Search

Control auto-tuning for industrial and robotic systems, when framed as an optimization problem, provides an excellent means to tune these systems. However, most optimization methods are computationally costly, and this is problematic for high-dimension control parameter spaces. In this paper, we present a multi-stage control tuning framework that decomposes control tuning into subtasks, each with a reduced-dimension search space. We show formally that this framework reduces the sample complexity of the control-tuning task. We empirically validate this result by applying a Bayesian optimization approach to tuning multiple PID controllers in an unmanned underwater vehicle benchmark system. We demonstrate an 86\% decrease in computational time and 36\% decrease in sample complexity.

Primary Immune thrombocytopenia (ITP) is a rare autoimmune disease characterised by immune-mediated destruction of peripheral blood platelets in patients leading to low platelet counts and bleeding. The diagnosis and effective management of ITP is challenging because there is no established test to confirm the disease and no biomarker with which one can predict the response to treatment and outcome. In this work we conduct a feasibility study to check if machine learning can be applied effectively for diagnosis of ITP using routine blood tests and demographic data in a non-acute outpatient setting. Various ML models, including Logistic Regression, Support Vector Machine, k-Nearest Neighbor, Decision Tree and Random Forest, were applied to data from the UK Adult ITP Registry and a general hematology clinic. Two different approaches were investigated: a demographic-unaware and a demographic-aware one. We conduct extensive experiments to evaluate the predictive performance of these models and approaches, as well as their bias. The results revealed that Decision Tree and Random Forest models were both superior and fair, achieving nearly perfect predictive and fairness scores, with platelet count identified as the most significant variable. Models not provided with demographic information performed better in terms of predictive accuracy but showed lower fairness score, illustrating a trade-off between predictive performance and fairness.

We propose a novel framework for modelling attack scenarios in cyber-physical control systems: we represent a cyber-physical system as a constrained switching system, where a single model embeds the dynamics of the physical process, the attack patterns, and the attack detection schemes. We show that this is compatible with established results in the analysis of hybrid automata, and, specifically, constrained switching systems. Moreover, we use the developed models to compute the impact of cyber attacks on the safety properties of the system. In particular, we characterise system safety as an asymptotic property, by calculating the maximal safe set. The resulting new impact metrics intuitively quantify the degradation of safety under attack. We showcase our results via illustrative examples., Comment: 8 pages, 5 figures, submitted for presentation to IFAC World Congress 2023, Yokohama, JAPAN

Feature extraction is a fundamental task in the application of machine learning methods to SAT solving. It is used in algorithm selection and configuration for solver portfolios and satisfiability classification. Many approaches have been proposed to extract meaningful attributes from CNF instances. Most of them lack a working/updated implementation, and the limited descriptions lack clarity affecting the reproducibility. Furthermore, the literature misses a comparison among the features. This paper introduces SATfeatPy, a library that offers feature extraction techniques for SAT problems in the CNF form. This package offers the implementation of all the structural and statistical features from there major papers in the field. The library is provided in an up-to-date, easy-to-use Python package alongside a detailed feature description. We show the high accuracy of SAT/UNSAT and problem category classification, using five sets of features generated using our library from a dataset of 3000 SAT and UNSAT instances, over ten different classes of problems. Finally, we compare the usefulness of the features and importance for predicting a SAT instance's original structure in an ablation study., Comment: 8 pages, 2 figures, code available at https://github.com/bprovanbessell/SATfeatPy

One of the grand challenges in heliophysics is the characterisation of coronal mass ejection (CME) magnetic structure and evolution from eruption at the Sun through heliospheric propagation. At present, the main difficulties are related to the lack of direct measurements of the coronal magnetic fields and the lack of 3D in-situ measurements of the CME body in interplanetary space. Nevertheless, the evolution of a CME magnetic structure can be followed using a combination of multi-point remote-sensing observations and multi-spacecraft in-situ measurements as well as modelling. Accordingly, we present in this work the analysis of two CMEs that erupted from the Sun on 28 April 2012. We follow their eruption and early evolution using remote-sensing data, finding indications of CME--CME interaction, and then analyse their interplanetary counterpart(s) using in-situ measurements at Venus, Earth, and Saturn. We observe a seemingly single flux rope at all locations, but find possible signatures of interaction at Earth, where high-cadence plasma data are available. Reconstructions of the in-situ flux ropes provide almost identical results at Venus and Earth but show greater discrepancies at Saturn, suggesting that the CME was highly distorted and/or that further interaction with nearby solar wind structures took place before 10 AU. This work highlights the difficulties in connecting structures from the Sun to the outer heliosphere and demonstrates the importance of multi-spacecraft studies to achieve a deeper understanding of the magnetic configuration of CMEs., Comment: 40 pages, 11 figures, accepted for publication in Journal of Geophysical Research: Space Physics

Stealthy attacks on Industrial Control Systems can cause significant damage while evading detection. In this paper, instead of focusing on the detection of stealthy attacks, we aim to provide early warnings to operators, in order to avoid physical damage and preserve in advance data that may serve as an evidence during an investigation. We propose a framework to provide grounds for suspicion, i.e. preliminary indicators reflecting the likelihood of success of a stealthy attack. We propose two grounds for suspicion based on the behaviour of the physical process: (i) feasibility of a stealthy attack, and (ii) proximity to unsafe operating regions. We propose a metric to measure grounds for suspicion in real-time and provide soundness principles to ensure that such a metric is consistent with the grounds for suspicion. We apply our framework to Linear Time-Invariant (LTI) systems and formulate the suspicion metric computation as a real-time reachability problem. We validate our framework on a case study involving the benchmark Tennessee-Eastman process. We show through numerical simulation that we can provide early warnings well before a potential stealthy attack can cause damage, while incurring minimal load on the network. Finally, we apply our framework on a use case to illustrate its usefulness in supporting early evidence collection.

Attacks on Industrial Control Systems (ICS) can lead to significant physical damage. While offline safety and security assessments can provide insight into vulnerable system components, they may not account for stealthy attacks designed to evade anomaly detectors during long operational transients. In this paper, we propose a predictive online monitoring approach to check the safety of the system under potential stealthy attacks. Specifically, we adapt previous results in reachability analysis for attack impact assessment to provide an efficient algorithm for online safety monitoring for Linear Time-Invariant (LTI) systems. The proposed approach relies on an offline computation of symbolic reachable sets in terms of the estimated physical state of the system. These sets are then instantiated online, and safety checks are performed by leveraging ideas from ellipsoidal calculus. We illustrate and evaluate our approach using the Tennessee-Eastman process. We also compare our approach with the baseline monitoring approaches proposed in previous work and assess its efficiency and scalability. Our evaluation results demonstrate that our approach can predict in a timely manner if a false data injection attack will be able to cause damage, while remaining undetected. Thus, our approach can be used to provide operators with real-time early warnings about stealthy attacks.

The Cassini mission entered the Grand Finale phase in April 2017 and executed 22.5 highly inclined, close-in orbits around Saturn before diving into the planet on September 15th 2017. Here we present our analysis of the Cassini Grand Finale magnetometer (MAG) dataset, focusing on Saturn's internal magnetic field. These measurements demonstrate that Saturn's internal magnetic field is exceptionally axisymmetric, with a dipole tilt less than 0.007 degrees (25.2 arcsecs). Saturn's magnetic equator was directly measured to be shifted northward by ~ 0.0468 +/- 0.00043 (1-sigma) $R_S$, 2820 +/- 26 km, at cylindrical radial distances between 1.034 and 1.069 $R_S$ from the spin-axis. Although almost perfectly axisymmetric, Saturn's internal magnetic field exhibits features on many characteristic length scales in the latitudinal direction. Examining Br at the a=0.75 $R_S$, c=0.6993 $R_S$ isobaric surface, the degrees 4 to 11 contributions correspond to latitudinally banded magnetic perturbations with characteristic width similar to that of the off-equatorial zonal jets observed in the atmosphere of Saturn. Saturn's internal magnetic field beyond 60 degrees latitude, in particular the small-scale features, are less well constrained by the available measurements, mainly due to incomplete spatial coverage in the polar region. A stably stratified layer thicker than 2500 km likely exists above Saturn's deep dynamo to filter out the non-axisymmetric internal magnetic field. A heat transport mechanism other than pure conduction, e.g. double diffusive convection, must be operating within this layer to be compatible with Saturn's observed luminosity. The latitudinally banded magnetic perturbations likely arise from a shallow secondary dynamo action with latitudinally banded differential rotation in the semi-conducting layer., Comment: Accepted for publication in Icarus

Saturn's magnetospheric magnetic field, planetary radio emissions, plasma populations and magnetospheric structure are all known to be modulated at periods close to the assumed rotation period of the planetary interior. These oscillations are readily apparent despite the high degree of axi-symmetry in the internally produced magnetic field of the planet, and have different rotation periods in the northern and southern hemispheres. In this paper we study the spatial structure of (near-) planetary period magnetic field oscillations in Saturn's equatorial magnetosphere. Extending previous analyses of these phenomena, we include all suitable data from the entire Cassini mission during its orbital tour of the planet, so as to be able to quantify both the amplitude and phase of these field oscillations throughout Saturn's equatorial plane, to distances of 30 planetary radii. We study the structure of these field oscillations in view of both independently rotating northern and southern systems, finding spatial variations in both magnetic fields and inferred currents flowing north-south that are common to both systems. With the greatly expanded coverage of the equatorial plane achieved during the latter years of the mission, we are able to present a complete survey of dawn-dusk and day-night asymmetries in the structure of the oscillating fields and currents. We show that the general structure of the rotating currents is simpler than previously reported, and that the relatively enhanced nightside equatorial fields and currents are due in part to related periodic vertical motion of Saturn's magnetotail current sheet.

We identify multiple periodic dusty structures in Saturn's Roche Division, a faint region spanning the $\sim3000$ km between the A and F rings. The locations and extent of these features vary over Cassini's tour of the Saturn system, being visible in 2006 and 2016-2017, but not in 2012-2014. These changes can be correlated with variations in Saturn's magnetospheric periods. In 2006 and 2016-2017, one of the drifting magnetospheric periods would produce a 3:4 resonance within the Roche Division, but in 2012-2014 these resonances would move into the A ring as the magnetospheric periods converged. A simple model of magnetic perturbations indicates that the magnetic field oscillations responsible for these structures have amplitudes of a few nanotesla, comparable to the magnetic field oscillation amplitudes of planetary period oscillations measured by the magnetometer onboard Cassini. However, some previously unnoticed features at higher radii have expected pattern speeds that are much slower than the magnetospheric periodicities. These structures may reflect an unexpectedly long-range propagation of resonant perturbations within dusty rings., Comment: Accepted for publication in Icarus