Search

Supervisor localization can be applied to distribute a monolithic supervisor into local supervisors. Performing supervisor localization can be computationally costly. In this work, we consider systems that evolve over time. We study how to reuse the results from a previous supervisor localization, to more efficiently compute local supervisors when the system is adapted. We call this approach transformational supervisor localization, and present algorithms for the procedure. The efficiency of the procedure is experimentally evaluated., Comment: Accepted for IEEE Control Systems Letters (L-CSS) (2023)

In this paper, we investigate the enforcement of opacity via supervisory control in the context of discrete-event systems. A system is said to be opaque if the intruder, which is modeled as a passive observer, can never infer confidentially that the system is at a secret state. The design objective is to synthesize a supervisor such that the closed-loop system is opaque even when the control policy is publicly known. In this paper, we propose a new approach for enforcing opacity using non-deterministic supervisors. A non-deterministic supervisor is a decision mechanism that provides a set of control decisions at each instant, and randomly picks a specific control decision from the decision set to actually control the plant. Compared with the standard deterministic control mechanism, such a non-deterministic control mechanism can enhance the plausible deniability of the controlled system as the online control decision is a random realization and cannot be implicitly inferred from the control policy. We provide a sound and complete algorithm for synthesizing a non-deterministic opacity-enforcing supervisor. Furthermore, we show that non-deterministic supervisors

As one of adaptive optimal controls, the Q-learning based supervisory control for hybrid electric vehicle (HEV) energy management is rarely studied for its adaptability. In real-world driving scenarios, conditions such as vehicle loads, road conditions and traffic conditions may vary. If these changes occur and the vehicle supervisory control does not adapt to it, the resulting fuel economy may not be optimal. To our best knowledge, for the first time, the study investigates the adaptability of Q-learning based supervisory control for HEVs. A comprehensive analysis is presented for the adaptability interpretation with three varying factors: driving cycle, vehicle load condition, and road grade. A parallel HEV architecture is considered and Q-learning is used as the reinforcement learning algorithm to control the torque split between the engine and the electric motor. Model Predictive Control, Equivalent consumption minimization strategy and thermostatic control strategy are implemented for comparison. The Q-learning based supervisory control shows strong adaptability under different conditions, and it leads the fuel economy among four supervisory controls in all three varying conditions.

In this paper we study a security problem of protecting secrets in discrete-event systems modeled by deterministic finite automata. In the system some states are defined as secrets, each of which is associated with a security level. The problem is to design an event-protecting policy such that any event sequence from the initial state that reaches a secret state contains a number of protected events no less than the required level of security. To solve this secret securing problem, we first develop a layered structure called the security automaton. Then we show that the problem is transformed to a supervisory control problem in the security automaton. We consider two criteria of optimality on protecting policies: (1) disruptiveness, i.e., protecting policies with a minimum degree of disturbance to legal users' normal operations; (2) cost, i.e., protecting policies with a minimal cost. For the optimality on disruptiveness, we prove that a minimally disruptive protecting policy is obtained by using the classical supervisory control theory in the security automaton. For the optimality on cost, we develop a method to obtain a protecting policy with minimal cost by finding a min-cut in the security automaton.

We study the security of control systems in the context of the supervisory control layer of stochastic discrete-event systems. Control systems heavily rely on correct communication between the plant and the controller. In this work, we consider that such communication is partially compromised by a malicious attacker. The attacker has the ability to modify a subset of the sensor readings and mislead the supervisor, with the goal of inducing the system into an unsafe state. We consider this problem from the attacker's viewpoint and investigate the synthesis of an attack strategy for systems modeled as probabilistic automata. Specifically, we investigate the synthesis of attack functions constrained by multiple objectives. First, we study the problem of synthesizing attack functions with the maximum likelihood of successfully reaching an unsafe state. Second, we consider the problem of synthesizing attack functions that have the maximum likelihood of successfully reaching an unsafe state while minimizing a cost function. Our solution methodology is based on mapping these problems to optimal control problems for Markov decision processes

This work investigates quantitative supervisory control with local mean payoff objectives on discrete event systems modeled as weighted automata. Weight flows are generated as new events occur, which are required to satisfy some quantitative conditions. We focus on the mean weight over a finite number of events, which serves as a measure to reflect the stability or robustness of the weight flows. The range of events to evaluate the mean payoff is termed the window, which slides as new events occur. Qualitative requirements such as safety and liveness are also necessary along with quantitative requirements. Supervisory control is employed to manipulate the operation of the system so that the requirements are satisfied. We consider two different scenarios based on whether the window size is fixed or not. Correspondingly, we formulate two supervisory control problems, both of which are solved sequentially by first tackling the qualitative issues and then the quantitative ones. The automaton model of the system is then transformed to a two-player game between the supervisor and the environment, where safety and liveness are enforced. Based on the intermediate results, several quantitative objectives are then defined to formulate two games, which correspond to the two proposed supervisory control problems. Finally we synthesize provably correct supervisors by solving the games and completely resolve both problems.

We investigate state estimation and safe controller synthesis for networked discrete-event systems (DES), where supervisors send control decisions to plants via communication channels subject to communication delays. Previous works on state estimation of networked DES are based on the open-loop system without utilizing the knowledge of the control policy. In this paper, we propose a new approach for online estimation and control of networked DES with control delays. We first propose a new state estimation algorithm for the closed-loop system utilizing the information of control decision history. The proposed state estimation algorithm can be implemented recursively upon the occurrence of each new observable event. Then we investigate how to predict the effect of control delays in order to calculate a control decision online at each instant. We show that the proposed online supervisor can be updated effectively and the resulting closed-loop behavior is safe. Furthermore, we compare the proposed online supervisor with the predictive supervisor proposed in the literature and show that our proposed online supervisor is more permissive than predictive supervisor in

In cyber-physical systems (CPSs), it is of great importance to handle network attack issues. In this article, we consider the supervisory control layer of CPSs, focusing on closed-loop control systems vulnerable to sensor-reading modification attacks (SM-attacks), which may disguise the occurrence of an event as a different event by modifying appropriately sensor readings in sensor communication channels. In particular, we consider the plant modeled as a bounded Petri net and the control specification consisting in liveness enforcing. Based on repeatedly computing a more restrictive liveness-enforcing supervisor under no attack and constructing a so-called basic supervisor, a method that synthesizes a liveness-enforcing supervisor tolerant to an SM-attack is proposed.

This work addresses the robust control problem of discrete event systems assuming that replacement attacks may occur, thus making it appear that an event that has occurred looks like another event. In particular, we assume that this is done by tampering with the sensor-readings in the sensor communication channel. Specifically, we use Petri nets as the reference formalism to model the plant and assume a control specification in terms of a Generalized Mutual Exclusion Constraint. We propose three different methods to derive a control policy that is robust to the possible replacement attacks. The first two methods lead to an optimal (i.e., maximally permissive) policy but are computationally inefficient when applied to large-size systems. On the contrary, the third method computes a policy more efficiently, and reveals more easily implementable in practice. However, this is done at the expense of optimality.

This letter handles a similarity control problem for nondeterministic discrete event systems. In this problem, the system and the modular specification are modeled as nondeterministic automata, where the modular specification is the synchronous composition of a certain set of elementary specifications. The similarity control problem requires us to synthesize a supervisor such that the supervised system is simulated by the modular specification. We solve this problem by applying a modular control architecture where, for each elementary specification, an elementary supervisor that enforces similarity is synthesized. We show that such elementary supervisors together achieve maximal permissiveness with respect to the modular specification if each of them is maximally permissive with respect to the corresponding elementary specification.

In industry the performance has to be optimized to allow its competitiveness. The goal then is to use the resources at their maximal capacity, reduce the production time and be flexible to adapt to the customers requests. The closed-loop behavior of a system under the Supervisory Control Theory (SCT) guarantees nonblockingness and safety requirements and, thus, it can be used as the search universe for a planning problem. SCT suffers with the "curse of dimensionality" when systems become bigger and more complex. This paper presents a set of sufficient conditions that allow to work with abstractions of the closed-loop behavior (supremal controllable sublanguage), instead of the closed-loop behavior itself, as the search universe to solve a planning problem. Such abstraction is the natural projection of the supervisor into the set of controllable events satisfying the observer property. This process leads to a reduction of the search space.

This article provides a detailed description of a supervisory control problem for a heavy-duty vehicle equipped with a diesel engine and an exhaust aftertreatment system (EATS). The control problem considers the coupling between the engine and EATS and assumes the availability of predictive driving information for a long horizon. The article also proposes a computationally efficient supervisory control algorithm, which is implemented in an existing engine electronic control unit. The algorithm is based on distributed economic nonlinear MPC. It minimizes total fuel consumption (by the diesel engine and EATS) and fulfils Euro VI real-driving and world harmonized transient cycle based NOx emission legislation for the heavy-duty diesel vehicle. The performance of the algorithm is evaluated using a high-fidelity simulation platform, which includes a validated GTPOWER model of a 13 L turbo compound diesel engine and the first-principles model of an EATS. Simulation results show that the supervisory controller has 1.7% less total fuel consumption, and 88.4% less NH3 slip, compared to a baseline controller, to achieve the same real-driving NOx emission.

In our previous paper, we have considered the supervisory control of the probabilistic discrete event systems (PDESs) under the centralized framework. In this paper, we continue to investigate the decentralized probabilistic supervisory control issue of PDESs. The intersection-based control architecture is adopted, and the notion of the probabilistic intersection-based co-observability and its verification algorithm are presented. It is shown that this observability and the probabilistic controllability act as the necessary and sufficient conditions for the existence of the intersection-based decentralized probabilistic supervisors. Moreover, a working example is provided to illustrate the proposed concepts, the obtained results, and the potential application of the proposed control approaches.

Real-time control applications are highly parallelizable and can be used to effectively harness the capacity of a given computing platform when appropriately scheduled. Given a multicore platform for executing a set of parallelizable applications, it is necessary to ensure proper functioning of the system even in the presence of transient processor faults. However, most existing scheduling approaches for parallel applications have been heuristic schemes which are often based only on the satisfaction of a set of sufficiency conditions and cannot take into consideration of all necessary schedulability requirements. Consequently, such schemes lead to suboptimal usage of resources resulting in higher design costs. Formal model-based safe design methodologies such as supervisory control are often desirable in the design of correct-by-construction fault-tolerant schedulers. This article proposes a supervisory control-based fault-tolerant scheduler synthesis scheme for real-time tasks modeled as precedence-constrained task graphs , executing on multicores. Further, we devise search strategies to obtain schedules that 1) maximize fault-tolerance and 2) minimize peak-power (MPP) dissipation. Conducted experiments using real-world benchmarks reveal the efficacy of our scheme.

This article investigates the enforcement of generalized mutual exclusion constraints (GMECs) and deadlock-freeness on a time Petri net (TPN) system with uncontrollable transitions, motivated by the fact that the existing methods enforcing GMECs may degrade the performance of a closed-loop system and lead to deadlock states. A supervisor enforcing a set of GMECs and deadlock-freeness on an underlying untimed Petri net system is assumed to be available. By exploiting timing information and mathematical programming, a control function is designed to restrict the firing intervals of transitions such that a TPN system can avoid entering forbidden states. The key idea behind the proposed approach is the online computation of a graph, called reduced modified state class graph (RMSCG), that is an extension of the partial modified state class graph recently introduced by the authors. Based on the RMSCG, an online control synthesis procedure is developed, which can enforce the originally given GMECs and deadlock-freeness in a maximally permissive way.

In microgrids with droop-controlled inverters, the supervisory controller must the tune the droop gains in real time to maintain stability as the system topology or the number of inverters change. Stability could be compromised due to delays in the dispatch of supervisory control inputs or due to communication link failures. This article proposes a filter deballasting control (FDC) method for maintaining stability in the interim period until the supervisory inputs become available. FDC employs a quasi-open-loop control structure to leverage the inductance of the LC filter as a stability ballast, which under normal operation does not contribute to the dynamic behavior of the system. To further improve the stability, a lead-lag compensator is implemented to compensate the lagging dynamics of the deballasted system. The proposed method does not affect the steady-state active and reactive power share of the inverters, enabling a smooth transition between the conventional and FDC operating modes. The effectiveness of the proposed control strategy is verified through theoretical and experimental means.

This paper studies the linear-time temporal logic (LTL) control of a class of fair discrete-event systems (DESs). It is motivated by the curious extent in which the use of LTL can be strengthened and differentiated in control theory development. Over a fair DES model, a marker-progressive supervisory control problem is formulated in LTL. The problem formulation admits a more flexible specification of multiple markers to distinguish different DES tasks, and seeks to find a supervisor—a passive control function by convention for specified temporal safety—such that a fair DES under its control is guaranteed to make constant progress to these markers. The problem is studied in terms of DES marker-controllability—a new controllability concept formulation of temporal safety for constant marker progress. This new formulation sheds light on how event fairness in DESs coachieves such marker progress with supervision that exists. It is shown that a solution supervisor may be found by canonical LTL verification. Three examples are provided for illustration.

A supervisory control for neutron scattering superconducting magnet is proposed in this paper, which adopts hierarchical control strategy. In the field layer, programmable logic controller (PLC) based on monitoring control and data acquisition system is used for automatic control of various working conditions. PLC and upper monitoring computer are independent of each other in the equipment control layer, and independent operation is not dependent on the upper computer, which ensures the independence and safety of the process. The running parameters of superconducting magnet can be detected in real-time with the coverage of local area network (LAN), and the significant equipment of neutron scattering superconducting magnet can be controlled by using real-time control method in the mobile environment. In case of abnormal situation, the backfeed control will be used to realize remote alarm and other functions, which ensures the normal running of neutron scattering superconducting magnet safely and reliably.

The energy management of the fast-charging stations (FCSs) in power systems, while maintaining the power quality and the grid codes, can be particularly challenging, given the highly dynamic load curve, the heterogeneous PEV charging, and the inconvenience to the FCS customers. In this article, a consecutive horizon-based energy management (CHEM) process is proposed for the optimal techno-economic operation of the FCSs integrated with battery energy storage systems (ESS). The proposed CHEM process employs a central supervisory control and a resilient distributed control, for the energy management of FCSs and the active and reactive power control to maintain the grid codes. As such, the grid operator can realize and direct the economic and technical operation of the FCSs with low computational burden, without the need for real-time collaboration with the FCSs. Moreover, the FCSs can autonomously cope with the volatility of PEV charging with the use of the ESS. Furthermore, a novel Volt/VAr control is proposed to confine the unpredicted voltage deviations, including voltage sags, within an acceptable limit. The heterogeneous PEV charging is modeled using the queuing theory, while the inherent characteristics of the lithium-ion batteries are considered. Through multiple case studies, it is shown that the developed CHEM process can provide lower costs and a leveled active power profile for the FCSs, while ensuring that the voltage deviations are well confined and the convenience of PEV owners can still be met.

We consider feedback control systems where sensor readings may be compromised by a malicious attacker intending on causing damage to the system. We study this problem at the supervisory layer of the control system, using discrete event systems techniques. We assume that the attacker can edit the outputs from the sensors of the system before they reach the supervisory controller. In this context, we formulate the problem of synthesizing a supervisor that is robust against the class of edit attacks on the sensor readings and present a solution methodology for this problem. This methodology blends techniques from games on automata with imperfect information with results from supervisory control theory of partially-observed discrete event systems. Necessary and sufficient conditions are provided for the investigated problem., Comment: Full version of preprint accepted at IEEE Transactions on Automatic Control

Automation regroups a variety of advanced tools meant to improve performance and decrease human workload. This article was designed to investigate how different automation solutions, engaging different human–machine cooperation modes, interact with expertise. Aircraft pilots (i.e., experts) and nonpilots (i.e., novices) were presented with a set of simplified flight piloting tasks monitored simultaneously using the Open MATB (Open Multiattribute Task Battery) in four different automation conditions (manual, assisted, cooperative, and supervisory control). Experts’ performances at the Open MATB were higher than those of novices. Experts also exhibited a lower level of workload. Apart from function delegation, where no human performance is required, automation solutions benefit experts and novices similarly. Participants were then asked to choose repeatedly between the four automation conditions. Experts and novices exhibited different strategies. Novices spread their choices over the different automation conditions, whereas experts clearly favor cooperative control, the automation solution that enhanced their expertise.

Electric transport systems and renewable energy sources (RESs) have recently attracted significant interest, because of the limitations and drawbacks of fossil fuels and the growing demand for utilization of clean energy. Suitable paradigms are needed to manage the atypical and heterogeneous load of electric vehicles (EVs), the intermittent nature of RESs, and the changeability of electrical loads in power networks. The technical and commercial operation of an integrated system comprising an electric transportation system with a battery-powered bus [electric bus (eBus)] charging station and an EV parking lot, integrated with solar photovoltaic (PV) generation, and combined with a battery storage system (BSS), as a virtual energy hub (VEH) is proposed. Moreover, a cooperative decision-making (CDM) strategy is proposed for the VEH, where the active and reactive power flows and the economic operation of the VEH are scheduled using a novel three-stage cooperative control system. A supervisory control system is developed using agents to realize the scheduled CDM of the VEH and to assign the control parameters for the provision of ancillary services. The VEH agent can operate autonomously during the real-time operation to cope with the volatility of the power system based on the optimal predetermined set points from the grid agent. The effectiveness of the CDM methodology is studied and evaluated through different realistic scenarios. The results show that the proposed CDM can organize the effective energy management (EM) with a minimum operational cost to the VEH. Furthermore, because of the designed supervisory control and the provision of ancillary services using the VEH, the grid requirements can be effectively managed.

This article presents a novel framework for the modeling and scheduling of real-time systems (RTS) processing both sporadic and (multiperiod) periodic tasks, based on nonblocking supervisory control of state-tree structures. Instead of assigning static priorities, a real-time scheduling mechanism, namely priority-free conditionally preemptive scheduling, is used to describe the preemption relation among the tasks processed in an RTS. As a dynamic priority scheduling mechanism, partially preemptive or nonpreemptive earliest-deadline first scheduling is also addressed in this article. Finally, the scheduling strategies are illustrated by real-world examples.

This paper presents and evaluates a discrete-event control-based Supervisory Controller (SC) for a grid-connected, battery-enabled, DC Fast-Charging (DCFC) station for Electric Vehicles’ (EVs). The SC enhances operational resiliency of the DCFC station and its host grid by decoupling the dynamics and limiting the steady-state power between the station and its host grid. The charging station includes multiple DCFC units, a Battery Energy Storage System (BESS), a solar-PV unit, and a grid-interface Voltage-Sourced Converter (VSC). The SC (i) limits the station-imported power from the grid based on the grid-imposed requirements, (ii) dynamically dispatches power flow among the grid, PV unit, DCFC units, and BESS to meet charging needs of EVs and when applicable charges BESS or exports power to the grid, (iii) protects the BESS by enforcing its state of charge upper and lower limits and the maximum charge/discharge rates, and (iv) enables the VSC to provide grid ancillary services, if required. Simulation results in PLECS software platform are given to assess the dynamic performance of the proposed SC in response to various discrete events/transitions. The SC algorithm is implemented on a Programmable Logic Controller (PLC) and verified based on real-time Hardware-In-the-Loop approach using an OPAL-RT platform.

A maximally permissive (or optimal) supervisory control of an automated manufacturing system (AMS) modeled by Petri nets (PNs) can be usually implemented by imposing constraints in the form of a set of linear inequalities. To find such a set of linear constraints, in the existing work, an integer linear programming (ILP) problem is generally formulated and solved for some dominant markings obtained by reachability analysis and vector covering theory, which is computationally inefficient due to the combinatorial nature of solving ILPs to decide the coefficients of optimal linear constraints. This paper addresses the deadlock prevention problem for AMSs by developing efficient methods to reduce the computational overhead through the establishment of conditions on deciding the maximal permissiveness of linear constraints imposed on a system. By taking the advantage of structural properties of a PN model under consideration, we identify the most part of minimal covered illegal markings that can be optimally controlled via policies obtained by specific linear inequality constraints. Also, algorithms are developed to implement the proposed policies. The proposed approach can verify the optimality of a linear constraint efficiently without solving ILP problems. A linear programming method is developed to deal with the markings that cannot be processed by the proposed structural analysis. It is shown that for the considered class of PNs, called system of simple sequential processes with resources (S3PR), no mixed ILP problem needs to solve and the computational burden is dramatically reduced. Two examples are employed to demonstrate the efficiency of the developed method.

This paper presents predictive energy management of hybrid electric vehicles (HEVs) via computationally efficient multi-layer control. First, we formulate an optimization problem by considering driveability and a penalty for using service brakes in the objective function to optimize gear, engine on/off, engine clutch state, and power-split decisions subject to constraints on the battery state of charge (SOC) and charge sustenance. Then, we split it into two control layers, including a supervisory control in a higher layer and a local power-split control in a lower layer. In the supervisory layer, a gear and powertrain mode manager (PM) is designed, and optimal gear, engine on/off and clutch states are obtained by using a combination of dynamic programming (DP) and Pontryagin's minimum principle (PMP). Moreover, a real-time iteration Secant method is proposed to calculate optimal battery costate such that the constraint on charge sustenance is satisfied. In the local controller layer, a linear quadratic tracking method (LQT) is used to optimally split power between the engine and the electric machine and keep battery SOC within its bounds.

In practical systems, due to reasons such as sensor limitations, sensor faults, and packet losses in networks, the observation of events becomes nondeterministic. In this article, we extend strong detectability and weak detectability to the case of nondeterministic observations and denote them as A-( $k_{1}$ , $k_{2}$ )-detectability and O-( $k_{1}$ , $k_{2}$ )-detectability, respectively. A-( $k_{1}$ , $k_{2}$ )-detectability says that, for any string, we can distinguish state pairs in the specification for all possible observations of the string. O-( $k_{1}$ , $k_{2}$ )-detectability says that, for at least one string, we can distinguish state pairs in the specification for all possible observations of the string. For A-( $k_{1}$ , $k_{2}$ )-detectability, we construct a transformed automaton and then translate the A-( $k_{1}$ , $k_{2}$ )-detectability problem into the traditional detectability problem that has been solved. We show that A-( $k_{1}$ , $k_{2}$ )-detectability can be used to solve the deterministic supervisory control problem. For O-( $k_{1}$ , $k_{2}$ )-detectability, we construct an augmented automaton that includes all the information of the given automaton and its state estimates. Based on the augmented automaton, we propose a depth-first search (DFS)-based algorithm to check O-( $k_{1}$ , $k_{2}$ )-detectability. Note to Practitioners —Nowadays, practical engineering systems become more and more complex. In these systems, the observation of events often becomes nondeterministic due to reasons such as sensor limitations, sensor faults, and packet losses in networks. Consider a mobile robot as an example. The availability of a sensor output may depend on the current location of the mobile robot. If the mobile robot is in an area where the wireless network is unreliable, the sensor output may not be received by the supervisor. In this article, we investigate the state estimation problem for practical engineering systems under nondeterministic observations within a discrete-event system framework. The results in this article provide not only insights for engineers in the automatic control field to understand nondeterministic observations in practical systems but also the methodology to estimate the current discrete state that is always an important issue. Therefore, we believe that the engineers in the automatic control field should be interested in this article and can benefit from it.

In this article, we consider the problem of synthesizing a nondeterministic supervisor for the plant and the specification modeled as nondeterministic automata. A similarity control problem requires us to synthesize a supervisor such that the supervised plant is simulated by the specification. Ideally, the synthesized supervisor should be as permissive as possible. We develop a method for synthesizing a maximally permissive supervisor that solves the similarity control problem.

Due to the complexity of a multimode plug-in hybrid electric vehicle (PHEV) powertrain, the energy management strategy of the said powertrain is a prime candidate for the application of optimal control methods. This article presents a predictive control strategy for optimal mode selection and powertrain control for a multimode PHEV capable of real-time control. This method utilizes predictions of future vehicle behavior in order to plan an optimal path of vehicle powertrain modes that minimizes energy consumption. This article also presents the integration of the developed optimal mode control strategy with an optimal powersplit strategy using nonlinear model predictive control to create a real-time integrated predictive powertrain controller (IPPC) responsible for all aspects of multimode PHEV powertrain supervisory control. The IPPC provides a real-time optimal solution to address the major challenge of a multimode HEV powertrain control: an integrated discrete and continuous optimization. Testing in simulation has shown the IPPC to be capable of reducing PHEV energy consumption by 4%–10% across real-world and standard drive cycles. In addition, the presented IPPC was deployed onto a rapid prototyping embedded controller where on-road, real-time testing has shown the IPPC to be capable of providing an energy reduction of 5%, thus confirming the energy savings observed in simulation.

In this article, we investigate security and privacy issues in networked supervisory control systems over multiple channel networks. We consider a networked discrete-event system controlled by a supervisor that receives information from sensors and sends control decisions to actuators via observation channels and control channels, respectively. The security problem is studied for the scenario where some of the communication channels are insecure in the sense there exists a passive intruder (eavesdropper) that can access the information-flow in those insecure communication channels. We adopt the concept of opacity, an information flow security property, to characterize the security status of the supervisory control system. Specifically, we assume that the system has a secret and the system is said to be opaque if the intruder can never determine the secret of the system unambiguously based on the information-flow in the insecure channels. We consider the notions of current-state opacity, $K$ -step opacity, and infinite-step opacity in the networked control setting. New network observers are proposed to estimate the state of the system with two-side incomparable channel information. We show that the opacity verification problems for the networked setting can be effectively solved using the proposed network observers.

The dual active bridge (DAB) is one of the useful interfaces for the bidirectional power exchange between two buses with galvanic isolation. In this article, a design framework is given in the analog domain that completely fulfills the operational features of a DAB under phase angle modulation. This is a cost-effective solution with configurable control parameters and easy to implement using well-established semiconductor integrated circuit (IC). Moreover, it provides an efficient way to study the behavior of power converters. A closed-loop control for bidirectional power exchange is established. For that, a control-to-output current transfer function is derived analytically considering the dc and first-order harmonic terms and verified experimentally. A small scale laboratory prototype of DAB is set up for the validation of the concept. It is shown that the design architecture has the facility for supervisory control that is useful for battery charging/discharging, microgrid application, etc.

Reinforcement learning (RL) is gradually being implemented in the hybrid electric vehicle (HEV) supervisory control. Even though RL exhibits significant fuel consumption saving, the long learning time makes it hardly applicable in real-world vehicles. This study aims to reduce the learning iterations of Q-learning in HEV application utilizing warm-start methods. Different from previous studies, which initiated Q-learning with zero or random Q values, this study initiates the Q-learning with different supervisory controls, and the detailed analysis is given. The results show that the proposed warm-start Q-learning requires 68.8% fewer iterations than cold-start Q-learning and improves 10%–16% MPG compared with equivalent consumption minimization strategy control. The results of this study can be used to facilitate the deployment of RL in vehicle applications.

In low-voltage dc distribution networks (DDNs), the load terminals typically employ voltage-source inverters to provide ac supply to the consumers. These load terminals can potentially experience poor voltage regulation especially in the presence of intermittent photovoltaic sources and uncoordinated charging patterns of plugged-in electric vehicles. Variations in power absorption and generation at different points can potentially create significant voltage drop/rise at the load terminals and may reduce the power quality at the ac side of the loads. In order to mitigate this issue, this article suggests a real-time coordinated supervisory control strategy that estimates the present state of the network and generates a set of references for the controllable sources such that the load voltage deviations are minimized. This article also proposes a power-electronic-converter-based soft open point (SOP) that provides a seamless connection of the remote end and gives an additional degree of freedom to control the circulating current among the feeders. Detailed description and control strategy of the SOP is presented. Steady-state analysis carried out on a three-feeder DDN is presented with and without the proposed SOP. Real-time simulation with an Ethernet-based transmission control protocol/Internet protocol communication system along with experimental results is presented to validate the implementation and effectiveness of the control strategy.

A novel framework is introduced for the supervisory control (SC) of timed discrete event systems, based on Time Petri nets. The method encompasses both logical (markings to reach or avoid) and temporal specifications (arrival and departure times in specific markings). It relies on the construction of a partial forward reachability graph, of the Modified State Class Graph type, and the formulation of integer linear programming problems to establish suitable firing time intervals (FTIs) for the controllable transitions. The SC algorithm provides for each enabled controllable transition the largest FTI that guarantees that the specifications are met, irrespectively of the firing times of the uncontrollable transitions.

In this study, a systematic solution for mobile manipulation with the application of autonomous trolley collection in the airport is presented. The developed system features a novel integrated hardware and software system, incorporating a specially designed manipulator capable of tightly catching the trolley. Particularly, we innovatively present a supervisory control guided coarse-to-fine visual trolley catching strategy in large-scale airport environments. At the coarsest level, an image-based visual servoing method is utilized to approach the trolley from far to near, where a learning-based object detection method is adopted to identify trolleys in the environment in real-time. At the finest level, we present an accurate point cloud registration based object pose estimation method to promote the object manipulation with a position-based visual servoing controller. The tightly integrated system and the subsystems are evaluated through conducting experiments in the trolley collection task and comparing the key module with the state of the art. Extensive experiments demonstrate that the proposed framework can find and fetch the trolley robustly. Notably, the proposed object detection and pose estimation methods exhibit higher accuracy, providing credible perceptual approaches for mobile manipulation. Moreover, the proposed framework can be easily extended to other mobile manipulation tasks for object search and retrieval.

In this article, we study the supervisory control problem of networked discrete event systems with timing structure assuming bounded communication delays and intermittent loss of observations. The communication between the plant and the supervisor is through a network that can have several channels, and so communication delays may change the order of event observations by the supervisor. We assume a priori knowledge of the minimal activation time of the plant transitions and the maximum communication delays, and propose an equivalent untimed model that takes into account all possible observation delays and also loss of observations. Based on this model, we formulate a networked supervisory control problem and show that it can be converted into an equivalent range control problem. We also address implementation, and, to this end, we present a representation for networked supervisors that is able to deal with possible loss of synchronization between event occurrence and its observation.

The supervisory control problem for discrete event system (DES) under control involves identifying the supervisor, if one exists, which, when synchronously composed with the DES, results in a system that conforms to the control specification. In this context, we consider a non-deterministic DES under complete observation and control specification expressed in action-based propositional $\mu$ -calculus. The key to our solution is the process of quotienting the control specification against the plan resulting in a new $\mu$ -calculus formula such that a model for the formula is the supervisor. Thus the task of control synthesis is reduced a problem of $\mu$ -calculus satisfiability. In contrast to the existing $\mu$ -calculus quotienting-based techniques that are developed in deterministic setting, our quotienting rules can handle nondeterminism in the plant models. Another distinguishing feature of our technique is that while existing techniques use a separate $\mu$ -calculus formula to describe the controllability constraint (that uncontrollable events of plants are never disabled by a supervisor), we absorb this constraint as part of quotienting which allows us to directly capture more general state-dependent controllability constraints. Finally, we develop a tableau-based technique for verifying satisfiability of quotiented formula and model generation. The runtime for the technique is exponential in terms of the size of the plan and the control specification. A better complexity result that is polynomial to plant size and exponential to specification size is obtained when the controllability property is state-independent. A prototype implementation in a tabled logic programming language as well as some experimental results are presented.

In this paper, a deadlock prevention policy for robotic manufacturing cells with uncontrollable and unobservable events is proposed based on a Petri net formalism. First, a Petri net for the deadlock control of such systems is defined. Its admissible markings and first-met inadmissible markings (FIMs) are introduced. Next, place invariants are designed via an integer linear program (ILP) to survive all admissible markings and prohibit all FIMs, keeping the underlying system from reaching deadlocks, livelocks, bad markings, and the markings that may evolve into them by firing uncontrollable transitions. ILP also ensures that the obtained deadlock-free supervisor does not observe any unobservable transition. In addition, the supervisor is guaranteed to be admissible and structurally minimal in terms of both control places and added arcs. The condition under which the supervisor is maximally permissive in behavior is given. Finally, experimental results with the proposed method and existing ones are given to show its effectiveness.

This article presents a new droop-based supervisory-level controller for nanogrids (NGs) in the islanded mode of operation. The proposed control system is able to significantly improve the dynamics of power electronic interfaces in NGs. Unlike the existing techniques, the proposed controller takes into account the input-side (i.e., source-side) and the output-side (i.e., load-side) parameters as well as the dc-bus voltage dynamics. Theoretical analysis, simulation results, and experimental results confirm the feasibility of the proposed control system and demonstrate its superior performance.

In this article, we propose a two-layer-structure called hierarchical basis reachability graph that is useful for marking estimation in Petri nets that contain unobservable transitions. On the basis of a hierarchical partition of the set of observable transitions, a hierarchical basis reachability graph consists of a primary and a secondary structures. Simulation shows that the time-consumption to compute a hierarchical basis reachability graph is in general much less than that of computing the corresponding basis reachability graph. A marking estimation algorithm is proposed by the structural analysis of the corresponding hierarchical basis reachability graph.

The hybrid ac–dc microgrid, consisting of ac subgrid and dc subgrid, is a promising structure to interconnect ac and dc systems to provide high quality and reliable electric energy. However, in single-phase microgrid or unbalanced three-phase microgrid, second-order ripples of the dc subgrid can lead to significant power quality issues. The ripple mitigation function can be added to existing bidirectional dc–dc converters (BDCs) or dc active power filters. Considering these ripple sources can be distributed in the dc subgrid, it is essential to develop ripple mitigation control schemes for distributed BDCs. In this article, a virtual resistor based ripple mitigation strategy is proposed. Autonomous accurate ripple current sharing can be achieved for BDCs installed closely, enabling flexible parallel of ripple mitigation converters. A comprehensive parameter design method is developed to ensure the accurate sharing for closely installed BDCs. For dc subgrid with significant line impedance, the control scheme provides the opportunity for supervisory control to further enhance ripple current sharing and mitigation performance. Both simulation and experimental results validate that the control method and the parameter design method are effective.

In this article, the distributed synchronization control problem is investigated for a class of uncertain nonlinear leader-following systems on directed graph, where unknown backlash-like hysteresis exists in the actuator of each follower and the topology of the graph is fixed. Based on basic graph theory, the principle of variable structure control and Lyapunov theory, by using the unknown smooth function approximation capability of type-2 fuzzy logic systems, a new distributed adaptive supervisory control design method is proposed such that all followers asymptotically synchronize to the leader and each synchronizing error is bounded. In the control method, with the help of supervisor control strategy, it is guaranteed that the resulting closed-loop signals including system states, respectively, belong to the corresponding compact sets, which is necessary for fuzzy approximation theory because it is on a compact set that an unknown smooth function can be approximated by fuzzy logic system. Furthermore, the adaptive compensation terms of the optimal approximation errors are adopted to reduce the effects of modeling errors. Finally, simulation results demonstrate the effectiveness of the proposed new design method.

Supervisory control for discrete event systems that is widely studied in the literature to tackle the Petri net model’s deadlock occurrence of flexible manufacturing systems. A supervisory structure constructed using control places imposed restrictions on the concurrent operation in the systems, which degrade the supervisor’s efficiency. This study proposed a new method to construct a supervisory structure using combine control places and control transitions to ensure flexible manufacturing systems’ smooth operation. Place-Transition controller is designed for each concurrent process of the systems. Three algorithms are proposed in this method. The algorithm 1 computes the loop markings at each process of the Petri net model. Algorithm 2 is used to sort the deadlock markings based on the concurrent processes of the Petri net model. The Transition-Place controller and Place-Transition controller are designed using algorithm 3. Transition-Place controller creates extra states in the reachability graph, which is called dump markings. Place-Transition controller projects the deadlock markings to the loop markings via dump markings. The number of states created by the Transition-Place controller depends on the number of concurrent states in the Petri net model. The final supervisory structure controls the deadlock occurrence of flexible manufacturing systems with zero restrictions of systems operation. This proposed method is efficient as it retains all the states generated in the controlled Petri net model.

This paper proposes a design of a supervisory model predictive controller for a heating-ventilation-air-conditioning (HVAC) control system. The control objective is to minimize the operating cost and take into account of electrical load shaving and thermal comfort of users. To ensure that thermal comfort is well regulated, we utilize the Predicted Mean Vote (PMV) as an indicator and determine an acceptable bound of a desired set-point temperature. The control design consists of two layers, namely, a supervisory control (SC) layer and a model predictive control (MPC) layer. For the SC layer, we explore a configuration for the SC layer including the choice of predesign controller, the analysis of steady-state response, and the analysis of the possible range of the set-point temperature. We incorporate the effect of set-point temperature, air velocity, outside air temperature, heat load inside zone onto the HVAC electrical power. Then, we search for an optimal profile of the set-point temperature that minimizes a weighted sum of a total operating cost (TOC) and a thermal comfort cost (TCC). Moreover, exploration of trade-off between TOC and TCC helps us to achieve both control objectives efficiently. For the MPC layer, we formulate the control design with the objective of tracking the optimal set-point temperature and minimizing the control inputs. We apply the proposed control design to a complex dynamical model of HVAC system with volumetric flow rate, electrical power of heat exchanger, and removed moisture as control variables. When the allowable PMV of the zone is specified within 0.5, the TOC of the proposed supervisory MPC is reduced by 12.46% compared with that of the nominal operation. The maximum electricity demand of HVAC system is reduced by 13.43% and the electrical power profile is smoothly shaved under the proposed control scheme. Furthermore, the zone relative humidity is well regulated which corresponds to the assumption in the supervisory control.

In recent years, there has been a growing concern for robust supervisory control policies that can handle both deadlock and blockage propagation in automated manufacturing systems in the event of resource failures. This work proposes a novel robust supervisory control policy for automated manufacturing systems with multiple unreliable resources without using central buffers. The policy permits legal states as many as possible and ensures that parts not requiring unreliable resources can be automatically processed without human intervention if one or multiple unreliable resources fail. It is based on the modified neighborhood policy, namely single route neighborhood, which handles the allocation of failure-prone resources in a system. To guarantee deadlock-free operations in the remaining parts of the system, monitors are designed for emptiable strictly minimal siphons. Through examples, the applicability of the proposed policy is demonstrated.

We address the problem of flight control in the presence of actuator anomalies. A supervisory control architecture that includes the actions of both a human pilot and an autopilot is proposed to ensure resilient tracking performance in the presence of anomalies. The pilot is tasked with supervisory, higher level decision-making tasks, such as anomaly detection, estimation, and command regulation. The autopilot is assigned a lower level task of accurate command following based on an adaptive control design. The main innovations in the proposed architecture are the use of human pilot in utilizing the concepts of capacity for maneuver (CfM) and graceful command degradation (GCD), both of which originate in cognitive sciences and a judicious combination of the pilot inputs and the autopilot control action. Together, they provide guidelines for a system to be resilient, which corresponds to the system’s readiness to respond to unforeseen events. The supervisory control architecture is shown to be capable of achieving maximum CfM while allowing minimal GCD, as well as satisfactory command following postanomaly, resulting in resilient flight capabilities. The proposed controller is analyzed in a simulation study of a nonlinear F-16 aircraft model under actuator anomalies. It is shown through numerical studies that under suitable inputs from the pilot, the overall controller is able to deliver resilient flight.

This paper considers supervisory control of discrete event systems with observation delays and control delays, which is termed deterministic networked supervisory control. Delay observability, together with delay controllability, is required for obtaining a networked supervisor so that the language generated by the supervised system is deterministic and is equal to a given specification language. In this article, we simplify this existence condition and prove such a networked supervisor also exists if controllability and delay observability are satisfied. A new language property called relative delay observability, which accommodates observation delays in relative observability, is further defined. Relatively delay observability implies delay observability and enjoys an important property that is closed under union. We then show how to check the relative delay observability in polynomial time complexity. A general algorithm used to calculate the supremal sublanguage satisfying controllability and another user-defined language property in a solution space of interest, is presented. As a special case, we calculate the supremal controllable and relatively delay observable sublanguage. Finally, a practical example is given to illustrate the application of relative delay observability.

This letter considers the problem of obstacle avoidance for multiple robotic agents moving in an environment with obstacles. A decentralized supervisory controller is synthesized based on control barrier functions (CBF) that guarantees obstacle avoidance with limited actuation capability. The proposed method is applicable to general nonlinear robot dynamics and is scalable to an arbitrary number of agents. Agent-to-agent communication is not required, yet a simple broadcasting scheme improves the performance of the algorithm. The key idea is based on a control barrier function constructed with a backup controller, and we show that by assuming other agents respecting the same CBF condition, the supervisory control algorithm can be implemented decentrally and guarantees obstacle avoidance for all agents.

In this article, we investigate the synthesis of a (possibly nondeterministic) supervisor, with partial information, such that the controlled nondeterministic plant is bisimulation equivalent to a nondeterministic specification. We derive a new necessary and sufficient condition for the existence of a supervisor and provide an algorithm for the supervisor construction when one exists. The algorithm is exponential in the size of the plant and specification, a substantial improvement over the double exponential cost from exhaustive search. We show that the cases of complete information and deterministic control are special cases of the presented theory.