SEARCH · Engineering Papers

Results for “Reducing agent”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 487 records · Page 27

All sources NASA NTRS DOE OSTI

DOE OSTI2023

How good are learning-based control v.s. model-based control for load shifting? Investigations on a single zone building energy system

Both model predictive control (MPC) and deep reinforcement learning control (DRL) have been presented as a way to approximate the true optimality of a dynamic programming problem, and these two have shown significant operational cost saving potentials for building energy systems. Furthermore, there is still a lack of in-depth quantitative studies on their approximation levels to the true optimality, especially in the building energy domain. To fill in the gap, this paper provides a numerical framework that enables the evaluation of the optimality levels of different controllers for building energy systems. This framework is then used to comprehensively compare the optimal control performance of both MPC and DRL controllers with given computation budgets for a single zone fan coil unit system. Note the optimality is estimated based on a user-specific selection of trade-off weights among energy costs, thermal comfort and control slew rates. Compared with the best optimality we can find through expensive optimization simulations, the best DRL agent can maximally approximate the optimality by 96.54%, which outperforms the best MPC whose optimality level is 90.11%. However, due to the stochasticity, the DRL agent is only expected to approximate the optimality by 90.42%, which is almost equivalent to the best MPC. Except for Proximal Policy Optimization (PPO), all DRL agents can have a better approximation to the optimality than the best MPC, and are expected to have better approximation than the MPC with a prediction horizon of 32 steps (15 min per step). In terms of reducing energy cost and thermal discomfort, MPC can outperform the rule-based control (RBC) by 18.47%–25.44%. DRL can be expected to outperform RBC by 18.95%–25.65% ,and the best DRL control policy can outperform RBC by 20.29%–29.72%. Although the comparison of the optimality level is performed in a perfect setting, e.g., MPC assumes perfect models, and DRL assumes a perfect offline training process and online deployment process, this can shed insight on their capabilities of approximating to the original dynamic programming problem.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2022

Mechanisms of woody-plant mortality under rising drought, CO 2 and vapour pressure deficit

Drought-associated woody plant mortality may have increased over the past several decades, and is projected to increase in the future, impacting terrestrial climate forcing, biodiversity, and resource availability. The mechanisms underlying such mortality, however, are debated owing to interactions between multiple drivers and mechanisms. In this work, we synthesize knowledge of drought-related tree mortality under a warming and drying atmosphere with rising atmospheric CO 2 . Drought-associated mortality results from the depletion of pools of water and carbon within a plant and declines in their fluxes relative to demand by living tissues. These pools and fluxes are interdependent and underlay plant defenses against biotic agents. Death via failure to maintain a positive water balance is particularly dependent on soil-to-root conductance, capacitance, vulnerability to hydraulic failure, cuticular water losses, and dehydration tolerance, all of which could be exacerbated by reduced carbon supply rates to support cellular survival e.g. the carbon starvation process. The depletion of plant water and carbon pools accelerates under rising vapor pressure deficit; however, increasing CO 2 can mitigate these impacts. Advancing knowledge and reducing predictive uncertainties requires studies that integrate carbon, water, and defensive processes, and utilize a range of experimental and modeling approaches.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS2000

A Unified Approach to Model-Based Planning and Execution

Writing autonomous software is complex, requiring the coordination of functionally and technologically diverse software modules. System and mission engineers must rely on specialists familiar with the different software modules to translate requirements into application software. Also, each module often encodes the same requirement in different forms. The results are high costs and reduced reliability due to the difficulty of tracking discrepancies in these encodings. In this paper we describe a unified approach to planning and execution that we believe provides a unified representational and computational framework for an autonomous agent. We identify the four main components whose interplay provides the basis for the agent's autonomous behavior: the domain model, the plan database, the plan running module, and the planner modules. This representational and problem solving approach can be applied at all levels of the architecture of a complex agent, such as Remote Agent. In the rest of the paper we briefly describe the Remote Agent architecture. The new agent architecture proposed here aims at achieving the full Remote Agent functionality. We then give the fundamental ideas behind the new agent architecture and point out some implication of the structure of the architecture, mainly in the area of reactivity and interaction between reactive and deliberative decision making. We conclude with related work and current status.

Muscettola, Nicola↗

NASA NTRS2003

Biological countermeasures in space radiation health

Exposure to the types of ionizing radiation encountered during space travel may cause a number of health-related problems, but the primary concern is related to the increased risk of cancer induction in astronauts. The major types of radiation considered to be of importance during space travel are protons and particles of high atomic number and high energy (HZE particles). It is now clear that biological countermeasures can be used to prevent or reduce the levels of biological consequences resulting from exposure to protons or HZE particles, including the induction of cancer, immunosuppression and neurological defects caused by these types of ionizing radiation. Research related to the dietary additions of agents to minimize the risks of developing health-related problems which can result from exposure to space radiations is reviewed.

NASA Discipline Radiation Health↗

DOE OSTI2020

Agent-based modeling of supply disruptions in the global rare earths market

Several independent assessments have identified rare earth elements (REEs) as critical materials, notably neodymium (Nd), praseodymium (Pr), and dysprosium (Dy) used in permanent magnets. Factors affecting their criticality include expected growth in demand arising from their unique performance-enhancing properties in consumer, energy, and military applications and the supply risk associated with China's dominance in their production. We demonstrate the use of Argonne's Global Critical Materials (GCMat) agent-based model to explore the possible consequences to REE market dynamics of different types of regional supply disruptions including a temporary loss of production, shutdown of capacity, and diversion of supply. Results suggest that supply disruptions may foster earlier and more REE mine starts outside of China, although some of these mines may not be able to sustain operations post disruption. Further, price and associated market responses such as production, capacity, and demand tended to extend beyond the disruption period. Such market impacts in the magnet supply chain could affect the costs and availability of a number of emerging clean energy technology applications such as electric vehicles and wind turbines. In the future, GCMat could be used to evaluate the effectiveness of mitigation strategies - including recycling, conservation, product substitution, and diversification of supplies - on reducing the severity of disruptions in REE markets.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Potential biological control agents of Geosmithia morbida restrict fungal pathogen growth via mycoparasitism and antibiosis

Abstract Thousand cankers disease of Juglans (walnut) and Pterocarya (wingnut) spp. (Fagales: Juglandaceae) is caused by the fungal pathogen Geosmithia morbida Kolarík, Freeland, Utley, and Tisserat (Hypocreales: Bionectriaceae) and bark beetle pest/vector, Pityophthorus juglandis Blackman (Coleoptera: Curculionidae). To further the development of biological management strategies for thousand cankers disease, we assessed the ability of 14 endophytic Trichoderma (Hypocreales: Hypocreaceae) isolates and the commercially available isolate T. afroharzianum strain KRL-AG2 to inhibit the in vitro growth of three different G. morbida isolates via mycoparasitism and antibiosis. To identify factors that may affect field success of candidate biological control agents, we quantified the growth responses of Trichoderma spp. and the commercially available entomopathogenic fungus, Beauveria bassiana (Bals.-Criv.) Vuill. (Hypocreales: Cordycipitaceae) strain GHA, to the plant secondary metabolite and antimicrobial compound, juglone in vitro. A total of 12 Trichoderma isolates (from six different Trichoderma species) demonstrated antagonistic activity towards G. morbida in dual-plate assays. Juglone consistently reduced the growth of B. bassiana strain GHA and 14 of the 15 screened Trichoderma isolates in vitro. Additionally, one metabolite-producing Trichoderma isolate, TN4-47, completely inhibited the growth of all three G. morbida isolates across all tested metabolite concentrations and had comparatively greater tolerance to juglone compared to other Trichoderma isolates. Future lines of research should focus on characterizing the active antagonistic compound present in the metabolite filtrates, determine the mode of action of the active component(s), and elucidate how abiotic and biotic factors may influence the growth, persistence, and antagonistic activity of candidate biological control agents in planta .

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2023

Designing a transactive electric vehicle agent with customer’s participation preference

The proliferation of electric vehicles (EVs) and their inherent flexibility in charging timings make them an asset to improve grid performance. In contrast to direct control by a utility or autonomous price-based charging, the transactive control framework not only provides benefits to both grid and customers but also ensures customer autonomy. In this work, we design a transactive electric vehicle (TEV) agent that incorporates the EV owner’s willingness to trade-off between savings and amenity in form of a slider, where the EV owner’s amenity is characterized as vehicle readiness. Further, a privacy-preserving bidding formulation is proposed that also represents the customer’s transactive preference. A transactive market mechanism is discussed that integrates the TEV Agents into the local retail market and reconciles with the current day-ahead and real-time market structure. It is demonstrated that the proposed slider is able to provide a preferred trade-off between savings and amenity to individual customers. At the same time, the market mechanism is shown to successfully reduce both peak prices and peak demand. A comparative investigation of V1G and V2G technologies with respect to the battery prices is also discussed. It reveals that the V2G does not offer significant additional benefits with current battery prices, but could be promising if battery costs decline in the future.

33 ADVANCED PROPULSION SYSTEMS↗

DOE OSTI2023

Traffic Signal Optimization by Integrating Reinforcement Learning and Digital Twins

Machine learning (ML) methods, especially reinforcement learning (RL), have been widely considered for traffic signal optimization in intelligent transportation systems. Most of these ML methods are centralized, lacking in scalability and adaptability in large traffic networks. Further, it is challenging to train such ML models due to the lack of training platforms and/or the cost of deploying and training in a real traffic networks. This paper presents an approach for the integration of decentralized graph-based multi-agent reinforcement learning (DGMARL) with a Digital Twin (DT) to optimize traffic signals for the reduction of traffic congestion and network-wide fuel consumption related to stopping. Specifically, the DGMARL agents learn traffic state patterns and make decisions regarding traffic signal control with assistance from a Digital Twin module, which simulates and replicates the traffic behaviors of a real traffic network. The proposed approach was evaluated using PTV-Vissim [1], a microscopic traffic simulation platform. PTV-Vissim is also the simulation engine of the DT, enabling emulation and optimization of the traffic signals on the MLK Smart Corridor in Chattanooga, Tennessee. Compared to an actuated signal control baseline approach, experiment results show that Eco_PI, a developed performance measure capturing the impact of stops on fuel consumption, was reduced by 44.27% in a 24-hour and an average of 29.88% in a PM-peak-hour scenario.

Multi-Agent Reinforcement Learning, Digital Twin, ↗

DOE OSTI2022

The DOE E3SM Model Version 2: Overview of the Physical Model and Initial Model Evaluation

Abstract This work documents version two of the Department of Energy's Energy Exascale Earth System Model (E3SM). E3SMv2 is a significant evolution from its predecessor E3SMv1, resulting in a model that is nearly twice as fast and with a simulated climate that is improved in many metrics. We describe the physical climate model in its lower horizontal resolution configuration consisting of 110 km atmosphere, 165 km land, 0.5° river routing model, and an ocean and sea ice with mesh spacing varying between 60 km in the mid‐latitudes and 30 km at the equator and poles. The model performance is evaluated with Coupled Model Intercomparison Project Phase 6 Diagnosis, Evaluation, and Characterization of Klima simulations augmented with historical simulations as well as simulations to evaluate impacts of different forcing agents. The simulated climate has many realistic features of the climate system, with notable improvements in clouds and precipitation compared to E3SMv1. E3SMv1 suffered from an excessively high equilibrium climate sensitivity (ECS) of 5.3 K. In E3SMv2, ECS is reduced to 4.0 K which is now within the plausible range based on a recent World Climate Research Program assessment. However, a number of important biases remain including a weak Atlantic Meridional Overturning Circulation, deficiencies in the characteristics and spectral distribution of tropical atmospheric variability, and a significant underestimation of the observed warming in the second half of the historical period. An analysis of single‐forcing simulations indicates that correcting the historical temperature bias would require a substantial reduction in the magnitude of the aerosol‐related forcing.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

Economic Storage Size Optimization for Electric Vehicle Extreme-Fast Charging Stations

En-route charging infrastructure for electric vehicles is critical to support transportation needs. These charging stations are likely to have high loads and especially sharp peak loads given fast charging capabilities needed to meet transportation schedules. In order to reduce both strain on distribution grid infrastructure and charging station operational costs, many stations are likely to employ behind the meter storage. This paper demonstrates a behind the meter storage sizing optimization that employs an open-source agent-based vehicle behavior model (BEAM) to determine the best sizing across many scenarios. This optimization and analysis is novel in that it examines how storage size impacts not only charging station cost and peak load, but also vehicle queue times. The optimization is also applied across a wide analysis region with sufficient diversity and numbers to provide novel statistical analysis of optimal sizes.

Aka, Julius↗

DOE OSTI2023

A Physics-Based Data-Driven Approach for Modeling of Environmental Degradation in Elastomers

Abstract Elastomers are now commonly used in a number of industries, including aerospace, structure, transportation, shipbuilding, and automotive, due to their excellent workability, formability, and flexibility. During their activity, elastomers are subjected to harsh environmental conditions, which decreases their resilience. False predictions made early in their lives can have major financial and environmental implications. Elastomers’ performance and properties, such as strength, durability, and density, are influenced by chemical changes in these materials, known as degradation, which occurs over time. This process can alter the morphology of a polymer matrix as well as cause chain scission and cross-linking, resulting in different behaviors than that of the unaged material. To demonstrate the effect of thermaloxidative aging on the mechanical behavior of elastomers, several experimental and theoretical models have been proposed. In view of the large volume of experimental data available on micro-structural evolution in the course of aging, we propose a physics-based data-driven approach to overcome the shortcomings of both phenomenological and micro-mechanical models. This work presents a novel thermodynamically consistent, multiagent machine-learned model for predicting the constitutive behavior of cross-linked elastomers during environmental aging, such as thermo-oxidative and hydrolytic aging for various states of deformation. Single mechanism degradation changes the polymer matrix over time where it is causing chain scission, reduction of cross-links, and morphology change. To capture the idealized Mullins effect and permanent set due to the effect of single aging mechanisms on nonlinear mechanical responses of elastomers, we propose a data-driven model for simulating inelastic elements in a polymer matrix. By using a sequential order reduction, we were able to reduce the 3D stress-strain tensor mapping problem to a small number of super-constrained 1D mapping problems. To systematically classify such mapping problems into a few categories, an assembly of multiple replicated conditional neural network learning agents (L-agents) is used based on our recent work. Each category is represented by a different type of agent. The effect of deformation history, aging time, and aging temperature is captured by this model. The model is validated using a broad collection of data, ranging from our experimental results to data from the literature. In addition, thermodynamic consistency and frame independence are investigated. The most significant achievements of this model are its precision, simplicity, and prediction of inelasticity under various states of deformation. The model’s accuracy and simplicity make it a good option for commercial and industrial applications. Conveniently, due to the model modular nature, it can be expanded in the future to include viscoelasticity and non-isotropic formation for better precision.

Ghaderi, Aref↗

NASA NTRS1979

The solubility of olivine in basaltic liquids - An ionic model

A model is presented which enables the temperature at which olivine is in equilibrium with any alkali-depleted basaltic compound to be calculated to within + or - 30 C. It is noted that the error increases substantially when applied to terrestrial basalts which contain several weight percent alkalis. In addition the model predicts and quantifies the reduced activity of SiO4(4-) monomers due to increasing SiO2 concentrations in the melt. It is shown that the coordination of alumina in melts which precipitate olivine only appears to be dominantly octahedral, while titanium acts as a polmerizing agent by interconnecting previously isolated SiO4(4-) monomers. It is concluded that the model is sufficiently sensitive to show that there are small repulsive forces between Mg(2+) and calcium ions which are in association with normative diopside in the melt.

Herzberg, C. T.↗

NASA NTRS1992

Human-like agents with posture planning ability

Human body models are geometric structures which may be ultimately controlled by kinematically manipulating their joints, but for animation, it is desirable to control them in terms of task-level goals. We address a fundamental problem in achieving task-level postural goals: controlling massively redundant degrees of freedom. We reduce the degrees of freedom by introducing significant control points and vectors, e.g., pelvis forward vector, palm up vector, and torso up vector, etc. This reduced set of parameters are used to enumerate primitive motions and motion dependencies among them, and thus to select from a small set of alternative postures (e.g., bend vs. squat to lower shoulder height). A plan for a given goal is found by incrementally constructing a goal/constraint set based on the given goal, motion dependencies, collision avoidance requirements, and discovered failures. Global postures satisfying a given goal/constraint set are determined with the help of incremental mental simulation which uses a robust inverse kinematics algorithm. The contributions of the present work are: (1) There is no need to specify beforehand the final goal configuration, which is unrealistic for the human body, and (2) the degrees of freedom problem becomes easier by representing body configurations in terms of 'lumped' control parameters, that is, control points and vectors.

Jung, Moon R.↗

DOE OSTI2023

Treatment of cholinergic‐induced status epilepticus with polytherapy targeting GABA and glutamate receptors

Abstract Despite new antiseizure medications, the development of cholinergic‐induced refractory status epilepticus (RSE) continues to be a therapeutic challenge as pharmacoresistance to benzodiazepines and other antiseizure medications quickly develops. Studies conducted by Epilepsia . 2005;46:142 demonstrated that the initiation and maintenance of cholinergic‐induced RSE are associated with trafficking and inactivation of gamma‐aminobutyric acid A receptors (GABA A R) thought to contribute to the development of benzodiazepine pharmacoresistance. In addition, Dr. Wasterlain's laboratory reported that increased N‐methyl‐ d ‐aspartate receptors (NMDAR) and alpha‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid receptors (AMPAR) contribute to enhanced glutamatergic excitation ( Neurobiol Dis . 2013;54:225; Epilepsia . 2013;54:78). Thus, Dr. Wasterlain postulated that targeting both maladaptive responses of reduced inhibition and increased excitation that is associated with cholinergic‐induced RSE should improve therapeutic outcome. We currently review studies in several animal models of cholinergic‐induced RSE that demonstrate that benzodiazepine monotherapy has reduced efficacy when treatment is delayed and that polytherapy with drugs that include a benzodiazepine (eg midazolam and diazepam) to counter loss of inhibition, concurrent with an NMDA antagonist (eg ketamine) to reduce excitation provide improved efficacy. Improved efficacy with polytherapy against cholinergic‐induced seizure is demonstrated by reduction in (1) seizure severity, (2) epileptogenesis, and (3) neurodegeneration compared with monotherapy. Animal models reviewed include pilocarpine‐induced seizure in rats, organophosphorus nerve agent (OPNA)‐induced seizure in rats, and OPNA‐induced seizure in two mouse models: (1) carboxylesterase knockout (Es1 −/− ) mice which, similarly to humans, lack plasma carboxylesterase and (2) human acetylcholinesterase knock‐in carboxylesterase knockout (KIKO) mice. We also review studies showing that supplementing midazolam and ketamine with a third antiseizure medication (valproate or phenobarbital) that targets a nonbenzodiazepine site rapidly terminates RSE and provides further protection against cholinergic‐induced SE. Finally, we review studies on the benefits of simultaneous compared with sequential drug treatments and the clinical implications that lead us to predict improved efficacy of early combination drug therapies. The data generated from seminal rodent studies of efficacious treatment of cholinergic‐induced RSE conducted under Dr. Wasterlain's guidance suggest that future clinical trials should treat the inadequate inhibition and temper the excess excitation that characterize RSE and that early combination therapies may provide improved outcome over benzodiazepine monotherapy.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2020

Phenanthroline-strapped calix[4]pyrroles: anion receptors displaying affinity reversal as a function of solvent polarity

Calix[4]pyrroles 1 and 2, diametrically strapped with a phenanthroline via ester and amide linkages, respectively, have been synthesized as anion receptors. It was revealed by 1 H NMR spectroscopic analysis that receptors 1 and 2 possessing both hydrogen bonding donors and acceptors were able to bind the bicarbonate anion (as its tetraethylammonium (TEA + ) or sodium salt) in CDCl 3 , as well as in 15% aqueous DMSO with high affinity and selectivity over other anions. The amide-based receptor 2 contains additional potential hydrogen bonding donors relative to its ester-based congener 1. Nevertheless, in CDCl 3 receptor 1 was found to display a higher affinity for all test anions than receptor 2. In contrast, in 15% aqueous DMSO solution the affinities of receptor 2 for anions, in particular chloride, bicarbonate, and dihydrogen phosphate, were enhanced, whereas those of receptor 1 were reduced dramatically with no appreciable interaction being seen in the case of most test anions considered in this study. These reversals in selectivity and affinities underscore the importance of solvent in regulating the recognition features of seemingly similiar anion binding agents.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS1991

Glass formation and crystallization in high-temperature glass-ceramics and Si3N4

The softening of glassy grain boundaries in ceramic matrix composites and Si3N4 at high temperatures reduces mechanical strength and the upper-use temperature. By crystallizing this glass to a more refractory crystalline phase, a material which performs at higher temperatures may result. Three systems were examined: a cordierite composition with ZrO2 as a nucleating agent; celsian compositions; and yttrium silicate glasses both in bulk and intergranular in Si3N4. For the cordierite compositions, a series of metastable phases was obtained. The crystallization of these compositions was summarized in terms of metastable ternary isothermal sections. Zircon formed at the expense of ZrO2 and spinel. In SiC composites, the transformations were slower. In celsian, two polymorphs were crystallized. One phase, hexacelsian, which always crystallized, even when metastable, had an undesirable volume change. The other phase, celsian, was very difficult to crystallize. In yttrium silicate bulk glasses, similar in composition to the intergranular glass in Si3N4, a number of polymorphs of Y2Si2O7 were crystallized. The conditions under which these polymorphs formed are compared with crystallization in Si3N4.

Drummond, Charles H., III↗

DOE OSTI2022

ELASTOMERIC MICROVASCULAR SELF-HEALING MATERIALS

Damaged elastomeric diaphragms within pneumatic controllers used in the oil and gas industry lead to an unintended release of methane. Self-healing microvascular materials capable of healing various types of damage have been fabricated. These microvascular materials are designed to replace currently available commercial diaphragms found in pneumatic controllers and provide a solution to reduce unintended methane leaks. Poly(dimethylsiloxane) (PDMS) was used as the main matrix material, with additional testing conducted on polyurethane and flexibilised epoxy materials. Microvascular networks were implemented into the elastomeric membranes to act as ves- sels to deliver healing agents to the damaged areas. The complex, interdigitated channel networks were created using a 3D printed custom compounded filament composed of polylactic acid (PLA) and tin(II) oxalate. Two-part liquid casting polymers were poured around the channels into a mould to create samples with the scaffolding of the microvascular channels intact. Hollow microvascular networks were created by placing the samples within a vacuum oven at 250◦C which causes the thermal depolymerisation of PLA into its gaseous monomers. The manufactured materials were placed in test stands and pressurized using nitrogen gas to determine their healing and mechanical behaviour. Self-healing behaviour was demonstrated using PDMS matrix materials and healing chemistries within a pressure test cell. This test cell was designed to detect any damage to a sample by record- ing an outlet pressure. Damages including puncture holes, diagonal cuts and star-shaped central iv cuts were applied to the samples and a reduction in the outlet pressure was recorded for all sam- ples. A variation in channel spacing and diameter was studied to determine the optimal design of the microvascular network for self-healing performance. The mechanical performance of elastomeric membrane materials within a diaphragmatic pressurised stand was observed. Deformations and strains around the channels and channel inter- sections at the midplane of the materials were recorded using digital image correlation. Normal stresses were calculated using Hooke’s law and the material properties of the PDMS matrix. A reduction in leak rate, implying the success of self-healing, was recorded within manu- factured diaphragm samples tested within a commercial valve. PDMS is the favourable material when compared to polyurethane and flexibilised epoxy for creating

03 NATURAL GAS↗

DOE OSTI2023

Two-Stage Deep Reinforcement Learning for Distribution System Voltage Regulation and Peak Load Management

The growing integration of distributed solar photovoltaic (PV) in distribution systems could result in adverse effects during grid operation. This paper develops a two-agent soft actor critic-based deep reinforcement learning (SAC-DRL) solution to simultaneously control PV inverters and battery energy storage systems for voltage regulation and peak demand reduction. The novel two-stage framework, featured with two different control agents, is applied for daytime and nighttime operations to enhance control performance. Comparison results with other control methods on a real feeder in Western Colorado demonstrate that the proposed method can provide advanced voltage regulation with modest active power curtailment and reduce peak load demand from feeder's head.

deep reinforcement learning↗