This paper presents an overview of technology related to on-board microgrids for the more electric aircraft. All aircraft use an isolated system, where security of supply and power density represents the main requirements. Different distribution systems (ac and dc) and voltage levels coexist, and power converters have the central role in connecting them with high reliability and high power density. Ensuring the safety of supply with a limited redundancy is one of the targets of the system design since it allows increasing the power density. This main challenge is often tackled with proper load management and advanced control strategies, as highlighted in this paper.

/pdf/on-board-microgrids-for-the-more-electric-aircraft-3lqh4el4ee.pdf

On-Board Microgrids for the More Electric Aircraft—Technology Review

The continuously growing amount of monitored data in the Industry 4.0 context requires strong and reliable anomaly detection techniques. The advancement of Digital Twin technologies allows for realistic simulations of complex machinery; therefore, it is ideally suited to generate synthetic datasets for the use in anomaly detection approaches when compared to actual measurement data. In this article, we present novel weakly supervised approaches to anomaly detection for industrial settings. The approaches make use of a Digital Twin to generate a training dataset, which simulates the normal operation of the machinery, along with a small set of labeled anomalous measurement from the real machinery. In particular, we introduce a clustering-based approach, called cluster centers (CC), and a neural architecture based on the Siamese Autoencoders (SAE), which are tailored for weakly supervised settings with very few labeled data samples. The performance of the proposed methods is compared against various state-of-the-art anomaly detection algorithms on an application to a real-world dataset from a facility monitoring system, by using a multitude of performance measures. Also, the influence of hyperparameters related to feature extraction and network architecture is investigated. We find that the proposed SAE-based solutions outperform state-of-the-art anomaly detection approaches very robustly for many different hyperparameter settings on all performance measures.

Real-World Anomaly Detection by Using Digital Twin Systems and Weakly Supervised Learning

Aircraft transportation holds a tremendous importance in today’s society. Considering the international roadmaps for the intensification of the air travel and the targets concerning the environmental impact, aircraft efficiency has been in the spotlight of scientific research for more than two decades. Electrification of the aircraft subsystems and, in the future, of the propulsion appears to be the way forward to reach these ambitious goals. This Special Section collects contributions related to the more electric aircraft (MEA) technologies, from the power systems to the actuation systems.

/pdf/on-board-microgrids-for-the-more-electric-aircraft-43pgzhos3k.pdf

On-Board Microgrids for the More Electric Aircraft

The threat of attack faced by cyber-physical systems (CPSs), especially when they play a critical role in automating public infrastructure, has motivated research into a wide variety of attack defence mechanisms. Assessing their effectiveness is challenging, however, as realistic sets of attacks to test them against are not always available. In this paper, we propose smart fuzzing, an automated, machine learning guided technique for systematically finding 'test suites' of CPS network attacks, without requiring any knowledge of the system's control programs or physical processes. Our approach uses predictive machine learning models and metaheuristic search algorithms to guide the fuzzing of actuators so as to drive the CPS into different unsafe physical states. We demonstrate the efficacy of smart fuzzing by implementing it for two real-world CPS testbeds---a water purification plant and a water distribution system---finding attacks that drive them into 27 different unsafe states involving water flow, pressure, and tank levels, including six that were not covered by an established attack benchmark. Finally, we use our approach to test the effectiveness of an invariant-based defence system for the water treatment plant, finding two attacks that were not detected by its physical invariant checks, highlighting a potential weakness that could be exploited in certain conditions.

https://ink.library.smu.edu.sg/cgi/viewcontent.cgi?article=5908&context=sis_research

Learning-guided network fuzzing for testing cyber-physical system defences

Data-centric approaches are becoming increasingly common in the creation of defense mechanisms for critical infrastructure such as the electric power grid and water treatment plants. Such approaches often use well-known methods from machine learning and system identification, i.e., the Multi-Layer Perceptron, Convolutional Neural Network, and Deep Auto Encoders to create process anomaly detectors. Such detectors are then evaluated using data generated from an operational plant or a simulator; rarely is the assessment conducted in real time on a live plant. Regardless of the method to create an anomaly detector, and the data used for performance evaluation, there remain significant challenges that ought to be overcome before such detectors can be deployed with confidence in city-scale plants or large electric power grids. This position paper enumerates such challenges that the authors have faced when creating data-centric anomaly detectors and using them in a live plant.

/pdf/challenges-in-machine-learning-based-approaches-for-real-48afh82r3o.pdf

Challenges in Machine Learning based approaches for Real-Time Anomaly Detection in Industrial Control Systems

Industrial Control Systems (ICS) are found in critical infrastructure such as for power generation and water treatment. When security requirements are incorporated into an ICS, one needs to test the additional code and devices added do improve the prevention and detection of cyber attacks. Conducting such tests in legacy systems is a challenge due to the high availability requirement. An approach using Timed Automata (TA) is proposed to overcome this challenge. This approach enables assessment of the effectiveness of an attack detection method based on process invariants. The approach has been demonstrated in a case study on one stage of a 6- stage operational water treatment plant. The model constructed captured the interactions among components in the selected stage. In addition, a set of attacks, attack detection mechanisms, and security specifications were also modeled using TA. These TA models were conjoined into a network and implemented in UPPAAL. The models so implemented were found effective in detecting the attacks considered. The study suggests the use of TA as an effective tool to model an ICS and study its attack detection mechanisms as a complement to doing so in a real plant–operational or under design.

Testing the Effectiveness of Attack Detection Mechanisms in Industrial Control Systems

An energy management system of a microgrid (MG) has several basic objectives; e.g. to maximize the utilization of renewable energy resources (RES), to protect the internal components from overloading, and to ensure that the MG operates reliably under any operating conditions. Although many control techniques are available in the literature to monitor and control the energy flows among distributed RES in MGs, formal verification of those techniques was not proposed yet. The emphasis of this paper is to design and validate energy management system for a MG which consists of a solar photovoltaic (PV) array, a pair of battery energy storage systems (BESes), a diesel generator (DG) and a load (LD). The physics and dynamics of the MG are defined as energy flow invariants and the designed behaviours are abstracted, modelled and validated in this work. Therefore, we have considered an invariant based flow technique to manage the energy flow in an MG. The results are validated and verified with UPPAAL, a powerful industrial tool which is commonly used to verify the correctness of real-time systems like supervisory controllers, communication protocols and others.

Formal validation of supervisory energy management systems for microgrids

This article presents the modeling and verification of supervisory energy-management systems (EMSs) for microgrids using timed automata (TA) and a formal verification approach. The EMS plays an essential role in managing the power flow among different components in the microgrid system for its safe and reliable operation. The modeling of the EMS is based on predefined invariants with allowable and nonallowable operating modes, which are the conditions that do not change over time. The failure of invariants could have severe effects on microgrid system functionality, which highlights the importance of verification during the initial stage of EMS design. Conventional approaches, such as simulation and/or experimental verification, require manual checking and skilled professional knowledge to check EMS design correctness. Also, there may be a corner case that could lead to system failure that goes unidentified by manual analysis.

/pdf/supervisory-energy-management-systems-for-microgrids-a3o0b3p35b.pdf

Supervisory Energy-Management Systems for Microgrids: Modeling and Formal Verification

Several methods exist for detecting process anomalies resulting from cyber-attacks on critical infrastructure. The assessment of such methods could be conducted using simulation or directly on a realistic operational testbed. While the results of an assessment on a testbed may be more authentic than those carried out using simulation, conducting such experiments is fraught with challenges such as the time required to set up and launch attacks thus limiting the variety and number of attacks launched. To overcome such limitations, while maintaining the reliability of the outcome of the assessment, an approach based on timed automata models of a critical infrastructure was investigated. The investigation involved development of a digital twin for a 6-stage water treatment plant. A design-centric anomaly detection method, as well as an attack launcher, were integrated with the model and experiments were performed. The outcome of this investigation reveals the value of the proposed approach in rapid assessment of a design-centric anomaly detection method.

Gayathri Sugumar

Papers

Testing the Effectiveness of Attack Detection Mechanisms in Industrial Control Systems

Formal validation of supervisory energy management systems for microgrids

Supervisory Energy-Management Systems for Microgrids: Modeling and Formal Verification

Assessment of a Method for Detecting Process Anomalies Using Digital-Twinning

A method for testing distributed anomaly detectors