Search

Supply chains operate in a highly distuptive environment where a SC master plan should be updated in line with disruptions to ensure that a high service level is provided to customers while total cost is minimised. There is an absence of knowledge of how a SC master plan should be updated to cope with disruptions using hybrid modelling. To fill this gap, we present a hybrid modelling framework to update a SC master plan in presence of disruptions. The proposed framework, which is a precursor to a SC digital twin, integrates simulation, machine learning, and optimisation to identify the production, storage, and distribution values that maximise SC service level while minimising total cost under disruptions. This approach proves effective in a SC disrupted by demand increase and lead time extension. Results show that employing hybrid modelling leads to a noticeable improvement in service level and total cost. The outcome of the new knowledge on using hybrid modelling for managing disruptions provides essential learning for the extension of modelling through a digital twin for SC master planning. We observe that in the presence of disruptions it is more economical to keep higher inventory at downstream SC members than the upstream SC members. [ABSTRACT FROM AUTHOR]

Streamflow simulation is crucial for flood mitigation, ecological protection, and water resource planning. Process-based hydrological models and machine learning algorithms are the mainstream tools for streamflow simulation. However, their inherent limitations, such as time-consuming and large data requirements, make achieving high-precision simulations challenging. This study developed a hybrid approach to simultaneously improve the accuracy and computational efficiency of streamflow simulation, which integrates Block-wise use of the TOPMODEL (BTOP) model into the eXtreme Gradient Boosting (XGBoost), i.e., BTOP_XGB. In this approach, BTOP generates simulated streamflow using the Latin hypercube sampling algorithm instead of the time-consuming calibration algorithms to reduce computational costs. Then, XGBoost combines BTOP simulated streamflow with multi-source data to reduce simulation errors. In which, serval input variable selection algorithms are employed to choose relevant inputs and remove redundant information for model. The hybrid approach is validated and compared with a standalone model at three hydrological stations in the Jialing River basin, China. The results show that the performance of BTOP_XGB is significantly better than the BTOP and XGBoost models. The NSE of BTOP_XGB at Beibei, Xiaoheba, and Luoduxi stations increases by 54%, 21%, and 83%, respectively. Meanwhile, the computational time of BTOP_XGB is saved by >90% compared to the original calibrated BTOP. BTOP_XGB is less affected by parameter sample sizes and data amounts, demonstrating the robustness of the hybrid model. This study simplifies the complexity of the hydrological model and enhances the stability of machine learning, jointly improving the reliability of streamflow simulation. The hybrid approach provides a potential shortcut for streamflow simulation over basins with large areas or limited observed data.

Copyright of BHM Berg- und Hüttenmännische Monatshefte is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Österreichische Wasser- und Abfallwirtschaft is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Hydrological modelling, essential for water resources management, can be very complex in karst catchments with different climatic and geologic characteristics. In this study, three combined conceptual models incorporating the snow module with machine learning models were used for hourly rainfall-runoff modelling in the mostly karst Ljubljanica River catchment, Slovenia. Wavelet-based Extreme Learning Machine (WELM) and Wavelet-based Regression Tree (WRT) machine learning models were integrated into the conceptual CemaNeige Génie Rural à 4 paramètres Horaires (CemaNeige GR4H). In this regard, the performance of the hybrid models was compared with stand-alone conceptual and machine learning models. The stand-alone WELM and WRT models using only meteorological variables performed poorly for hourly runoff forecasting. The CemaNeige GR4H model as stand-alone model yielded good performance; however, it overestimated low flows. The hybrid CemaNeige GR4H-WELM and CemaNeige-WRT models provided better simulation results than the stand-alone models, especially regarding the extreme flows. The results of the study demonstrated that using different variables from the conceptual model, including the snow module, in the machine learning models as input data can significantly affect the performance of rainfall-runoff modelling. The hybrid modelling approach can potentially improve runoff simulation performance in karst catchments with diversified geological formations where the rainfall-runoff process is more complex. [ABSTRACT FROM AUTHOR]

The practice of deceased-donor solid-organ transplantation is unique among clinical therapies because it substantially benefits recipients, The allocation process is a complex System-of-Systems (SoS) with multiple stakeholders, components, rules, and policies. While we model and simulate such SoS, they are challenging to verify and validate thoroughly due to a large number of use cases and the sensitivity to initial conditions. Furthermore, subject matter experts within communities of practice do not trust such models because they are difficult to explain. This paper proposes a trust-centric approach to validation focusing on experimentally exploring the simulation's behavior space and demonstrating its utility to a community of experts. We describe a three-step validation process to show how modelling and simulation professionals can foster trust in simulations of a complex SoS. We apply the framework to verify and validate a simulation model of the Kidney–Pancreas (KP) allocation process within the United States. [ABSTRACT FROM AUTHOR]

An important aspect in the development of small molecules as drugs or agrochemicals is their systemic availability after intravenous and oral administration. The prediction of the systemic availability from the chemical structure of a potential candidate is highly desirable, as it allows to focus the drug or agrochemical development on compounds with a favorable kinetic profile. However, such predictions are challenging as the availability is the result of the complex interplay between molecular properties, biology and physiology and training data is rare. In this work we improve the hybrid model developed earlier (Schneckener in J Chem Inf Model 59:4893–4905, 2019). We reduce the median fold change error for the total oral exposure from 2.85 to 2.35 and for intravenous administration from 1.95 to 1.62. This is achieved by training on a larger data set, improving the neural network architecture as well as the parametrization of mechanistic model. Further, we extend our approach to predict additional endpoints and to handle different covariates, like sex and dosage form. In contrast to a pure machine learning model, our model is able to predict new end points on which it has not been trained. We demonstrate this feature by predicting the exposure over the first 24 h, while the model has only been trained on the total exposure. [ABSTRACT FROM AUTHOR]

This study utilizes machine learning and deep learning algorithms, including the Decision Tree, Support Vector Machine (SVM), Conventional Neural Network (CNN), Long Short-Term Memory (LSTM), and a novel Hybrid Machine Learning model to predict student performance. The performance of these models is assessed and compared based on their accuracy. Notably, the Hybrid Machine Learning model achieved the highest accuracy, demonstrating the potential of hybrid modeling in educational data mining. While the study primarily aims to enhance the precision of student performance prediction, it also critically examines the pros and cons of various machine learning and deep learning techniques in the context of educational settings. Moreover, the study underscores the importance of further research in this area, focusing on model interoperability and generalization across different datasets. It is expected that the insights gained from this study will guide educators and researchers in refining student performance predictions, thereby bolstering student success. [ABSTRACT FROM AUTHOR]

Earth observation techniques represent a reliable and faster alternative to in-situ measurements by providing spatio-temporal information on crop status. In this framework, a study was conducted to assess the performance of hybrid approaches, either standard (HYB) or exploiting an active learning optimisation strategy (HYB-AL), to estimate leaf area index (LAI) and canopy nitrogen content (CNC) from Sentinel–2 (S2) data, in rice crops. To achieve this, the PROSAIL-PRO Radiative Transfer Model (RTM) was tested. Results demonstrate that a wide range of rice spectra, simulated according to realistic crop parameters, are reliable when appropriate field background conditions are considered. Simulations were used to train a Gaussian Process Regression (GPR) algorithm. Both cross-validation and validation results showed that HYB-AL approach resulted the best performing retrieval schema. LAI estimation achieved good performance (R2=0.86; RMSE=0.54) and resulted very promising for model application in operational monitoring systems. CNC estimations showed moderate performance (R2=0.63; RMSE=0.89) due to a saturation behaviour limiting the retrieval accuracy for moderate/high CNC values, approximately above 4 [g m−2]. S2 maps of LAI and CNC provided spatio-temporal information in agreement with crop growth, nutritional status and agro-practices applied to the study area, resulting in an important contribution to precision farming applications. [ABSTRACT FROM AUTHOR]

Soil total nitrogen is one of the most important basic indicators for fertiliser decision making, but tens of millions of soil total nitrogen sampling data have been accumulated, forming a huge database. In this large database, there is a large amount of anomalous data, which can interfere with data analysis, affect the construction of spatial interpolation and prediction models, and then affect the accuracy of nutrient management decisions. The traditional method of identifying soil total nitrogen anomalies based on boxplots suffers from the problems of not being able to identify local anomalies, which can easily lead to misclassification of soil total nitrogen data anomalies, and the detection efficiency is not high. We propose a method to identify soil total nitrogen outliers by combining the Isolation Forest algorithm and local spatial autocorrelation analysis, which can simultaneously detect global and local outliers from large amounts of data and combine organic matter as an auxiliary indicator in the spatial analysis to help judge local outliers. Finally, the results of global and local anomalies were combined to provide a comprehensive assessment of the soil nitrogen data, avoiding the misjudgement or omission of judgement that can occur when using a single method. Using 25,930 soil test data from Yunnan Province in 2009 as an example, we compared and analysed the typical boxplot method and the unsupervised OneClassSVM method and evaluated the performance of each method in terms of correct detection rate, false positive rate and false negative rate. The results show that the proposed method has a correct detection rate (TR) of 99.97%, a false positive rate (FPR) of 8.06% and a false negative rate (FNR) of 0.01% on the data, which shows high validity and accuracy; it is also comparable to the independent isolated forests (FNR = 4.76%), boxplot (FNR = 3.90%) and OneClassSVM (FNR = 4.77%), and the false negative rate is reduced by 4.75%, 3.89% and 4.76%, respectively. [ABSTRACT FROM AUTHOR]

This paper presents an efficient flood early warning system developed for the city of Mandra, Greece which experienced a devastating flood event in November 2017 resulting in significant loss of life. The location is of particular interest due to both its small-sized water basin (20 km 2 upstream of the studied cross-section), necessitating a rapid response time for effective flood warning calculations, and the lack of hydrometric data. To address the first issue, a database of pre-simulated flooding events with a 2D hydrodynamic model corresponding to synthetic precipitations with different return periods was established. To address the latter issue, the hydrological model was calibrated using qualitative information collected after the catastrophic event, compensating for the lack of hydrometric data. The case study demonstrates the establishment of a hybrid (online–offline) flood early warning system in data-scarce environments. By utilizing pre-simulated events and qualitative information, the system provides valuable insights for flood forecasting and aids in decision-making processes. This approach can be applied to other similar locations with limited data availability, contributing to improved flood management strategies and enhanced community resilience. [ABSTRACT FROM AUTHOR]

Extensive benefits of biodiesel amalgamation with diesel engine have prompted several researches towards suitable optimization of operating parameters of production process. Ultrasonic reactors are acclaimed instruments in generating biodiesel from raw oil. The contemporary research has varied the operating parameters of the ultrasonic reactor for maximum yield with the aid of artificially intelligent software's. Eucalyptus oil combined with ethanol and sulphuric acid were used as reactants to generate biodiesel. Prime input factors considered in this study comprises of reaction time, molar ratio, frequency, power and temperature. The study's results are analysed and compared with models created using intelligent hybrid prediction approaches including adaptive neuro-fuzzy inference system (ANFIS), response surface methodology (RSM) - genetic algorithm (GA). The parameters were varied and optimized for maximum biodiesel yield by employing best operating conditions for ultrasonic reactor, furnished by ANFIS and GA in MATLAB software and RSM in Minitab software. All the models performed exceptionally well, with ANFIS performing slightly better with RSME value of 0.0017 while RSM achieved a RSME value of 0.0023. Combining the precision of ANFIS's prediction with the efficiency of GA-optimization gives a reliable and thorough evaluation. Enhancing the efficiency of biodiesel production can decrease the world's fuel consumption by reducing the reliance on fossil fuels and concurrently reducing the carbon footprint by vehicles. [ABSTRACT FROM AUTHOR]

Extensive benefits of biodiesel amalgamation with diesel engine have prompted several researches towards suitable optimization of operating parameters of production process. Ultrasonic reactors are acclaimed instruments in generating biodiesel from raw oil. The contemporary research has varied the operating parameters of the ultrasonic reactor for maximum yield with the aid of artificially intelligent software’s. Eucalyptus oil combined with ethanol and sulphuric acid were used as reactants to generate biodiesel. Prime input factors considered in this study comprises of reaction time, molar ratio, frequency, power and temperature. The study’s results are analysed and compared with models created using intelligent hybrid prediction approaches including adaptive neuro-fuzzy inference system (ANFIS), response surface methodology (RSM) - genetic algorithm (GA). The parameters were varied and optimized for maximum biodiesel yield by employing best operating conditions for ultrasonic reactor, furnished by ANFIS and GA in MATLAB software and RSM in Minitab software. All the models performed exceptionally well, with ANFIS performing slightly better with RSME value of 0.0017 while RSM achieved a RSME value of 0.0023. Combining the precision of ANFIS’s prediction with the efficiency of GA-optimization gives a reliable and thorough evaluation. Enhancing the efficiency of biodiesel production can decrease the world’s fuel consumption by reducing the reliance on fossil fuels and concurrently reducing the carbon footprint by vehicles.

Resilience, when defined as the capacity of an animal to respond to short-term environmental challenges and to return to the prechallenge status, is a dynamic and complex trait. Resilient animals can reinforce the capacity of the herd to cope with often fluctuating and unpredictable environmental conditions. The ability of modern technologies to simultaneously record multiple performance measures of individual animals over time is a huge step forward to evaluate the resilience of farm animals. However, resilience is not directly measurable and requires mathematical models with biologically meaningful parameters to obtain quantitative resilience indicators. Furthermore, interpretive models may also be needed to determine the periods of perturbation as perceived by the animal. These applications do not require explicit knowledge of the origin of the perturbations and are developed based on real-time information obtained in the data during and outside the perturbation period. The main objective of this paper was to review and illustrate with examples, different modelling approaches applied to this new generation of data (i.e., with high-frequency recording) to detect and quantify animal responses to perturbations. Case studies were developed to illustrate alternative approaches to real-time and post-treatment of data. In addition, perspectives on the use of hybrid models for better understanding and predicting animal resilience are presented. Quantification of resilience at the individual level makes possible the inclusion of this trait into future breeding programmes. This would allow improvement of the capacity of animals to adapt to a changing environment, and therefore potentially reduce the impact of disease and other environmental stressors on animal welfare. Moreover, such quantification allows the farmer to tailor the management strategy to help individual animals to cope with the perturbation, hence reducing the use of pharmaceuticals, and decreasing the level of pain of the animal.

Assessment and modelling of land degradation are crucial for the management of natural resources and sustainable development. The current study aims to evaluate land degradation by integrating various parameters derived from remote sensing and legacy data with analytical hierarchy process (AHP) combined machine learning models for the Mandovi river basin of western India. Various land degradation conditioning factors comprising of topographical, vegetation, pedological, and climatic variables were considered. Integration of the factors was performed through weighted overlay analysis to generate the AHP-based land degradation map. The output of AHP was then used with land degradation conditioning factors to build AHP combined gradient boosting machine (AHP-GBM), random forest (AHP-RF), and support vector machine (AHP-SVM) model. The model performances were assessed through an area under the receiver operating characteristic (AUC). The AHP-RF model recorded the highest AUC (0.996) followed by AHP-SVM (0.987), AHP (0.977), and AHP-GBM (0.975). The study revealed that AHP combined with RF could significantly improve the model performance over solo AHP. High rainfall with high slopes and improper land use were the major causes of land degradation in the study area. The findings of the current study will aid the policymakers to formulate land degradation action plans through implementing appropriate soil and water conservation measures. [ABSTRACT FROM AUTHOR]

Many chemical and industrial processes are complex, and the dynamics of such processes cannot be explained using a partial differential equation (PDE) or a system of PDEs with constant coefficients. Parametric PDEs, that is, PDEs with their coefficients varying across time or space, are utilized for this purpose. The non‐availability of data at all spatial locations and partially available process knowledge add to the complexity of modelling such processes. This paper proposes a framework to discover parametric PDEs using data‐driven and hybrid modelling approaches with the temperature dynamics of steam‐assisted gravity drainage (SAGD) process in an oil reservoir as the system under study. We utilize an ensemble of 200 realizations of the temperature dynamics generated using the variogram for the PDE discovery. Permeability, which is one of the oil reservoir's petrophysical properties, is used to develop the hybrid models. We infer that utilizing partial process knowledge aids in improving the model's accuracy. [ABSTRACT FROM AUTHOR]

Cancer cells metabolize glucose through metabolic pathways that differ from those used by healthy and differentiated cells. In particular, tumours have been shown to consume more glucose than their healthy counterparts and to use anaerobic metabolic pathways, even under aerobic conditions. Nevertheless, scientists have still not been able to explain why cancer cells evolved to present an altered metabolism and what evolutionary advantage this might provide them. Experimental and computational models have been increasingly used in recent years to understand some of these biological questions. Multicellular tumour spheroids are effective experimental models as they replicate the initial stages of avascular solid tumour growth. Furthermore, these experiments generate data which can be used to calibrate and validate computational studies that aim to simulate tumour growth. Hybrid models are of particular relevance in this field of research because they model cells as individual agents while also incorporating continuum representations of the substances present in the surrounding microenvironment that may participate in intracellular metabolic networks as concentration or density distributions. Henceforth, in this review, we explore the potential of computational modelling to reveal the role of metabolic reprogramming in tumour growth.

Earth observation techniques represent a reliable and faster alternative to in-situ measurements by providing spatio-temporal information on crop status. In this framework, a study was conducted to assess the performance of hybrid approaches, either standard (HYB) or exploiting an active learning optimisation strategy (HYB-AL), to estimate leaf area index (LAI) and canopy nitrogen content (CNC) from Sentinel–2 (S2) data, in rice crops. To achieve this, the PROSAIL-PRO Radiative Transfer Model (RTM) was tested. Results demonstrate that a wide range of rice spectra, simulated according to realistic crop parameters, are reliable when appropriate field background conditions are considered. Simulations were used to train a Gaussian Process Regression (GPR) algorithm. Both cross-validation and validation results showed that HYB-AL approach resulted the best performing retrieval schema. LAI estimation achieved good performance (R2=0.86; RMSE=0.54) and resulted very promising for model application in operational monitoring systems. CNC estimations showed moderate performance (R2=0.63; RMSE=0.89) due to a saturation behaviour limiting the retrieval accuracy for moderate/high CNC values, approximately above 4 [g m−2]. S2 maps of LAI and CNC provided spatio-temporal information in agreement with crop growth, nutritional status and agro-practices applied to the study area, resulting in an important contribution to precision farming applications.

The blast furnace is an important part of the steelmaking process, and its main function is to melt and reduce oxygen from the iron ore for subsequent processing into the steel-ironmaking process. Due to its complexity, Blast Furnaces need to operate near their practical limits because of economic and environmental constraints. The capacity to monitor and regulate the process's thermal condition is, however, constrained by the harsh operating conditions inside the furnace. The amount of silicon present in pig iron, which is the metallic iron generated by the blast furnace process, serves as a crucial indicator of the furnace's thermal condition. Therefore, the creation of a predictive model is essential to assist in proactive control of the furnace's thermal condition because measurements of this crucial variable can only be sampled at sporadic and irregular intervals and analysis of the sample introduces a substantial delay. In this paper, an improved hybrid modelling methodology is introduced for blast furnace operation, which integrates physical and data-driven models. Deep Learning based Autoencoders are used for the prediction of the changes in silicon concentration with respect to time and that helps users to avoid running frequent and costly feature pre-processing procedures and correlation studies. Integrating the physical model improved the prediction accuracy compared to a purely data-driven model. [ABSTRACT FROM AUTHOR]

• A new mathematical model for rubber–tyred gantry cranes is developed considering the elastic structure. • Two advanced sliding mode controls are proposed to validate dynamic model. • Experiment shows the sufficient accuracy and effective application of modeling and control to cranes. A gantry crane holds strict underactuation and elastic multibody structure, and thus faces challenges in control system design. For model–based control, we develop a three–dimensional dynamic model of rubber–tyred gantry cranes considering the distributed mass and bending deformation of gantry. Such modelling hybridises the lumped model of trolley–bridge–payload motions and continuous model of gantry vibration. We consider a complex operation in which three motors composed of travelling trolley, pushing bridge and hoisting container run simultaneously. We propose two versions of the robust controller, namely, dynamic sliding mode and fast terminal sliding mode for tracking the three motions, eliminating swings of container, and reducing the bending vibration of elastic gantry. Simulation and experiment reveal precision and effective application of our crane modelling to control approaches. The control system effectively tracks and stabilises all outputs and achieves robustness when a crane suffers from the structural and parametric uncertainties. [ABSTRACT FROM AUTHOR]

For strategic planning of healthcare services, understanding and accounting for future demand can support resource management. A hybrid modelling approach is well-placed to address these issues using demand projections with simulation for capacity planning. We present a methodology which conceptualises each stage as a collaborative activity in a strategic process. Medium-to-long term demand is uncertain, and depends upon internal and external factors. Our novel demand projections use both drivers of internal demand, and environmental scanning to identify external influences on demand. These create demand trajectories for each of three procedures that are used with discrete-event simulation to inform strategic planning for endoscopic diagnostic services. The study involved collaboration with a major Healthcare Trust in England, and demonstrates the applicability of our methodology to support strategic change at the regional service level by aligning external demand with internal resource planning. [ABSTRACT FROM AUTHOR]

Bone mechanobiology is the study of the physical, biological and mechanical processes that continuously affect the multiscale multicellular system of the bone from the organ to the molecular scale. Current knowledge derives from experimental studies, which are often limited to gathering qualitative data in a cross-sectional manner, up to a restricted number of time points. Moreover, the simultaneous collection of information about 3D bone microarchitecture, cell activity as well as protein distribution and level is still a challenge. In silico models can expand qualitative information with hypothetical quantitative systems, which allow quantification, testing and comparison to existing quantifiable experimental data. An overview of multiscale, multiphysics, agent-based and hybrid techniques and their applications to bone mechanobiology is provided in the present review. The study analysed how mechanical signals, cells and proteins can be modelled in silico to represent bone remodelling and adaptation. Hybrid modelling of bone mechanobiology could combine the methods used in multiscale, multiphysics and agent-based models into a single model, leading to a unified and comprehensive understanding of bone mechanobiology. Numerical simulations of in vivo multicellular systems aided in hypothesis testing of such in silico models. Recently, in silico trials have been used to illustrate the mechanobiology of cells and signalling pathways in clinical biopsies and animal bones, including the effects of drugs on single cells and signalling pathways up to the organ level. This improved understanding may lead to the identification of novel therapies for degenerative diseases such as osteoporosis.

Cryogenic-assisted hard turning emerges as sustainable machining process. Nonetheless, the control of surface integrity for avoidance of surface defects such as white layers and tensile residual stresses is still a challenge. In the present study a hybrid surface integrity model is introduced. The model is comprised by a 3D Finite-Difference Method (FDM) thermal model, able to simulate the cryogenic cooling conditions in addition to the tool flank wear effect. Afterwards surface hardness and material phase transformations (white & dark layers) are predicted using empirical models. The effect of tool wear and cryogenic cooling on the surface integrity are examined. [ABSTRACT FROM AUTHOR]

The planning of transport infrastructure must consider both the short-term capacity demand and the equipment's maintenance needs, as well as the long-term evolution of the transportation system. In the freight transportation system, long-term evolutions stem from the individual behaviours of several stakeholders and their dynamic relationships. Their actions may change how the infrastructure is used, giving rise to bottlenecks, or novel development opportunities. However, existing project planning techniques tend to make static assumptions on these agents' behaviours when predicting future developments. Furthermore, there is a need for models able to simulate short-term developments in transport operations alongside long-term system evolutions. This work aims to contribute towards solving these gaps, particularly in what concerns the consideration of stakeholder adaptation strategies when confronted with infrastructure alterations. We propose a novel a dual-approach Agent-based Modelling concept, based on Hybrid Modelling methodologies. This framework is comprised of two distinct modules: the micro module reproduces short-term freight transport operations in the physical network, while the macro module captures the stakeholders' decision-making in the long-term. These modules continuously communicate with one another, updating critical information regarding system conditions throughout the simulation period. The developed modelling concept was assessed through a set of simulation trials, which revealed its sensitivity to different scenarios of railway project implementation and demonstrated its potential for capturing the possible outcomes of distinct infrastructure projects, stimulating the responses of stakeholders to the new market conditions, and identifying network bottlenecks. [ABSTRACT FROM AUTHOR]

Physical model testing is a crucial step in the validation process of floating offshore wind concepts that can be used to contribute to the worldwide decarbonization objectives. It is also important to understand the coupled performance of all the sub-systems that conform to a floating wind concept: wind turbines, mooring systems or floating platform hydrodynamic responses. Few experimental tests had been conducted, and no real-time hybrid modelling had been performed in these tests. In this paper, an open-source experimental dataset is presented, and the dynamics of a concrete spar-based floating concept withstanding the IEA 15-MW Reference Wind Turbine are investigated. An experiment on a 1:55 scale concrete-based spar (WindCrete) at IHCantabria's ocean basin was developed by means of the hardware-in-the-loop technique under the COREWIND EU-H2020 project. The available data consist of the hydrodynamic characterizations and seakeeping tests with a total of 57 available datasets. The proper behaviour of the WindCrete platform and the designed mooring system is confirmed by the maximum motion values and accelerations below the established acceptance criteria defined. Moreover, this test program identifies the importance of the wind loads over the dynamic performance of the concept. • Floating offshore wind platform dynamics. • On the importance of coupled tests over FOWT performance analysis. • Real-time Hybrid modelling of FOWT. [ABSTRACT FROM AUTHOR]

• Progressive failure process is considered in deformation safety diagnosis. • HM is improved to calibrate elastic state and shows satisfactory performances. • ABC is improved by integrating multiple strategies to promote HM performance. • Progressive diagnosis criteria have rigorous probabilistic and physical meanings. • The methodology realizes the 5-level safety control for in-service high arch dam. Conventional methods cannot effectively characterize the progressive failure process, which limits their use in the deformation safety diagnosis of in-service high arch dam. This work proposes a novel method utilizing the frontier theories of mathematics, mechanics, and dam safety monitoring. First, considering the implicit assumption of traditional method, hybrid model (HM) is improved to calibrate the elastic deformation state of high arch dam. Second, the adaptive proportion selection and the neighborhood search strategy are applied to improve artificial bee colony (ABC) algorithm. The undetermined parameters of HM are optimized by the improved ABC. Third, the dangerous degrees of single-point deformation and multi-point deformation are characterized using HM analysis results and information entropy. On this basis, the progressive diagnosis criteria are constructed based on probability principle. Finally, two case studies are conducted to validate the proposed methodology. The analysis results demonstrate that the performance of HM is better than that of the statistical model (SM); the improved ABC promotes the HM performance; and the progressive diagnosis criteria compared with the traditional confidence interval criteria have stricter probabilistic and physical meanings. The 5-level safety control is realized, promoting the emergency response ability of high arch dam in operating condition. [ABSTRACT FROM AUTHOR]

In the context of climate change, droughts, increasingly frequent and severe, necessitate effective monitoring. Existing methods, such as drought indices and data-driven models, face important limitations. Drought indices are built on prior expert knowledge but lack calibration based on actual drought events, while data-driven models prioritize goodness of fit over real event identification, undermining their credibility and generalization, and also struggling to generalize from regional to large-scale contexts. To address these challenges, here we introduce a hybrid machine learning framework for time series that combines domain knowledge and observational data in a variational recurrent neural network. The network models the joint distribution of total precipitation, air temperature, and real drought events, providing accurate predictions and uncertainty estimates. Extensive experiments focusing on a wide range of European drought events from 2011 to 2018 consistently show that our hybrid model surpasses both drought indices and data-driven models in terms of accuracy in drought detection, underlining its effectiveness, robustness, and stability. Our model achieves the best ROC-AUC (%) results in Afghanistan (79.7 ± 0.5), Italy (84.3 ± 0.6), Russia (89.4 ± 0.2), Europe-0 (84.3 ± 0.1), and Europe-1 (82.8 ± 0.4), effectively capturing the starting and ending times of drought events with lower uncertainty, and also generalizing better for unseen locations. • We propose DK-VRN, a novel hybrid framework for drought monitoring. • DK-VRN takes ECVs as input and integrates multi-scalar SPEI as domain knowledge prior. • DK-VRN is end-to-end supervised by historical drought events. • DK-VRN can detect droughts over diverse timeframes and across continental scales. • Experiments on European events from 2011 to 2018 show the effectiveness of DK-VRN. [ABSTRACT FROM AUTHOR]

Streamflow simulation is crucial for flood mitigation, ecological protection, and water resource planning. Process-based hydrological models and machine learning algorithms are the mainstream tools for streamflow simulation. However, their inherent limitations, such as time-consuming and large data requirements, make achieving high-precision simulations challenging. This study developed a hybrid approach to simultaneously improve the accuracy and computational efficiency of streamflow simulation, which integrates Block-wise use of the TOPMODEL (BTOP) model into the eXtreme Gradient Boosting (XGBoost), i.e., BTOP_XGB. In this approach, BTOP generates simulated streamflow using the Latin hypercube sampling algorithm instead of the time-consuming calibration algorithms to reduce computational costs. Then, XGBoost combines BTOP simulated streamflow with multi-source data to reduce simulation errors. In which, serval input variable selection algorithms are employed to choose relevant inputs and remove redundant information for model. The hybrid approach is validated and compared with a standalone model at three hydrological stations in the Jialing River basin, China. The results show that the performance of BTOP_XGB is significantly better than the BTOP and XGBoost models. The NSE of BTOP_XGB at Beibei, Xiaoheba, and Luoduxi stations increases by 54%, 21%, and 83%, respectively. Meanwhile, the computational time of BTOP_XGB is saved by >90% compared to the original calibrated BTOP. BTOP_XGB is less affected by parameter sample sizes and data amounts, demonstrating the robustness of the hybrid model. This study simplifies the complexity of the hydrological model and enhances the stability of machine learning, jointly improving the reliability of streamflow simulation. The hybrid approach provides a potential shortcut for streamflow simulation over basins with large areas or limited observed data. • A hybrid approach is proposed by combining the hydrological model (BTOP) with machine learning (XGBoost) and multi-source data. • The hybrid approach significantly improves the accuracy of streamflow simulation. • The Latin hypercube sampling algorithm reduce the computing time of original BTOP by >90%. • Input variable selection algorithms can reduce data redundancy and choose more relevant information for the hybrid model. • The hybrid approach is less affected by the data volume and parameter sample sizes, providing a potential shortcut to simulate streamflow in data-scarce basins. [ABSTRACT FROM AUTHOR]

Soil total nitrogen is one of the most important basic indicators for fertiliser decision making, but tens of millions of soil total nitrogen sampling data have been accumulated, forming a huge database. In this large database, there is a large amount of anomalous data, which can interfere with data analysis, affect the construction of spatial interpolation and prediction models, and then affect the accuracy of nutrient management decisions. The traditional method of identifying soil total nitrogen anomalies based on boxplots suffers from the problems of not being able to identify local anomalies, which can easily lead to misclassification of soil total nitrogen data anomalies, and the detection efficiency is not high. We propose a method to identify soil total nitrogen outliers by combining the Isolation Forest algorithm and local spatial autocorrelation analysis, which can simultaneously detect global and local outliers from large amounts of data and combine organic matter as an auxiliary indicator in the spatial analysis to help judge local outliers. Finally, the results of global and local anomalies were combined to provide a comprehensive assessment of the soil nitrogen data, avoiding the misjudgement or omission of judgement that can occur when using a single method. Using 25,930 soil test data from Yunnan Province in 2009 as an example, we compared and analysed the typical boxplot method and the unsupervised OneClassSVM method and evaluated the performance of each method in terms of correct detection rate, false positive rate and false negative rate. The results show that the proposed method has a correct detection rate (TR) of 99.97%, a false positive rate (FPR) of 8.06% and a false negative rate (FNR) of 0.01% on the data, which shows high validity and accuracy; it is also comparable to the independent isolated forests (FNR = 4.76%), boxplot (FNR = 3.90%) and OneClassSVM (FNR = 4.77%), and the false negative rate is reduced by 4.75%, 3.89% and 4.76%, respectively.

This paper presents an efficient flood early warning system developed for the city of Mandra, Greece which experienced a devastating flood event in November 2017 resulting in significant loss of life. The location is of particular interest due to both its small-sized water basin (20 km2 upstream of the studied cross-section), necessitating a rapid response time for effective flood warning calculations, and the lack of hydrometric data. To address the first issue, a database of pre-simulated flooding events with a 2D hydrodynamic model corresponding to synthetic precipitations with different return periods was established. To address the latter issue, the hydrological model was calibrated using qualitative information collected after the catastrophic event, compensating for the lack of hydrometric data. The case study demonstrates the establishment of a hybrid (online–offline) flood early warning system in data-scarce environments. By utilizing pre-simulated events and qualitative information, the system provides valuable insights for flood forecasting and aids in decision-making processes. This approach can be applied to other similar locations with limited data availability, contributing to improved flood management strategies and enhanced community resilience.

Prognostics and health management (PHM) of systems usually depends on appropriate prior knowledge and sufficient condition monitoring (CM) data on critical components’ degradation process to appropriately estimate the remaining useful life (RUL). A failure of complex or critical systems such as heating, ventilation, and air conditioning (HVAC) systems installed in a passenger train carriage may adversely affect people or the environment. Critical systems must meet restrictive regulations and standards, and this usually results in an early replacement of components. Therefore, the CM datasets lack data on advanced stages of degradation, and this has a significant impact on developing robust diagnostics and prognostics processes; therefore, it is difficult to find PHM implemented in HVAC systems. This paper proposes a methodology for implementing a hybrid model-based approach (HyMA) to overcome the limited representativeness of the training dataset for developing a prognostic model. The proposed methodology is evaluated building an HyMA which fuses information from a physics-based model with a deep learning algorithm to implement a prognostics process for a complex and critical system. The physics-based model of the HVAC system is used to generate run-to-failure data. This model is built and validated using information and data on the real asset; the failures are modelled according to expert knowledge and an experimental test to evaluate the behaviour of the HVAC system while working, with the air filter at different levels of degradation. In addition to using the sensors located in the real system, we model virtual sensors to observe parameters related to system components’ health. The run-to-failure datasets generated are normalized and directly used as inputs to a deep convolutional neural network (CNN) for RUL estimation. The effectiveness of the proposed methodology and approach is evaluated on datasets containing the air filter’s run-to-failure data. The experimental results show remarkable accuracy in the RUL estimation, thereby suggesting the proposed HyMA and methodology offer a promising approach for PHM.

Universal differential equations are a concept in scientific machine learning that leverages the potential of the universal approximator theorem and the physical knowledge of a given system. Creating this level of hybridization within a stiff partial differential equation system is a challenge. On the other hand, adsorption phenomenological models have sink/source terms that describe the adsorption equilibrium through a well‐known simplified model (e.g., Langmuir; Sips; and Brunauer, Emmet, Teller [BET]). These suitable mechanistic assumptions are identified through experiments, providing the parameters of the sink/source model. However, these mechanistic assumptions are a simplification of the system phenomenology. Therefore, the resulting model is limited by its premises. In this scenario, the universal ordinary differential equations (UODE) is presented as an approach that conciliates the potential of artificial neural networks to learn given phenomena without conceptual simplifications. On the other hand, keeping into consideration the system physics. This work proposes a UODE system to solve the multicomponent separation by adsorption in a fixed bed column. Experimental data is used to identify the hybrid model. The required amount of data used in the model identification demonstrates that hybrid models can use a few data points to precisely describe the system. Furthermore, the obtained model can describe competitive adsorption with higher precision than the Langmuir model. [ABSTRACT FROM AUTHOR]

Mammalian cell cultures make a significant contribution to the pharmaceutical industry. They produce many of the biopharmaceuticals obtaining FDA-approval each year. Motivated by quality-by-design principles, various modelling methodologies are frequently trialled to gain insight into these bioprocesses. However, these systems are highly complex and uncertain, involving dynamics at different scales, both in time and space, making them challenging to model in a comprehensive and fully mechanistic manner. This study develops a machine-learning-supported multiscale modelling framework of cell cultures, linking the macroscale bioprocess dynamics to the microscale metabolic flux distribution. As a relevant biopharmaceutical case study, we consider the production of Trastuzumab by Chinese Hamster Ovary (CHO) cells in batch. A macroscale hybrid model is constructed by integrating macro-kinetic and machine-learning approaches. Enzyme-constrained Dynamic Metabolic Flux Analysis (ecDMFA) is adopted to calculate flux distributions based on the dynamic predictions of the hybrid model. Uncertainty estimation of the multiscale model is conducted through bootstrapping. Judging from experimental data, our hybrid model can reduce the modelling error of the macroscale dynamics to 8.0%; a 70% reduction from the purely mechanistic model. In addition, the predicted dynamic flux distribution aligns with observations seen in literature, highlighting important metabolic changes throughout the process. Model uncertainty is maintained at a low level, demonstrating the trustworthiness of the predictions. Overall, our comprehensive modelling framework has the potential to facilitate the development of digital twins in the biopharmaceutical industry., Competing Interests: Declaration of competing interest The authors declare that they have no competing interest., (Copyright © 2024 The Authors. Published by Elsevier Inc. All rights reserved.)

Data-driven process monitoring is an important tool to ensure safe production and smooth operation. Generally, implicit information can be mined through data processing and analysis algorithms to detect process disturbances on the basis of historical production data. In industrial practice, signals with different sources of disturbance show different distribution patterns along with the time domain and frequency domain, that is, noise and pulse-type changes are usually contained in the high-frequency portion while most process dynamic is contained in the low-frequency portion. However, feature extraction is usually implemented at a single scale in traditional multivariate statistical algorithms. With this concern, a novel multi-scale process monitoring method is proposed in this work, by which wavelet packet decomposition is first employed for time-frequency analysis. After decomposition, multivariate statistical models are established for each scale to construct process statistics. For the high-frequency part, the classical principal component analysis (PCA) algorithm is adopted to construct squared prediction error (SPE) and Hotelling T2(T2) statistics. While for the low-frequency part, the slow feature analysis (SFA) algorithm is adopted to construct T2, Te2, S2 and Se2 statistics for the extraction of the long-term slowly changing trend. Then the monitoring statistics, obtained from each method at different scales, are integrated by a support vector data description (SVDD) method to give a final fault detection decision. The performance of the proposed method is verified on the benchmark Tennessee Eastman Process (TEP) and an industrial continuous catalytic reforming heat exchange unit by comparing with related multivariate statistical methods, which only focus on a single scale.

Integrated modelling of biological systems is becoming a necessity for constructing models containing the major biochemical processes of such systems in order to obtain a holistic understanding of their dynamics and to elucidate emergent behaviours. Hybrid modelling methods are crucial to achieve integrated modelling of biological systems. This paper reviews currently popular hybrid modelling methods, developed for systems biology, mainly revealing why they are proposed, how they are formed from single modelling formalisms and how to simulate them. By doing this, we identify future research requirements regarding hybrid approaches for further promoting integrated modelling of biological systems. [ABSTRACT FROM AUTHOR]

The estimation of indoor thermal comfort and the associated occupant feedback in office buildings is important to provide satisfactory and safe working environments, enhance the productivity of personnel, and to reduce complaints. The assessment of thermal comfort is a difficult task due to many environmental, physiological, and cultural variables that influence occupants' thermal perception and the way they judge their working environment. Traditional physics-based methods for evaluating thermal comfort have shown shortcomings when compared to actual responses from the occupants due to the incapacity of these methods to incorporate information of various natures. In this paper, a hybrid approach based on machine learning and building dynamic simulation is presented for the prediction of indoor thermal comfort feedback in an office building in Le Bour-get-du-Lac, Chambéry, France. The office was equipped with Internet of Things (IoT) environmental sensors. Occupant feedback on thermal comfort was collected during an experimental campaign. A calibrated building energy model was created for the building. Various machine learning models were trained using information from the occupants, environmental data, and data extracted from the calibrated dynamic simulation model for the prediction of thermal comfort votes. When compared to traditional predictive approaches, the proposed method shows an increase in accuracy of about 25%. [ABSTRACT FROM AUTHOR]

Automation and technological innovations pushed manufacturing companies to integrated plant configurations, where sub-systems are highly intertwined, though easily reconfigurable thanks to modularization. Frequent reconfigurations change the way sub-systems interact amongst each other more often than in the past. However, in large manufacturing systems digital sub-system models may be still ran independently, limiting their support in decision-making. This work proposes a methodology for the hybrid digital modelling of large manufacturing systems, where hybrid stands for multi-technique modelling, to achieve: (i) reduction of modelling complexity, (ii) portability, (iii) optimal modelling choice, (iv) hybrid modelling integration. An industrial case study in the electrical equipment sector shows the validity of the proposed approach. [ABSTRACT FROM AUTHOR]