Hussein Abdeltawab
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Hussein Abdeltawab has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Yasser Abdel-Rady Ibrahim Mohamed
University of Alberta
21 shared publications - Amirhossein Dolatabadi
Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, Canada
5 shared publications - Mohammad Jadidbonab
Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, Canada
4 shared publications - Ahmed A. Shalaby
Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, Canada
3 shared publications - Hamed Moayyed
Department of Electrical and Computer Engineering, University of Porto, Porto, Portugal
3 shared publications - Ali Bidram
Department of Electrical and Computer Engineering, University of New Mexico, Albuquerque, NM, USA
1 shared publication
Latest Publications
- 3 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 20 (2) , 586-598
A Group-Relative DRL Strategy for Optimizing Multiarea Hydrogen-Integrated Power and Traffic Networks
This article presents a multiagent framework for optimal scheduling of multiarea hydrogen-integrated power and traffic (HPT) networks. The framework utilizes twin-delayed deep deterministic policy gradient (TD3) reinforcement learning enhanced by group-relative policy optimization (GRPO). The proposed multiarea GRPO TD3 (MGRPO-TD3) incorporates a Bayesian-tuned hybrid forecasting method based on a graph transformer network to capture spatial–temporal patterns and accurately predict traffic flows and hydrogen vehicle demands. Nonlinear physical and operational constraints are embedded. The multicritic, group-relative design improves robustness to disturbances, forecasting errors, and facility outages, while agent grouping reduces the state dimensionality and computational cost for large systems. On a Calgary-inspired three-area HPT network, the forecaster reduces traffic-flow mean absolute percentage error by up to 21.3% compared with graph and graph-transformer benchmarks. For the same network, the MGRPO-TD3 increases daily profit by about 25.6% relative to a multiagent proximal policy optimization baseline. The approach also delivers robust performance in realistic contingency scenarios: when a hydrogen charging station is out of service, profit loss is limited to 1.2%, compared with 14.5%–41.7% for competing methods. In addition, scalability is confirmed under expanded traffic and power conditions with a larger number of agents.
- 3 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 20 (2) , 586-598
A Group-Relative DRL Strategy for Optimizing Multiarea Hydrogen-Integrated Power and Traffic Networks
This article presents a multiagent framework for optimal scheduling of multiarea hydrogen-integrated power and traffic (HPT) networks. The framework utilizes twin-delayed deep deterministic policy gradient (TD3) reinforcement learning enhanced by group-relative policy optimization (GRPO). The proposed multiarea GRPO TD3 (MGRPO-TD3) incorporates a Bayesian-tuned hybrid forecasting method based on a graph transformer network to capture spatial–temporal patterns and accurately predict traffic flows and hydrogen vehicle demands. Nonlinear physical and operational constraints are embedded. The multicritic, group-relative design improves robustness to disturbances, forecasting errors, and facility outages, while agent grouping reduces the state dimensionality and computational cost for large systems. On a Calgary-inspired three-area HPT network, the forecaster reduces traffic-flow mean absolute percentage error by up to 21.3% compared with graph and graph-transformer benchmarks. For the same network, the MGRPO-TD3 increases daily profit by about 25.6% relative to a multiagent proximal policy optimization baseline. The approach also delivers robust performance in realistic contingency scenarios: when a hydrogen charging station is out of service, profit loss is limited to 1.2%, compared with 14.5%–41.7% for competing methods. In addition, scalability is confirmed under expanded traffic and power conditions with a larger number of agents.
- 23 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industry Applications
- Vol. PP (00939994) , 1-21
A DRL-Based Framework for Optimized Scheduling and Delivery in a Green Hydrogen Hub
This paper presents a novel model-free deep reinforcement learning (DRL) framework for optimal hydrogen production and delivery operations in a green hydrogen hub (GHH), where produced hydrogen is delivered to customer locations via a fleet of portable tanks. The problem formulation takes into account (1) realistic nonlinear physical constraints and characteristics of the GHH components, such as the proton exchange membrane (PEM) electrolyzer and storage system, and (2) hydrogen delivery challenges, including hydrogen demand and traffic flow uncertainty. The proposed framework adopts 1) a safe proximal policy optimization algorithm that embeds an interior-point log-barrier in the actor loss, which ensures hard operational constraint satisfaction during learning over high-dimensional continuous states and actions, and 2) a deep neural network-based forecasting algorithm that actively monitors traffic flows and travel times between the GHH and consumer locations and estimates hydrogen demand and renewable power generation. Moreover, the DRL method can track electrolyzer degradation by using the predicted remaining useful life of the PEM unit, thereby introducing more realistic conditions that account for the impact of the PEM unit's practical lifetime on GHH operation. The proposed framework effectively schedules the GHH and manages portable hydrogen tank distribution under uncertainties. Comparisons with model-based information-gap decision theory optimization and state-of-the-art DRL methods validate the effectiveness of the proposed framework across various conditions.
- 23 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industry Applications
- Vol. PP (00939994) , 1-21
A DRL-Based Framework for Optimized Scheduling and Delivery in a Green Hydrogen Hub
This paper presents a novel model-free deep reinforcement learning (DRL) framework for optimal hydrogen production and delivery operations in a green hydrogen hub (GHH), where produced hydrogen is delivered to customer locations via a fleet of portable tanks. The problem formulation takes into account (1) realistic nonlinear physical constraints and characteristics of the GHH components, such as the proton exchange membrane (PEM) electrolyzer and storage system, and (2) hydrogen delivery challenges, including hydrogen demand and traffic flow uncertainty. The proposed framework adopts 1) a safe proximal policy optimization algorithm that embeds an interior-point log-barrier in the actor loss, which ensures hard operational constraint satisfaction during learning over high-dimensional continuous states and actions, and 2) a deep neural network-based forecasting algorithm that actively monitors traffic flows and travel times between the GHH and consumer locations and estimates hydrogen demand and renewable power generation. Moreover, the DRL method can track electrolyzer degradation by using the predicted remaining useful life of the PEM unit, thereby introducing more realistic conditions that account for the impact of the PEM unit's practical lifetime on GHH operation. The proposed framework effectively schedules the GHH and manages portable hydrogen tank distribution under uncertainties. Comparisons with model-based information-gap decision theory optimization and state-of-the-art DRL methods validate the effectiveness of the proposed framework across various conditions.
- Open Access
- 7 January 2026
- Published by Elsevier in Journal of Power Sources
- Vol. 667
Accelerated battery state-of-health estimation using conditional autoencoders for intelligent frequency band selection
- 7 January 2026
- Published by Elsevier in Journal of Power Sources
- Vol. 667
Open AccessAccelerated battery state-of-health estimation using conditional autoencoders for intelligent frequency band selection
- Open Access
- 21 November 2025
- Published by Elsevier in Reliability Engineering & System Safety
- Vol. 268
Innovative defense strategies: Fusion deep learning approach to counter false data injection attacks in power systems
- 21 November 2025
- Published by Elsevier in Reliability Engineering & System Safety
- Vol. 268
Open AccessInnovative defense strategies: Fusion deep learning approach to counter false data injection attacks in power systems
- Open Access
- 10 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 193356-193376
Adaptive Federated Learning for Robust PV Power Forecasting Under Weather-Induced Concept Drift
Photovoltaic (PV) power forecasting is vital for renewable energy integration but remains vulnerable to concept drift caused by dynamic weather conditions such as temperature, irradiance, and wind. This paper proposes an Adaptive Federated Learning (AFL) framework that combines federated learning with adaptive retraining to manage both abrupt and gradual drifts in decentralized PV systems. AFL integrates dual drift detectors, Rényi divergence for abrupt shifts and Wasserstein distance for gradual trends, together with adaptive thresholding and selective client retraining to ensure timely and efficient adaptation. Evaluations on real-world PV datasets show that AFL achieves 6–9% lower RMSE and ≈30% shorter recovery time than leading federated baselines (FedAvg, FedProx, FedDyn, APFL, and FedCurv), while maintaining consistent performance across non-stationary periods. The approach also demonstrates R2 ≈ 0.91 under drift conditions and R2 ≈ 0.96 during cross-site transfer tests between a smart-building system in Portugal and a microgrid in Norway, confirming strong generalization to unseen environments. AFL operates efficiently on CPU-based edge devices (8–16 GB RAM) without GPU acceleration, enabling practical deployment in distributed solar networks and microgrid controllers. Overall, the framework provides a scalable and computation-efficient solution for real-world non-IID PV forecasting, capable of continuous adaptation under seasonal variability, anomalies, and evolving operational conditions.
- 10 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 193356-193376
Open AccessAdaptive Federated Learning for Robust PV Power Forecasting Under Weather-Induced Concept Drift
Photovoltaic (PV) power forecasting is vital for renewable energy integration but remains vulnerable to concept drift caused by dynamic weather conditions such as temperature, irradiance, and wind. This paper proposes an Adaptive Federated Learning (AFL) framework that combines federated learning with adaptive retraining to manage both abrupt and gradual drifts in decentralized PV systems. AFL integrates dual drift detectors, Rényi divergence for abrupt shifts and Wasserstein distance for gradual trends, together with adaptive thresholding and selective client retraining to ensure timely and efficient adaptation. Evaluations on real-world PV datasets show that AFL achieves 6–9% lower RMSE and ≈30% shorter recovery time than leading federated baselines (FedAvg, FedProx, FedDyn, APFL, and FedCurv), while maintaining consistent performance across non-stationary periods. The approach also demonstrates R2 ≈ 0.91 under drift conditions and R2 ≈ 0.96 during cross-site transfer tests between a smart-building system in Portugal and a microgrid in Norway, confirming strong generalization to unseen environments. AFL operates efficiently on CPU-based edge devices (8–16 GB RAM) without GPU acceleration, enabling practical deployment in distributed solar networks and microgrid controllers. Overall, the framework provides a scalable and computation-efficient solution for real-world non-IID PV forecasting, capable of continuous adaptation under seasonal variability, anomalies, and evolving operational conditions.
- 26 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 16 (6) , 5065-5082
Safe Deep Reinforcement Learning for Resilient Self-Proactive Distribution Grids Against Wildfires
With the growing risks and frequency of wildfires, power distribution systems (PDS) face significant challenges in maintaining reliability and security. Existing literature primarily focuses on post-event service restoration using stochastic optimization methods. Nevertheless, such approaches fall short in managing the dynamic and uncertain nature of wildfires, particularly when it comes to taking proactive measures to mitigate power outages. To address this problem, this paper introduces a wildfire smart resilience controller (WF-SRC) that utilizes a model-assisted safe Deep Reinforcement Learning (DRL) mechanism to reduce the impacts of wildfire-induced disruptions. The WF-SRC continuously monitors and analyzes both the status of the PDS and the spatiotemporal dynamics of wildfires, then executes preemptive actions to prevent wildfires from compromising distribution lines. These actions include optimally dispatching stationary and mobile distributed energy resources (DERs) that operate under a master-slave control scheme. While recent works assume full observability and formulate the PDS resilience problem as a Markov Decision Process (MDP), this approach leads to an impractically large state space and limited realism. In contrast, our approach models the problem as a Partially Observable Markov Decision Process (POMDP). This explicitly captures real-world sensing limitations, such as noisy measurements that arise during extreme events. The POMDP is tackled using an LSTM-TD3 DRL agent, enabling effective sequential decision-making in environments with incomplete information. Using real-world data from Alberta wildfires, simulation results demonstrate the effectiveness of the proposed solution in reducing wildfire impact, optimizing energy distribution, and enhancing robustness to uncertainties, ultimately contributing to a more resilient and proactive power grid.
- 26 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 16 (6) , 5065-5082
Safe Deep Reinforcement Learning for Resilient Self-Proactive Distribution Grids Against Wildfires
With the growing risks and frequency of wildfires, power distribution systems (PDS) face significant challenges in maintaining reliability and security. Existing literature primarily focuses on post-event service restoration using stochastic optimization methods. Nevertheless, such approaches fall short in managing the dynamic and uncertain nature of wildfires, particularly when it comes to taking proactive measures to mitigate power outages. To address this problem, this paper introduces a wildfire smart resilience controller (WF-SRC) that utilizes a model-assisted safe Deep Reinforcement Learning (DRL) mechanism to reduce the impacts of wildfire-induced disruptions. The WF-SRC continuously monitors and analyzes both the status of the PDS and the spatiotemporal dynamics of wildfires, then executes preemptive actions to prevent wildfires from compromising distribution lines. These actions include optimally dispatching stationary and mobile distributed energy resources (DERs) that operate under a master-slave control scheme. While recent works assume full observability and formulate the PDS resilience problem as a Markov Decision Process (MDP), this approach leads to an impractically large state space and limited realism. In contrast, our approach models the problem as a Partially Observable Markov Decision Process (POMDP). This explicitly captures real-world sensing limitations, such as noisy measurements that arise during extreme events. The POMDP is tackled using an LSTM-TD3 DRL agent, enabling effective sequential decision-making in environments with incomplete information. Using real-world data from Alberta wildfires, simulation results demonstrate the effectiveness of the proposed solution in reducing wildfire impact, optimizing energy distribution, and enhancing robustness to uncertainties, ultimately contributing to a more resilient and proactive power grid.
- 11 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 40 (4) , 2915-2929
Towards Resilient Self-Proactive Distribution Grids Against Wildfires: A Dual Rolling Horizon-Based Framework
Wildfires have recently posed a significant danger to the security and reliability of electrical power systems, prompting the development of innovative frameworks for enhanced resilience. Existing literature has primarily focused on service restoration using stochastic optimization models that lack the necessary dynamism to tackle uncertainties effectively. To address this problem, a Dual Rolling horizon optimization (DRHO) is utilized to devise a smart resilience controller (SRC) that mitigates distribution network outages caused by wildfire disruptions. The proposed SRC can dynamically monitor and analyze the spatiotemporal behaviors of wildfires, such as their intensity, arrival time, and binding pathways from their ignition sources to electrical equipment. In doing so, it can proactively take real-time corrective measures before the wildfire reaches power distribution lines. The primary objective is to minimize load shedding and operational costs through re-configuring the network and employing a mix of stationary and mobile distributed energy resources (MDERs) that operate under a master-slave control scheme. The effectiveness of the proposed DRHO-based SRC is validated through diverse case studies and is compared against scenario-based approaches. Considering real-world data of Alberta wildfires, simulation results demonstrate the proposed solution's robustness to uncertainties, significantly reducing power outages and ensuring enhanced resilience.
- 11 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 40 (4) , 2915-2929
Towards Resilient Self-Proactive Distribution Grids Against Wildfires: A Dual Rolling Horizon-Based Framework
Wildfires have recently posed a significant danger to the security and reliability of electrical power systems, prompting the development of innovative frameworks for enhanced resilience. Existing literature has primarily focused on service restoration using stochastic optimization models that lack the necessary dynamism to tackle uncertainties effectively. To address this problem, a Dual Rolling horizon optimization (DRHO) is utilized to devise a smart resilience controller (SRC) that mitigates distribution network outages caused by wildfire disruptions. The proposed SRC can dynamically monitor and analyze the spatiotemporal behaviors of wildfires, such as their intensity, arrival time, and binding pathways from their ignition sources to electrical equipment. In doing so, it can proactively take real-time corrective measures before the wildfire reaches power distribution lines. The primary objective is to minimize load shedding and operational costs through re-configuring the network and employing a mix of stationary and mobile distributed energy resources (MDERs) that operate under a master-slave control scheme. The effectiveness of the proposed DRHO-based SRC is validated through diverse case studies and is compared against scenario-based approaches. Considering real-world data of Alberta wildfires, simulation results demonstrate the proposed solution's robustness to uncertainties, significantly reducing power outages and ensuring enhanced resilience.
- 19 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 18 (3) , 1581-1592
A Hybrid Traffic Flow Forecasting and Risk-Averse Decision Strategy for Hydrogen-Based Integrated Traffic and Power Networks
This article develops an operational framework for hydrogen microgrids integrated with traffic and power networks to optimize decision-making strategies. It tackles challenges in traffic flow prediction exacerbated by the rise of electric and hydrogen vehicles, which significantly affect power systems and hydrogen microgrids. We employ a risk-averse information gap decision theory to ensure secure operations under uncertain traffic conditions. Our framework utilizes a hybrid deep-learning forecasting method, combining a 1-D convolutional neural network and bidirectional long short-term memory to accurately predict traffic flow for origin–destination pairs in Edmonton, Canada. Enhanced by a Bayesian algorithm for hyperparameter tuning, this method improves prediction accuracy and operational efficiency. The framework also integrates operational strategies with urban travel plans to optimize charging for electric and hydrogen vehicles, thereby enhancing energy efficiency and supporting thermal demands. Validated in Edmonton's power and traffic networks, our framework enhances optimal charging, routing, and operation conditions, surpassing traditional methods to maintain secure operations during outages and improve the overall system robustness.
- 19 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 18 (3) , 1581-1592
A Hybrid Traffic Flow Forecasting and Risk-Averse Decision Strategy for Hydrogen-Based Integrated Traffic and Power Networks
This article develops an operational framework for hydrogen microgrids integrated with traffic and power networks to optimize decision-making strategies. It tackles challenges in traffic flow prediction exacerbated by the rise of electric and hydrogen vehicles, which significantly affect power systems and hydrogen microgrids. We employ a risk-averse information gap decision theory to ensure secure operations under uncertain traffic conditions. Our framework utilizes a hybrid deep-learning forecasting method, combining a 1-D convolutional neural network and bidirectional long short-term memory to accurately predict traffic flow for origin–destination pairs in Edmonton, Canada. Enhanced by a Bayesian algorithm for hyperparameter tuning, this method improves prediction accuracy and operational efficiency. The framework also integrates operational strategies with urban travel plans to optimize charging for electric and hydrogen vehicles, thereby enhancing energy efficiency and supporting thermal demands. Validated in Edmonton's power and traffic networks, our framework enhances optimal charging, routing, and operation conditions, surpassing traditional methods to maintain secure operations during outages and improve the overall system robustness.
- Open Access
- 27 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Intelligent Transportation Systems
- Vol. 5, 223-237
An Optimal Routing Framework for an Integrated Urban Power–Gas–Traffic Network
This paper develops a risk-averse-based framework for optimizing the operation of an integrated power, gas, and traffic (PGT) network with an application to a typical PGT network in downtown Edmonton, the forefront of Canada’s transition to electric vehicles and sustainable urban travel options. The developed non-probabilistic framework provides decision-makers with various secure options to avoid worst-case scenarios and promote social and environmental benefits. The integration of different energy systems allows operators to pursue optimal strategies in critical situations, such as facility outages, maintaining the system within a secure operational range without resorting to expensive workarounds. The proposed algorithm and integrated structure can select optimal travel routes to minimize gas-emission effects and locate charging options to reduce electric vehicle users’ travel time. It can mitigate challenges posed by distributed generator outages and roadway closures. The numerical results from implementing the framework on different case studies and the solar-based PGT network of Edmonton indicate its feasibility and effectiveness.
- 27 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Intelligent Transportation Systems
- Vol. 5, 223-237
Open AccessAn Optimal Routing Framework for an Integrated Urban Power–Gas–Traffic Network
This paper develops a risk-averse-based framework for optimizing the operation of an integrated power, gas, and traffic (PGT) network with an application to a typical PGT network in downtown Edmonton, the forefront of Canada’s transition to electric vehicles and sustainable urban travel options. The developed non-probabilistic framework provides decision-makers with various secure options to avoid worst-case scenarios and promote social and environmental benefits. The integration of different energy systems allows operators to pursue optimal strategies in critical situations, such as facility outages, maintaining the system within a secure operational range without resorting to expensive workarounds. The proposed algorithm and integrated structure can select optimal travel routes to minimize gas-emission effects and locate charging options to reduce electric vehicle users’ travel time. It can mitigate challenges posed by distributed generator outages and roadway closures. The numerical results from implementing the framework on different case studies and the solar-based PGT network of Edmonton indicate its feasibility and effectiveness.
- Open Access
- 10 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 7674-7688
SFNAS-DDPG: A Biomass-Based Energy Hub Dynamic Scheduling Approach via Connecting Supervised Federated Neural Architecture Search and Deep Deterministic Policy Gradient
The transition to a near-zero-emission power and energy industry for facing up to global warming issues is dominated by the use of renewable energy resources such as bioenergy and solar energy. When these resources are coordinated within an energy hub framework, the system’s flexibility is increased and dispatchable energy is provided by enhancing the share of renewable-dominated power. This paper proposes a dynamic scheduling framework for an energy hub with a biomass-solar hybrid renewable system. A hybrid forecasting model based on convolutional neural networks (CNNs) and Gated recurrent units (GRUs) is developed first to capture solar-related uncertainty sensibly, which will provide a great opportunity for the learning-based controller to determine an effective operation strategy in an optimal manner, especially on a cloudy-weather day. Then, a supervised federated neural architecture search (SFNAS) technique has been presented to eliminate the need for manual engineering of deep neural network models and the unnecessary computational burden associated with them. Finally, the deep deterministic policy gradient (DDPG), as an actor-critic deep reinforcement learning (DRL) methodology, enables the biomass-based energy hub to achieve cost-effective dynamic control strategies by addressing the decision-making problem as a highly dynamic continuous state-action model. The major conclusions of the numerical results show the effectiveness of the proposed SFNAS-DDPG method from average operating cost reduction up to 7.31% compared to the conventional DDPG model.
- 10 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 7674-7688
Open AccessSFNAS-DDPG: A Biomass-Based Energy Hub Dynamic Scheduling Approach via Connecting Supervised Federated Neural Architecture Search and Deep Deterministic Policy Gradient
The transition to a near-zero-emission power and energy industry for facing up to global warming issues is dominated by the use of renewable energy resources such as bioenergy and solar energy. When these resources are coordinated within an energy hub framework, the system’s flexibility is increased and dispatchable energy is provided by enhancing the share of renewable-dominated power. This paper proposes a dynamic scheduling framework for an energy hub with a biomass-solar hybrid renewable system. A hybrid forecasting model based on convolutional neural networks (CNNs) and Gated recurrent units (GRUs) is developed first to capture solar-related uncertainty sensibly, which will provide a great opportunity for the learning-based controller to determine an effective operation strategy in an optimal manner, especially on a cloudy-weather day. Then, a supervised federated neural architecture search (SFNAS) technique has been presented to eliminate the need for manual engineering of deep neural network models and the unnecessary computational burden associated with them. Finally, the deep deterministic policy gradient (DDPG), as an actor-critic deep reinforcement learning (DRL) methodology, enables the biomass-based energy hub to achieve cost-effective dynamic control strategies by addressing the decision-making problem as a highly dynamic continuous state-action model. The major conclusions of the numerical results show the effectiveness of the proposed SFNAS-DDPG method from average operating cost reduction up to 7.31% compared to the conventional DDPG model.