Risk-Averse Goal-Oriented Optimal Experimental Design Using Non-linear Models
Explore the source record for details and available documents.
Jakeman, John Davis↗
SEARCH · Search NASA
Results for “risk aversion”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
An Inexact Trust-Region Algorithm for Nonsmooth Risk-Averse Optimization
Explore the source record for details and available documents.
Kouri, Drew Philip↗
An Inexact Trust-Region Algorithm for Nonsmooth Risk-Averse Optimization
Explore the source record for details and available documents.
Kouri, Drew Philip [Sandia National Laboratories (↗
An Inexact Trust-Region Algorithm for Nonsmooth Risk-Averse Optimization
Explore the source record for details and available documents.
Kouri, Drew Philip [Sandia National Laboratories (↗
Economics of Crop Rotations With and Without Carinata for Sustainable Aviation Fuel Production in the SE United States
In 2019, the aviation sector in the United States emitted 255 million metric tons of carbon dioxide (CO 2 ) emissions, i.e., about five percent of the total domestic CO 2 emissions from the energy sector. The sustainable aviation fuel (SAF) derived from carinata (Brassica carinata) could reduce CO 2 emissions of the aviation sector in the United States. Therefore, it is important to estimate changes in farm economics with and without carinata for ascertaining its production feasibility. In this context, the current study first assesses a combination of 12 popular rotations of corn, cotton, peanut, and soybean with winter crops of winter wheat and carinata in South Georgia over 4 years. Then, the net present values (NPVs) of 292 feasible cropping systems over 4 years are calculated. Finally, this study develops a risk model for ascertaining the probability distributions of NPVs for selected cropping systems subject to uncertainties related to prices and yields of summer and winter crops. Carinata in the corn-corn-soybean rotation has the highest NPV ($\$$2,996/ha). The least risky rotation is cotton-cotton-peanut, with a 58.9% probability of a positive NPV. Carinata can decrease the risk level of crop rotations by 8.1%, only if a contract price of $\$$440.9/t is offered. Therefore, a risk averse, risk neutral, or risk acceptant farmer can potentially include carinata in the rotation. Overall, carinata would increase the profitability of farm operations and decrease risk in the SE United States, and therefore, a high likelihood exists, that farmers would adopt it for meeting the growing demand for SAF in the United States.
09 BIOMASS FUELS↗
Industrial battery operation and utilization in the presence of electrical load uncertainty using Bayesian decision theory
Behind the meter battery storage is becoming increasing popular in all sectors, though enthusiasm has recently lagged in the industrial sector. Even though there may be many factors contributing to this including lack of innovation, prohibitive costs, and undesirable rate structures, a difficulty arises in accounting for uncertainty of electrical load in industrial facilities while still attempting to utilize battery storage as much as possible all while trying to achieve fiscal profitability. Here this study utilizes Gaussian process regression and Bayesian decision theory to organize load data and quantify electrical load uncertainty to properly and effectively discharge industrial battery storage. The study employs a simulation model to set battery load setpoints for the span of the utility billing period according to the degree of risk aversion. This combination of economic analysis according to utility billing period and utilization of degree of risk aversion to make decisions on the uncertainty of the data has not before been applied to battery storage. The method resulted in an annual average reduction of peak demand by 3.8 % at the lowest amount of savings and lowest risk aversion. The highest risk aversion resulted in an annual average reduction of peak demand of 7.5 %. The maximum reduction of peak load in any month was 13.8 % in the month of December with a relatively high risk aversion. With a the highest amount risk aversion tested, the model reduced demand ten of the twelve months of the year.
25 ENERGY STORAGE↗
Role of reinforcement learning for risk-based robust control of cyber-physical energy systems
Critical infrastructures such as cyber-physical energy systems (CPS-E) integrate information flow and physical operations that are vulnerable to natural and targeted failures. Safe, secure, and reliable operation and control of CPS-E is critical to ensure societal well-being and economic prosperity. Automated control is key for real-time operations and may be mathematically cast as a sequential decision-making problem under uncertainty. Emergence of data-driven techniques for decision making under uncertainty, such as reinforcement learning (RL), have led to promising advances for addressing sequential decision-making problems for risk-based robust CPS-E control. However, existing research challenges include understanding the applicability of RL methods across diverse CPS-E applications, addressing the effect of risk preferences across multiple RL methods, and development of open-source domain-aware simulation environments for RL experimentation within a CPS-E context. This article systematically analyzes the applicability of four types of RL methods (model-free, model-based, hybrid model-free and model-based, and hierarchical) for risk-based robust CPS-E control. Problem features and solution stability for the RL methods are also discussed. We demonstrate and compare the performance of multiple RL methods under different risk specifications (risk-averse, risk-neutral, and risk-seeking) through the development and application of an open-source simulation environment. Motivating numerical simulation examples include representative single-zone and multizone building control use cases. Finally, six key insights for future research and broader adoption of RL methods are identified, with specific emphasis on problem features, algorithmic explainability, and solution stability.
24 POWER TRANSMISSION AND DISTRIBUTION↗
A primal–dual algorithm for risk minimization
In this paper, we develop an algorithm to efficiently solve risk-averse optimization problems posed in reflexive Banach space. Such problems often arise in many practical applications as, e.g., optimization problems constrained by partial differential equations with uncertain inputs. Unfortunately, for many popular risk models including the coherent risk measures, the resulting risk-averse objective function is nonsmooth. Here, this lack of differentiability complicates the numerical approximation of the objective function as well as the numerical solution of the optimization problem. To address these challenges, we propose a primal–dual algorithm for solving large-scale nonsmooth risk-averse optimization problems. This algorithm is motivated by the classical method of multipliers and by epigraphical regularization of risk measures. As a result, the algorithm solves a sequence of smooth optimization problems using derivative-based methods. We prove convergence of the algorithm even when the subproblems are solved inexactly and conclude with numerical examples demonstrating the efficiency of our method.
97 MATHEMATICS AND COMPUTING↗
Navigating Epistemic Uncertainty in the Management of Flash Droughts
Abstract: Flash droughts, characterized by their rapid onset, sharply contrast with the typically gradual development of traditional droughts. These events are triggered by a combination of low rainfall and high evaporation rates, driven by elevated temperatures, making them particularly challenging to predict and prepare for. As a relatively new concept, flash drought is not well understood, which introduces significant epistemic uncertainties regarding their nature and detection methods. This under-detection hinders planners' ability to effectively manage these events. Despite these uncertainties, flash droughts can have significant impacts, raising the question: how can decision-makers prepare for such events given the current knowledge gaps? To address this, we propose a methodological framework aimed at enhancing flash drought preparedness by guiding the selection of appropriate indicators based on their detection capabilities and the decision-makers' level of risk aversion. Our approach involves evaluating six different flash drought indicators and analyzing the level of agreement among them. Additionally, we consider the decision-makers' risk aversion, distinguishing between those who require consensus across all methods (risk-takers) and those who act based on a single method's indication (risk-averse). The insights gained from this study offer a pathway towards more informed decision-making processes regarding flash droughts, potentially mitigating their adverse effects through better preparedness and response strategies.
climate resilience↗
Designing Cost‐Effective Carbon Payments to Induce Cellulosic Feedstock Production for Sustainable Aviation Fuel
Perennial bioenergy crops, such as miscanthus and switchgrass, and crop residues have the potential to scale up sustainable aviation fuel (SAF) production and mitigate carbon emissions. However, high establishment costs, delayed returns, and risk–return profiles that diverge from those of conventional crops can hinder incentives to adopt bioenergy crops. We develop an economic model that incorporates spatially varying joint yield and price distributions for the multiple crop choices a farmer faces and apply it to examine the incentives for risk-averse, present-biased, and credit-constrained farmers to produce cellulosic feedstocks under various biomass prices. We link this model to a biogeochemical model to quantify the spatially varying carbon mitigation benefits from these feedstocks in the rainfed region of the United States. We also analyze the cost-effectiveness of two carbon payment policies: annual and upfront. We find that risk-averse, present-biased, or credit-constrained farmers prefer to grow the lower-yielding but less risky switchgrass and harvest corn stover instead of the lower carbon, higher-yielding but riskier feedstock miscanthus, resulting in lower SAF production. Upfront carbon payments incentivize higher quantities of less carbon-intensive SAF production by risk-averse, credit-constrained, and present-biased farmers because they offset a part of the establishment costs of miscanthus. We also find that when farmers are credit-constrained, upfront payments are more cost-effective in terms of carbon mitigation per dollar spent. In contrast, annual payments are more cost-effective when farmers can access credit.
09 BIOMASS FUELS↗
Extreme Risk Mitigation in Reinforcement Learning using Extreme Value Theory
Risk-sensitive reinforcement learning (RL) has garnered significant attention in recent years due to the growing interest in deploying RL agents in real-world scenarios. A critical aspect of risk awareness involves modelling highly rare risk events (rewards) that could potentially lead to catastrophic outcomes. These infrequent occurrences present a formidable challenge for data-driven methods aiming to capture such risky events accurately. While risk-aware RL techniques do exist, they suffer from high variance estimation due to the inherent data scarcity. Our work proposes to enhance the resilience of RL agents when faced with very rare and risky events by focusing on refining the predictions of the extreme values predicted by the state-action value distribution. To achieve this, we formulate the extreme values of the state-action value function distribution as parameterized distributions, drawing inspiration from the principles of extreme value theory (EVT). We propose an extreme value theory based actor-critic approach, namely, Extreme Valued Actor-Critic (EVAC) which effectively addresses the issue of infrequent occurrence by leveraging EVT-based parameterization. Importantly, we theoretically demonstrate the advantages of employing these parameterized distributions in contrast to other risk-averse algorithms. Our evaluations show that the proposed method outperforms other risk averse RL algorithms on a diverse range of benchmark tasks, each encompassing distinct risk scenarios.
Wang, Yu↗
Designing a Robust MEA-Based Post-Combustion Carbon Capture Process with Capture Rate Guarantees
This work presents an application of the nonlinear two-stage robust optimization solver PyROS to the model-based design and operation of a monoethanolamine scrubbing process for CO<sub>2</sub> capture under epistemic uncertainty. Through this application, risk-averse process designs are successfully obtained for CO<sub>2</sub> capture targets ranging from 90% to over 99%. In particular, the risk-averse solutions for CO<sub>2</sub> capture targets of up to 98% are shown to be only marginally more expensive than their nominally optimal counterparts. Thus, the results demonstrate the utility of recently developed nonlinear robust optimization approaches for the solution of large-scale chemical process models under uncertainty.
20 FOSSIL-FUELED POWER PLANTS↗
Projected land ice contributions to twenty-first-century sea level rise
The land ice contribution to global mean sea level rise has not yet been predicted using ice sheet and glacier models for the latest set of socio-economic scenarios, nor using coordinated exploration of uncertainties arising from the various computer models involved. Two recent international projects generated a large suite of projections using multiple models, but primarily used previous-generation scenarios and climate models, and could not fully explore known uncertainties. Here we estimate probability distributions for these projections under the new scenarios using statistical emulation of the ice sheet and glacier models. In this work, we find that limiting global warming to 1.5 degrees Celsius would halve the land ice contribution to twenty-first-century sea level rise, relative to current emissions pledges. The median decreases from 25 to 13 centimetres sea level equivalent (SLE) by 2100, with glaciers responsible for half the sea level contribution. The projected Antarctic contribution does not show a clear response to the emissions scenario, owing to uncertainties in the competing processes of increasing ice loss and snowfall accumulation in a warming climate. However, under risk-averse (pessimistic) assumptions, Antarctic ice loss could be five times higher, increasing the median land ice contribution to 42 centimetres SLE under current policies and pledges, with the 95th percentile projection exceeding half a metre even under 1.5 degrees Celsius warming. This would severely limit the possibility of mitigating future coastal flooding. Given this large range (between 13 centimetres SLE using the main projections under 1.5 degrees Celsius warming and 42 centimetres SLE using risk-averse projections under current pledges), adaptation planning for twenty-first-century sea level rise must account for a factor-of-three uncertainty in the land ice contribution until climate policies and the Antarctic response are further constrained.
54 ENVIRONMENTAL SCIENCES↗
Extreme-scale stochastic optimization and simulation via learning-enhanced decomposition and parallelization (Final Technical Report)
Stochastic optimization and simulation models ubiquitously arise in designing and operating complex service/engineering systems. They can be extreme in scale due to high-dimensional data and decisions, and can also involve decisions made sequentially in response to newly revealed data, both causing significant computational challenge. The objective of this research is to explore a unified framework that integrates machine learning with discrete optimization and risk-averse modeling, to improve the efficiency of decomposition paradigms for stochastic optimization and simulations at extreme scale. The models we consider represent a broad class of complex decision-making problems, where 0-1 or continuous decisions are made before and/or after knowing multiple sources of uncertainties that could be correlated. We will employ machine learning methods to dynamically decide and prioritize computational procedures, including cut generation, branching, and bounding of the optimal objective. Furthermore, the research will shed new lights on the traditional decomposition algorithms for extreme-scale computing. Deliverables of the research include new modeling and computational methods for advancing the state-of-the-art research in optimization and simulation, bringing many relevant risk-averse, data-driven optimization problems in practice within the range of tractability. Examples include distributed computing server scheduling and sensor deployment for monitoring critical infrastructures. Success in this effort will enable progress in solving multiple extreme-scale problems in the complex system design and operations arising from DoE missions in energy, environment, and national security.
24 POWER TRANSMISSION AND DISTRIBUTION↗
Modular supply chain optimization considering demand uncertainty to manage risk
Supply chain under demand uncertainty has been a challenging problem due to increased competition and market volatility in modern markets. Flexibility in planning decisions makes modular manufacturing a promising way to address this problem. We report the problem of multiperiod process and supply chain network design is considered under demand uncertainty. A mixed integer two-stage stochastic programming problem is formulated with integer variables indicating the process design and continuous variables to represent the material flow in the supply chain. The problem is solved using a rolling horizon approach. Benders decomposition is used to reduce the computational complexity of the optimization problem. To promote risk-averse decisions, a downside risk measure is incorporated in the model. The results demonstrate the several advantages of modular designs in meeting product demands. A pareto-optimal curve for minimizing the objectives of expected cost and downside risk is obtained.
42 ENGINEERING↗
A robust offering strategy for wind producers considering uncertainties of demand response and wind power
This paper proposes a risk-constrained decision-making approach for a wind power producer participating in the day-ahead market. In the developed model, a flexible demand response trading scheme between the wind power producer and different customers is employed. Through the proposed demand response mechanism, the wind power producer is able to trade demand response resource internally with different customers, and then trade energy externally with the market to increase the expected profit and the wind energy utilization. The uncertainties in the wind power and demand response are modeled by using the information gap decision theory approach from risk averse (robust) and risk-seeking (opportunistic) perspectives. The objective of the robust model is to maximize the robust level while satisfying the desired profit, whereas the opportunistic model aims to evaluate the possibility of achieving windfall profits with favorable uncertainties. The overall offering strategy problem is modeled as a bi-objective mixed integer nonlinear programming, which is linearized by proper techniques and solved efficiently by using the normal boundary intersection technique. In this work, simulation results show that utilizing demand response resource to mitigate wind power deviations can increase a wind power producer's profit and reduce potential risks. In addition, the results demonstrate that the proposed bi-objective optimization approach enables the wind power producer to select appropriate offering decisions with respect to uncertainties.
17 WIND ENERGY↗
From Stewardship to Transformation: Toward a Nuclear Enterprise for the 21st Century
The erosion of the global security environment over the last two decades has been rapid and dramatic. It appears also to be accelerating. There are deep and widening concerns in the United States and among its allies and partners that deterrence too has eroded at both the conventional and nuclear levels. The Strategic Posture Commission captured the essence of the problem in its October 2023 report—the United States seeks to meet the challenges of complex multipolar nuclear rivalry in the 21 st century with a slimmed down version of a nuclear deterrent designed in the middle of the 20 th century. With great urgency, the Commission recommended that something more and/or different is needed than simply the replacement of legacy systems with modern variants. But the U.S. nuclear security enterprise was reshaped after the Cold War to support the stockpile stewardship mission—not renewed designed, engineering, and production at scale. It has long fallen short of the agility, flexibility, and adaptiveness periodically asked of it by national leaders. In recent years, however, the message of urgency has been clearly received in the enterprise. Its culture has begun to shift. It has also been received by those in Washington D.C. responsible for funding and guiding the enterprise. The result is an enterprise in transition. Stockpile stewardship has given way to stockpile transformation. Risk aversion is giving way to risk management. Innovation and adaptation have been embraced. Experience to date has yielded some important lessons about what works and what doesn’t in terms of accelerating and delivering the needed innovation and adaptation. It has revealed the enduring character of some challenges. But it has also revealed that a much more agile and effective enterprise is within the national reach—something that will pay important dividends for deterrence, assurance, strategic stability, and peace.
45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗
Data-Driven Compositional Optimization in Misspecified Regimes
With a manifold growth in the scale and intricacy of systems, the challenges of parametric misspecification become pronounced. These concerns are further exacerbated in compositional settings, which emerge in problems complicated by modeling risk and robustness. In “Data-Driven Compositional Optimization in Misspecified Regimes,” the authors consider the resolution of compositional stochastic optimization problems, plagued by parametric misspecification. In considering settings where such misspecification may be resolved via a parallel learning process, the authors develop schemes that can contend with diverse forms of risk, dynamics, and nonconvexity. They provide asymptotic and rate guarantees for unaccelerated and accelerated schemes for convex, strongly convex, and nonconvex problems in a two-level regime with extensions to the multilevel setting. Surprisingly, the nonasymptotic rate guarantees show no degradation from the rate statements obtained in a correctly specified regime and the schemes achieve optimal (or near-optimal) sample complexities for general T-level strongly convex and nonconvex compositional problems.
Business & Economics↗