The development of data sensing technology has generated a vast amount of high-dimensional data, posing great challenges for machine learning models. Over the past decades, despite demonstrating its effectiveness in data classification, genetic programming (GP) has still encountered three major challenges when dealing with high-dimensional data: 1) solution diversity; 2) multiclass imbalance; and 3) large feature space. In this article, we have developed a problem-specific multiobjective GP framework (PS-MOGP) for handling classification tasks with high-dimensional data. To reduce the large solution space caused by high dimensionality, we incorporate the recursive feature elimination strategy based on mining the archive of evolved GP solutions. A progressive domination Pareto archive evolution strategy (PD-PAES), which optimizes the objectives in a specific order according to their objectives, is proposed to evaluate the GP individuals and maintain a better diversity of solutions. Besides, to address the seriously imbalanced class issue caused by traditional binary decomposition (BD) one versus rest (OVR) for multiclass classification problems, we design a method named BD with a similar positive and negative class size (BD-SPNCS) to generate a set of auxiliary classifiers. Experimental results on benchmark and real-world datasets demonstrate that our proposed PS-MOGP outperforms state-of-the-art traditional and evolutionary classification methods in the context of high-dimensional data classification.

Power grid damage and blackouts are increasing with climate change. Load forecasting methods that integrate climate resilience are therefore essential to facilitate timely and accurate network reconfiguration during periods of extreme stress. Our paper proposes a generalised Wildfire Resilient Load Forecasting Model (WRLFM) to predict electricity load based on operational data of a Distribution Network (DN) in Australia during wildfire seasons in 2015–2020. We demonstrate that load forecasting during wildfire seasons is more challenging than during non-wildfire seasons, motivating an imperative need to improve forecast performance during wildfire seasons. To develop the robust WRLFM, comprehensive comparative analyses were conducted to determine proper Machine Learning (ML) forecast structures and methods for incorporating multiple factors. Bi-directional Gated Recurrent Unit (Bi-GRU) and Vision Transformer (ViT) were selected as they performed the best among all 13 recently trending ML methods. Multi-factors were incorporated to contribute to forecast performance, including input sequence structures, calendar information, flexible correlation-based temperature conditions, and categorical Fire Weather Index (FWI). High-resolution categorical FWI was used to build a forecasting model with climate resilience for the first time, significantly enhancing the average stability of forecast performances by 42%. A sensitivity analysis compared data set patterns and model performances during wildfire and non-wildfire seasons. The improvement rate of load forecasting performance during wildfire seasons was more than two times greater than in non-wildfire seasons. This indicates the significance and effectiveness of applying the WRLFM to improve forecast accuracy under extreme weather risks. Overall, the WRLFM reduces the Mean Absolute Percentage Error (MAPE) of the forecast by 14.37% and 20.86% for Bi-GRU and ViT-based models, respectively, achieving an average forecast MAPE of around 3%.

Clays exhibit complex mechanical behaviour with significant viscous, nonlinear, and hysteric characteristics, beyond the prediction capacity of the well-known modified cam clay (MCC) model. This paper extends the MCC model to address these important limitations. The proposed family of models is constructed entirely within the hyperplasticity framework deduced from thermodynamic extremal principles. More specifically, the previously developed MCC hyper-viscoplastic model based on the isotache concept is extended to incorporate multiple internal variables and to capture recent loading history, hysteresis, and smooth response of the material. This is achieved by defining an inelastic free energy and an element that implements a bounding surface within hyperplasticity, resulting in pressure dependency in both reversible and irreversible processes with a unique critical state envelope, and only eight material parameters with a readily measurable viscous parameter. A kinematic hardening in the logistic differential form in stress space is derived that enables the proposed model to function effectively across a wide range of stresses. Based on this kinematic hardening rule, the current stress state acts as an asymptotic attractor for the back/shift stresses whose evolution rates are proportional to their current state.

This paper describes experiments addressing the buckling and collapse behaviour of cold-formed stainless steel cruciform section columns. Doubly symmetrical flanged cruciform section columns were built using six individual press-braked plain channels made of austenitic grade EN 1.4307 assembled back-to-back. Three different column lengths were tested – namely, short (600 mm), intermediate (1200 mm) and long (2400 mm) lengths, and two channel geometries were used in all tests – with cross-section depths of 200 mm and 100 mm. Tests were carried out under pure axial compression and with fixed end support conditions. Tests were repeated two times for each length. It was observed that the buckling patterns were affected both by the global slenderness and by the spacing between the fasteners. Stocky columns experienced local buckling of the individual channel sections. The plastic failure mechanism was dependent on the fasteners spacing. Intermediate slenderness columns were characterised by interaction between global torsional buckling and local buckling whereas more slender columns failed predominantly through torsional buckling. For the latter, the spacing between the fasteners had a minor influence on the ultimate capacity.

Today, wire arc additive manufacturing (WAAM) can be used to fabricate critical structural steel components. The process allows to fabricate sophisticated shapes, thereby achieving very high levels of material capacity utilization. Key welding parameters inevitably influence the mechanical resistance of components made by WAAM, submitted to static or cyclic loading. In this research, the fatigue behaviour of wire arc additively manufactured carbon steel elements is investigated. Firstly, samples are manufactured by cold metal transfer (CMT) welding process using carbon steel 3Dprint AM 35 (S355) grade and following a welding procedure optimised to limit the welding-induced imperfections. A series of microstructural investigations and mechanical experiments are carried out on milled samples including: (i) hardness tests, (ii) static tensile tests, and (iii) cyclic fatigue tests. Both transverse and longitudinal directions are tested. The obtained fatigue test results are then compared against existing research on equivalent details. A database containing all similar test results and own experimental results is then used to calculate the fatigue detail categories, and to assess the applicability of the current Eurocode and IIW provisions for fatigue. The reliability levels of the proposed fatigue classes are then validated through the use of Weibull models, commonly used in survival analysis.

The electrification of heat necessitates the development of innovative domestic heat batteries to effectively balance energy demand with renewable power supply. Thermochemical heat storage systems show great promise in supporting the electrification of heating, thanks to their high thermal energy storage density and minimal thermal losses. Among these systems, salt hydrate-based thermochemical systems are particularly appealing. However, they do suffer from slow hydration kinetics in the presence of steam, which limits the achievable power density. Additionally, their relatively high dehydration temperature hinders their application in supporting heating systems. Furthermore, there are still challenges regarding the appropriate thermodynamic, physical, kinetic, chemical, and economic requirements for implementing these systems in heating applications. This study analyzes a proposal for thermochemical energy storage based on the direct hydration of sodium acetate with liquid water. The proposed scheme satisfies numerous requirements for heating applications. By directly adding liquid water to the salt, an unprecedented power density of 5.96 W/g is achieved, nearly two orders of magnitude higher than previously reported for other salt-based systems that utilize steam. Albeit the reactivity drops as a consequence of deliquescence and particle aggregation, it has been shown that this deactivation can be effectively mitigated by incorporating 10 % silica, achieving lower but stable energy and power density values. Furthermore, unlike other salts studied previously, sodium acetate can be fully dehydrated at temperatures within the ideal range for electrified heating systems such as heat pumps (40 °C – 60 °C). The performance of the proposed scheme in terms of dehydration, hydration, and multicyclic behavior is determined through experimental analysis.

RationaleHeterogeneity of the host response within sepsis, acute respiratory distress syndrome (ARDS) and more widely critical illness, limits discovery and targeting of immunomodulatory therapies. Clustering approaches using clinical and circulating biomarkers have defined hyper-inflammatory and hypo-inflammatory subphenotypes in ARDS associated with differential treatment response. It is unknown if similar subphenotypes exist in sepsis populations where leucocyte transcriptomic-defined subphenotypes have been reported.ObjectivesWe investigated whether inflammatory clusters based on cytokine protein abundance were seen in sepsis, and the relationships with previously described transcriptomic subphenotypes.MethodsHierarchical cluster and latent class analysis were applied to an observational study (UK Genomic Advances in Sepsis (GAinS)) (n=124 patients) and two clinical trial datasets (VANISH, n=155 and LeoPARDS, n=484) in which the plasma protein abundance of 65, 21, 11 circulating cytokines, cytokine receptors and regulators were quantified. Clinical features, outcomes, response to trial treatments and assignment to transcriptomic subphenotypes were compared between inflammatory clusters.Measurements and main resultsWe identified two (UK GAinS, VANISH) or three (LeoPARDS) inflammatory clusters. A group with high levels of pro-inflammatory and anti-inflammatory cytokines was seen that was associated with worse organ dysfunction and survival. No interaction between inflammatory clusters and trial treatment response was found. We found variable overlap of inflammatory clusters and leucocyte transcriptomic subphenotypes.ConclusionsThese findings demonstrate that differences in response at the level of cytokine biology show clustering related to severity, but not treatment response, and may provide complementary information to transcriptomic sepsis subphenotypes.Trial registration numberISRCTN20769191, ISRCTN12776039.