking the trajectory obtained by the planner. The proposed approach is evaluated in simulation and results show the effectiveness of the proposed planner for optimizing the lap time and especially for maximizing the SoC of the battery.

Keywords: Mining and mineral, Quality monitoring, Soft computing
Abstract: An enhanced control strategy for a multiple hearth furnace for the purpose of kaolin production is developed and presented in this paper. Mineralogy-driven machine learning algorithms play a key role in the optimization strategy of the furnace. First, the capacity and temperature setpoints for furnace control are determined based on the feed ore mineralogy. Next, the capacity is optimized by combining the prediction of soluble alumina content and mullite content, while maintaining the quality of the product. The stabilizing control level compensates the disturbances with a feedforward control, which uses a spinel phase reaction rate soft sensor, aimed at minimizing the energy use of the furnace. The control concept is successfully tested by simulation using industrial data. Finally, a sampling campaign and software testing of the soft sensors and machine learning algorithms are performed at the industrial site. The results are presented and discussed in the paper.

Keywords: Networked control system, Fault tolerant control / fault recovery
Abstract: This paper deals with networked control systems in which all subsystems have to fulfill a cooperative task. Based on the capabilities of each subsystem, a coordination unit evaluates how much a particular subsystem can contribute to the cooperative task. In case of a fault, the contribution of the faulty subsystem is re-evaluated and adjusted if necessary. As a result, the proposed method ensures the fulfillment of the cooperative task even in case of faults. A numerical example illustrates the functionality for a transportation system.

Keywords: Data driven methods, Chemical processes, Power plants / energy transport
Abstract: This work presents a process data analytics platform built around the concept of industry 4.0. The platform utilizes the state-of-the-art industry internet of things (IIoT) platforms, machine learning (ML) algorithms and big-data software tools. The industrial applicability of the platform was demonstrated by the development of soft sensors for use in a waste-to-energy (WTE) plant. In the case study, the work studied data-driven soft sensors to predict syngas heating value and hot flue gas temperature. From data-driven models, the neural network based nonlinear autoregressive with external input (NARX) model demonstrated better performance in prediction of both syngas heating value and flue gas temperature in a WTE process.

Keywords: Supervisory control, Building supervision, Model-based methods
Abstract: The remote supervision for a class of nonlinear systems in the presence of additive disturbances and measurement noises is considered in this paper. The communication network may introduce time delays while exchanging data among sites connected to the network medium (i.e., the data acquisition site and the remote plant site as shown in Fig. 1. Two different approaches are investigated in this paper: the conventional estimator-based predictor when the uncertainties are supposed to be known and the interval estimation-based predictor evaluating the set of admissible values for the state when the uncertainties are unknown but bounded by known bounds. The state prediction techniques are used to compensate the effect of network-induced delays. Simulation results are introduced to illustrate the efficiency of the proposed techniques.

Keywords: Fault detection and isolation, Discrete event and hybrid systems
Abstract: The complexity of critical systems does not tolerate an unplanned stop of attempted functionalities, which destabilizes their operation and causes material and human damage. This makes failure prognosis a requirement that scientific communities are constantly responding to through developing new techniques that reduce the impact of failures on the system. The control of failure events and the system repair date are very important in the discrete event systems (DES) area to meet this requirement. Therefore, in the first part of this paper we propose a prognostic approach based on extended Petri Nets (EPN), a technique which makes it possible to determine the occurrence of a future failure event tau-time units in advance. For multi-resource prognosis, we will use several clocks.

Keywords: Design for reliability and safety, Mechatronic and robotics, Fault tolerant control / fault recovery
Abstract: This paper deals with the functional performance analysis of various allowed configurations of a redundant mechatronics system. A new indicator called functionability is proposed, which evaluates the performance level of each possible configuration of a redundant system. This analysis is an off-line method, which is further extended for on-line reconfiguration of the system in case of faulty components. The proposed functionability analysis is illustrated through an example of a redundant electrical system. Finally, the analysis is applied to a four wheeled mobile robot called Robucar.

Keywords: Transportation systems, Automobile, Model-based methods
Abstract: A heuristic approach for the detection of the road environment-type (RET) that appears around, and sweeps past an ego-car, while the car is being driven in an urban area, is presented in the paper. The proposed RET detection relies on the traffic signs (TSs) encountered and detected along the route. In an expedient implementation, the TS data is gathered and stored by an on-board TS recognition system. The stored TS data is continually updated, aggregated, and evaluated in a multi-scale manner, i.e., over short, medium and longer patches of the road. Heuristic rules are set up to handle various competition situations between RETs put forward by the evaluation programs responsible for the different scales. These programs make use of the current aggregated TS data, as well as the aggregated TS and RET data gathered earlier during data collection drives. The competition situations between RETs are similar to those that arise in group tournaments between sport teams, or athletes. For this reason, labels, comparison and score-calculation techniques ‒ used by sport competition juries ‒ have been borrowed for the purpose. The RET detection results computed in this manner are compared to those acquired via another TS-based detection approach.

Keywords: Mechatronic and robotics, Model-based methods
Abstract: This paper presents an improved mathematical model of nonlinear actuator, as an example of a three-phase asynchronous motor, being an integral part of automatic mooring winch. For such type of technical objects, a mathematical model of auxiliary parametric sensitivity was developed, which becomes the real unifying factor that make it possible to put the task of analyzing of nonlinear technical systems on the common basis of the general theory of nonlinear differential equations combining into one algorithms the all stages of object's analysis: the calculation of the transitive and steady-state processes, determination of static stability and parametric sensitivity.

Keywords: Power plants / energy transport, Fault detection and isolation, Model-based methods
Abstract: IGBTs, in a Neutral Point Clamped inverter, can be affected by open-circuit or short-circuit faults. Two model-based diagnosis methods are proposed. The Fast Fourier Transform is used in the first one to compute the spectrum of the calculated voltage residuals. By monitoring the magnitude of the fundamental harmonic, failures can be detected and the faulty outer or inner couple of IGBTs can be located. Then, the measured currents are analyzed to isolate the impaired transistor. In the second method, the residuals of the currents are inspected to detect and isolate the fault. The combination of the mean, maximum and minimum values of these residuals yields a unique signature for each impaired IGBT. The two proposed methods are compared. The results are discussed and the best one is recommended for implementation.

Keywords: Fault detection and isolation, Power plants / energy transport, Data driven methods
Abstract: This paper extends the authors’ previous work on a diagnosis algorithm for single open-circuit faults in inverters to the double fault case. The proposed approach is entirely data-driven, using measurements of the stator currents. The first contribution of the present work resides in new choices for the feature variables which allow us to detect double open-circuit faults. These feature data are clustered to different classes corresponding to faulty modes using Auto-adaptive and Dynamical Clustering (AUDyC). These classes are then labelled appropriately which allows us to determine conditions for the fault detection and isolation. Moreover, the values of the algorithm parameters and their influence on the detection time are presented. The proposed improvements are validated on simulation data.

Keywords: Fault-forecasting methods, Health monitoring, Automobile
Abstract: This paper aims to develop an integrated methodology that correlates the slow degradation of the output filter capacitor with the state of a DC-DC Boost converter used in electric vehicles (EVs). Thus, an intelligent Model-Based prognostics technique is applied to predict the Remaining Useful Life (RUL) of this dynamic model with time-variant parameters. In order to capture the smooth degradation as well as reducing the innovation error which might not be detected by the sensors, we proposed an Adaptive Joint Extended Kalman Filter (AJEKF) and an Adaptive Dual Kalman Filter (ADKF) techniques to overcome this issue. The feasibility and effectiveness of the utilized techniques are demonstrated in simulation.

Keywords: Fault-forecasting methods, Health monitoring, Automobile
Abstract: Power electronics devices are widely used especially in critical applications. Their failure affects the reliability and the functionality of the whole system. Understanding their failure mechanisms and degradation for prognostics and health management is becoming more and more interesting in the last few years. Due to environmental operation conditions, these devices are subject to degradation process that leads to complete failure of the system. However, before a catastrophic failure occurs some early signs can be observed to indicate the level of degradation known as fault precursors. For the Metal Oxide Semiconductor Field Effect Transistor (MOSFET), the ON-resistance (Rds,(ON)) is considered as the main precursor parameter for die-attached failure. Hence, the latter occurs when the MOSFET faces thermal over-stress which can be simulated by accelerated aging experiments. In this article, the degradation models of the MOSFET component of a DC-DC converter are presented based on experimental data, and the deterioration is estimated by the state-parameter joint estimation using Extended Kalman Filter (EKF).

Keywords: Networked control system
Abstract: After successful cyber attacks, it has become growingly important to strengthen the cyber physical systems security. The need of security become a priority for people to protect their information from adversary’s hands. A replay attack represents one of the famous network attack and their existence can make a huge damage. In this paper we will study the stabilization problem of cyber physical system (CPS) exposed to a replay attack. The study will take into consideration the effect of packet losses. Linear matrix inequality (LMI) formulation is used to give a sufficient condition for stabilization of the system. Finally the theory is demonstrated by a numerical example.

Keywords: Fault detection and isolation
Abstract: This paper presents a watermarking signal injection method that compensates its effect in the loop, avoiding thus the signal reinjection. Similar to a virtual actuator scheme, the proposed methodology masks the presence of the authentication signal to the system controller, that do not need to be retuned as it remains immunized. Furthermore, a set-based analysis concerning the effect that the performance loss imposed by a watermarking signal has in the detectability of a replay attack is performed for the stationary, assuming that a standard state observer is used in order to monitor the plant. Finally, a numerical application example is used to illustrate the proposed approach.

Keywords: Networked control system, Fault detection and isolation, Model-based methods
Abstract: This paper addresses the design of a secure interval state estimator for linear continuous-time systems in the bounded error context with discrete-time measurements subject to external attacks. The attacker capabilities are assumed limited in the sense that only a subset of all the sensors can be attacked although this subset is unknown. For a given upper bound on the number of attacked sensors, we propose a new selection strategy, which is able to achieve resiliency to attacks, using the width of estimated intervals. The interval observer is modelled as an impulsive system, where impulsive corrections are made periodically using measurement. The nonnegativity of the observation error between two successive measurements is preserved by applying the internal positivity of the system. The theoretical result is supported by numerical simulations.

Keywords: Model-based methods, Fault detection and isolation
Abstract: In this paper, we consider a family of discrete-time systems with input and output. A new technique, which is reminiscent of a construction of finite-time observers for continuous-time systems, is proposed for discrete-time case. The systems are affected by additive disturbances and noises. Exact estimation or approximate estimation can be achieved depending on the absence or the presence of unknown but bounded uncertainties, respectively.

Keywords: Fault detection and isolation, Model-based methods, Decision making
Abstract: Aiming to enhance the performance of state estimation algorithms in the bounded error context, we propose in this work two kinds of combination of two types of setmembership estimators. The first estimator use interval analysis to over-approximate efficiently the estimation error, whereas the second estimator is based on set invariance principle coupled with ellipsoidal computation. Two hybrid (interval-ellipsoidal) structures are introduced in this paper. The first one has the advantage to need less online computation; while the second one due to its data fusion policy provides accurate state enclosures over all the working period. Illustrative numerical examples are presented to show the performance of the proposed approaches and to support the theoretical results.