Geosciences is a field of great societal relevance that requires solutions to several urgent problems facing our humanity and the planet. As geosciences enters the era of big data, machine learning ...(ML)—that has been widely successful in commercial domains—offers immense potential to contribute to problems in geosciences. However, geoscience applications introduce novel challenges for ML due to combinations of geoscience properties encountered in every problem, requiring novel research in machine learning. This article introduces researchers in the machine learning (ML) community to these challenges offered by geoscience problems and the opportunities that exist for advancing both machine learning and geosciences. We first highlight typical sources of geoscience data and describe their common properties. We then describe some of the common categories of geoscience problems where machine learning can play a role, discussing the challenges faced by existing ML methods and opportunities for novel ML research. We conclude by discussing some of the cross-cutting research themes in machine learning that are applicable across several geoscience problems, and the importance of a deep collaboration between machine learning and geosciences for synergistic advancements in both disciplines.
•Modeling of the physical state and functionality of deteriorating infrastructure.•Effects of deterioration on the ability of infrastructure to recover after shocks.•Incorporating the governing ...physical laws in the proposed formulation.•Emphasis on the modeling of potable water infrastructure.•Example: Reliability, resilience analysis of water infrastructure of Seaside OR.
The spatial and temporal extent of disruptions to services provided by infrastructure following disruptive events is directly related to the instantaneous state of the infrastructure and their post-disruption recovery. This paper develops a novel formulation to model the effects of infrastructure deterioration on their time-varying ability to recover after disruptive events. By unifying available models for deterioration and recovery, the paper proposes a general formulation to model the physical state and functionality of deteriorating infrastructure throughout its service life. The paper further develops resilience measures to quantify the temporal and spatial variations of infrastructure's ability to recover after disruptive events. The proposed formulation has a hierarchical structure that enables exploiting readily available data at the lower level of hierarchy to improve the prediction capability of models at the infrastructure level. Incorporating the governing physical laws in the proposed formulation also enables customizing the models to emulate the reality of infrastructure deterioration and recovery. While the formulation is general, the emphasis is on modeling potable water infrastructure as a case in which the deterioration of pipelines grows mostly undetectable until extensively developed. By the time the deterioration becomes visible, a substantial portion of the infrastructure service life has already been depleted, and costly repair or replacement would be inevitable. To illustrate, the proposed formulation has been implemented to model the time-varying reliability and resilience of the potable water infrastructure of the city of Seaside in Oregon, United States. The example highlights the effects of spatially varying exposure conditions and pipelines’ age on the reliability, functionality, recovery, and resilience of the potable water infrastructure.
Summary
Intelligent battery management systems (BMSs) require immediate and precise state of charge (SoC) estimation of the lithium‐ion cell for efficient use in electric vehicles (EVs). To solve the ...problem of real‐time SoC estimation for physics‐based lithium‐ion cell models in BMSs, rigorous physics‐based models are required to be reasonably order‐reduced. In this paper, an improved order‐reduced physics‐based model is proposed and validated with experimental data. Next, state equations of the reduced model are established by adopting finite difference approaches. Moreover, three reduced model‐based estimators including the extended Kalman filter (EKF), the uniform reaching law sliding‐mode observer (SMO‐URL), and the exponential reaching law sliding‐mode observer (SMO‐ERL) are developed to estimate battery SoC by using input current and external measured voltage. To demonstrate the effectiveness of the proposed SoC estimation algorithms, their performances are evaluated and compared via experimentation under the hybrid pulse power characteristic (HPPC) test. The results show that the proposed order‐reduced model‐based SMO‐ERL algorithm has faster convergence ability and higher estimation accuracy compared with the other two algorithms under incorrect SoC initial conditions.
Three model‐based estimators are designed to estimate lithium‐ion battery SoC for electric vehicles. To verify the proposed SoC estimation algorithms, their performances are evaluated and compared by experimentation.
•The relative capabilities knowledge-based, machine learning-based, and mechanistic surrogate models are compared to estimate the life-cycle seismic loss of concrete frame buildings based on their ...geometry and design parameters.•Two scenarios of incomplete (early design) and complete (design development) design information are considered to evaluate the impact of design data availability on surrogate models’ prediction.•A sequential framework is proposed to implement surrogate models with lower fidelity in earlier stages of design exploration, where each prior model guide parametrization of the posterior model.•Data-driven models are a versatile tool to demarcate seismic loss under incomplete design data. Mechanistic surrogate models can better extract the relationship between design parameters and seismic loss as the scale of analysis shrinks.
A performance-based early design must assess the life cycle performance of a sizable design space at low computational cost and limited data. This paper evaluates the relative capabilities of three surrogate modeling approaches to estimate seismic loss under complete and incomplete design information scenarios. Three surrogate models of knowledge-based (i.e., from prior published assessments), data-driven (i.e., support vector machine trained on a simulation-based building inventory), and physics-based (i.e., equivalent single-degree-of-freedom systems) are systematically used to estimate bounds on direct seismic loss for four hypothetical concrete office construction projects in Charleston, South Carolina. Subsequently, a framework is presented that implements these surrogate models as a sequence to explore design alternatives consistent with divergence-convergence cycles of early design exploration. The results show that all different surrogate models provide reasonable accuracy for the complete design information case, whereas data-driven models provided higher accuracy than the other models. For incomplete design information, the data-driven models demarcated the performance space and estimated the same median loss values as detailed loss analysis. In contrast, physics-based surrogate models were more accurate in capturing the relationship between loss and design parameters for smaller sets of design alternatives. Nevertheless, all different surrogate modeling techniques were inadequate to capture loss variability between different designs of the same geometry.
Despite peatlands' important feedbacks on the climate and global biogeochemical cycles, predicting their dynamics involves many uncertainties and an overwhelming variety of available models. This ...paper reviews the most widely used process-based models for simulating peatlands' dynamics, i.e., the exchanges of energy and mass (water, carbon, and nitrogen). ‘Peatlands’ here refers to mires, fens, bogs, and peat swamps both intact and degraded. Using a systematic search (involving 4900 articles), 45 models were selected that appeared at least twice in the literature. The models were classified into four categories: terrestrial ecosystem models (biogeochemical and global dynamic vegetation models, n = 21), hydrological models (n = 14), land surface models (n = 7), and eco-hydrological models (n = 3), 18 of which featured “peatland-specific” modules. By analysing their corresponding publications (n = 231), we identified their proven applicability domains (hydrology and carbon cycles dominated) for different peatland types and climate zones (northern bogs and fens dominated). The studies range in scale from small plots to global, and from single events to millennia. Following a FOSS (Free Open-Source Software) and FAIR (Findable, Accessible, Interoperable, Reusable) assessment, the number of models was reduced to 12. Then, we conducted a technical review of the approaches and associated challenges, as well as the basic aspects of each model, e.g., spatiotemporal resolution, input/output data format and modularity. Our review streamlines the process of model selection and highlights: (i) standardization and coordination are required for both data exchange and model calibration/validation to facilitate intercomparison studies; and (ii) there are overlaps in the models' scopes and approaches, making it imperative to fully optimize the strengths of existing models rather than creating redundant ones. In this regard, we provide a futuristic outlook for a ‘peatland community modelling platform’ and suggest an international peatland modelling intercomparison project.
Display omitted
•An extensive systematic search identified many process-based models being used for peatland dynamics.•Only about a quarter of the models meet the FOSS and FAIR criteria for an open-source and active community.•It is essential to standardize model calibration/validation and data sharing.•No one-size-fits-all model exists, but the models overlap in scope and approach.•A peatland community modelling platform can optimally exploit the strengths of existing models.
The ever expanding mobile consumer electronic market has accelerated the need for safe and efficient fast‐charging approaches that improve the overall speed of battery charging without hastened ...deterioration of the battery performance. Herein, the impact of a resource inexpensive, physics‐based, electrochemically optimized fast‐charging algorithm (charging time < 2 h) for mobile devices is investigated. A critical difference in the amount and morphology of lithium deposits on the anode for cells fast‐charged without an optimized algorithm is observed and found to be the main cause of capacity decay. An in‐depth study of the LiCoO2 cathode regions opposite to pronounced lithium deposits on the anode reveals a “mirroring” phenomenon, i.e., a frozen monoclinic phase, and inactivity to relithiation. In operando hard X‐ray absorption spectroscopy reveals that degraded spots on harvested cathodes seem to be activated again and participate in the intercalation process when lithiated at low rates from lithium foil counter electrodes. On the other hand, tests at higher C‐rates, closer to the actual fast‐charging rate, reveal only negligible oxidation state changes and therefore poor performance.
Impact on degradation between a nonoptimized and an electrochemical model‐based fast‐charging algorithm. Heavy Li deposits on the anode leading to degradation in opposing cathode regions (“mirroring”) are the consequence, if a non‐optimized, fast‐charging protocol is used for continuous charging, which can be avoided when using a model‐based fast‐charging protocol.
•We propose the two-stage physics-based Wiener process models.•An online stage division principle is developed to detect the change point.•An online remaining useful life prediction framework is ...constructed.•A dataset of wheel tread vibration demonstrates the superiority of the proposed method.
Due to most failure mechanisms, such as fatigue crack growth and fatigue spall, the degradation process of rotating machinery commonly exhibits two-stage features in engineering practice. Other minor factors are also the key issues affecting the health evolution process, including the component structure, assembly accuracy, and working environment. Ignoring such a mechanism may lead to imprecise in degradation modeling, life prognostic, and ultimately lead to safety risk. Besides, achieving high accuracy of prognostic emphasizes the influence of random effect in the degradation process. The contribution of this study lies in addressing this issue by proposing two-stage physics-based Wiener process models integrating: (a) fatigue crack mechanism and crack growth law, and (b) other minor factors. A general prognostic framework is formulated by jointly employing the online change point detection, parameter estimation, and remaining useful life (RUL) prediction, which has good statistic inference and applicability in two general nonlinear systems, i.e., power-law and exponential-law. A joint implement of offline two-step parameter estimation method and the online Bayesian update method is executed, making full advantage of historical and in-service data, based on which the RUL prediction transcends into an imperative PHM module. A practical case study on the vibration dataset of wheel treads demonstrates the practically implement ability of the proposed method in achieving high accuracy of RUL prediction.
This work presents the literature review of the methods used to model the heating, ventilation, and air conditioning (HVAC) systems. The model development is necessary for the study of the energy ...consumption of HVAC systems. Models are also required to simulate the different supervisory and local loop control strategies to improve the energy consumption efficiency. HVAC systems have complex structures consisting of heat and mass transfer equipment such as chiller, boiler, heating/cooling coils, and supply air ducts. HVAC systems also consist of several sensors and controllers for regulating the controllable variables such as zone temperature, supply air temperature, supply air fan speed, duct static pressure, and chilled water temperature at their set-points. To predict the energy consumption by the HVAC systems accurately, one needs to model the individual components either from the measured data or based on the knowledge of the underlying physical phenomenon. This results in three broad classes of the models known as data driven, physics based, and grey box models. In this paper, major data driven, physics based, and grey box modeling techniques reported in the recent literature are reviewed.
•Reviewed the major HVAC modeling techniques reported in the recent literature.•Physics based, data driven and grey box models are reviewed in detail.•Discussed the weaknesses and strengths of different modeling methods.•Comparison based on the robustness, ease of tuning and number of parameters etc.
A robot identification method which is based on the simulation modelling and analysis of the robots as a mechatronic system is presented in this study. The ever-increasing use of robotic systems in ...production lines necessitates the improvement of their accuracy. To this direction, the proposed approach leads to the enhancement of the accuracy using a closed loop estimation system of the robot’s dynamic parameters. Inaccurate positioning of the robot end effector mainly depends on the elastic behavior of the robot structure as well as the friction phenomena which occur in the motor’s gear box. This study presents an easy to use method for the identification of an industrial robot’s dynamic parameters based on physics-based simulation model. Using the robot motion data from both the digital and real robot, an intelligent algorithm is used to estimate the robot’s dynamic parameters and eventually adjust the control of the robot motion for achieving higher accuracy. The implementation procedure is analyzed in this work and a set of experiments is presented to validate the proposed methodology in an industrial robotic cell.
Digitalization of lithium-ion batteries can significantly advance the performance improvement of lithium-ion batteries by enabling smarter controlling strategies during operation and reducing risk ...and expenses in the design and development phase. Accurate physics-based models play a crucial role in the digitalization of lithium-ion batteries by providing an in-depth understanding of the system. Unfortunately, the high accuracy comes at the cost of increased computational cost preventing the employment of these models in real-time applications and for parametric design. Machine learning models have emerged as powerful tools that are increasingly being used in lithium-ion battery studies. Hybrid models can be developed by integrating physics-based models and machine learning algorithms providing high accuracy as well as computational efficiency. Therefore, this paper presents a comprehensive review of the current trends in integration of physics-based models and machine learning algorithms to accelerate the digitalization of lithium-ion batteries. Firstly, the current direction in explicit modeling methods and machine learning algorithms used in battery research are reviewed. Then a thorough investigation of contemporary hybrid models is presented addressing both battery design and development as well as real-time monitoring and control. The objective of this work is to provide details of hybrid methods including the various applications, type of employed models and machine learning algorithms, the architecture of hybrid models, and the outcome of the proposed models. The challenges and research gaps are discussed aiming to provide inspiration for future works in this field.
Display omitted
•Physics-based battery models are computationally expensive.•Machine learning models facilitate fast computation of battery states.•Machine learning models do not provide insight about the physical mechanisms.•Hybrid models allow battery monitoring, control and design.