Akie Uehara
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Akie Uehara has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Tomonobu Senjyu
Department of Electrical and Electronics Engineering, University of the Ryukyus, Okinawa, Japan
7 shared publications - Masahiro Furukakoi
Department of Electrical and Electronic Engineering, Sanyo-Onoda City University, Sanyo-Onoda, Yamaguchi, Japan
4 shared publications - Jane Wambui Chege3 shared publications
Tomonobu Senjyu
3 shared publications- Adnan Akhunzada
University of Doha for Science and Technology
2 shared publications - Ahmad Sami Al-Shamayleh
Department of Data Science and Artificial Intelligence, Faculty of Information Technology, Al-Ahliyya Amman University, Amman 19111, Jordan
2 shared publications
Latest Publications
- Open Access
- 18 August 2026
- Published by Elsevier in Energy Reports
- Vol. 16
Surrogate model-based fast control of wind farms with consideration of voltage stability
- 18 August 2026
- Published by Elsevier in Energy Reports
- Vol. 16
Open AccessSurrogate model-based fast control of wind farms with consideration of voltage stability
- Open Access
- 10 February 2026
- Published by Elsevier in SSRN Electronic Journal
Long-term thermal power reduction in Okinawa’s island grid: Risk-aware multi-objective optimization of second-life batteries with degradation and circularity assessment
Okinawa, Japan, is an island with a grid heavily dependent on imported liquefied natural gas (LNG), oil, and coal. This exposes it to volatile global fuel prices and a high risk of greenhouse gas emissions. Simultaneously, the rapid adoption of electric vehicles in Japan creates challenges related to the disposal of retired batteries. This study examines the potential of second-life batteries (SLB) to reduce reliance on thermal power and to improve risk-aware planning for the Okinawa grid. A multi-objective optimization framework employing the Non-Dominated Sorting Genetic Algorithm III is applied to minimize total cost intensity (TCOI), operational emissions intensity (OEI), and Conditional Value-at-Risk at the 95\% confidence level (CVaR$_{0.95}$). Battery degradation and replacement behavior are explicitly modeled, and a post-optimization circularity assessment quantifies material mass flows, recovery potential, and avoided virgin material demand arising from battery second-life and replacement dynamics. Four scenarios are examined, including thermal-dominated systems and hybrid configurations with photovoltaic, wind, and battery storage using either SLBs or new batteries. The optimal SLB-based configuration achieved a TCOI of 184.21 USD/MWh, an OEI of 0.44 tCO$_2$/MWh, and a CVaR$_{0.95}$ of 85.29 USD/MWh. The scenario also attained a 7 percentage point higher circularity score relative to the new battery scenario and a 17\% thermal reduction relative to the baseline scenario. These findings demonstrate that second-life batteries provide a cost-effective and circular pathway for reducing thermal dependence in island power systems. The proposed framework is transferable to other islands and remote grids.
- 10 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessLong-term thermal power reduction in Okinawa’s island grid: Risk-aware multi-objective optimization of second-life batteries with degradation and circularity assessment
Okinawa, Japan, is an island with a grid heavily dependent on imported liquefied natural gas (LNG), oil, and coal. This exposes it to volatile global fuel prices and a high risk of greenhouse gas emissions. Simultaneously, the rapid adoption of electric vehicles in Japan creates challenges related to the disposal of retired batteries. This study examines the potential of second-life batteries (SLB) to reduce reliance on thermal power and to improve risk-aware planning for the Okinawa grid. A multi-objective optimization framework employing the Non-Dominated Sorting Genetic Algorithm III is applied to minimize total cost intensity (TCOI), operational emissions intensity (OEI), and Conditional Value-at-Risk at the 95\% confidence level (CVaR$_{0.95}$). Battery degradation and replacement behavior are explicitly modeled, and a post-optimization circularity assessment quantifies material mass flows, recovery potential, and avoided virgin material demand arising from battery second-life and replacement dynamics. Four scenarios are examined, including thermal-dominated systems and hybrid configurations with photovoltaic, wind, and battery storage using either SLBs or new batteries. The optimal SLB-based configuration achieved a TCOI of 184.21 USD/MWh, an OEI of 0.44 tCO$_2$/MWh, and a CVaR$_{0.95}$ of 85.29 USD/MWh. The scenario also attained a 7 percentage point higher circularity score relative to the new battery scenario and a 17\% thermal reduction relative to the baseline scenario. These findings demonstrate that second-life batteries provide a cost-effective and circular pathway for reducing thermal dependence in island power systems. The proposed framework is transferable to other islands and remote grids.
- Open Access
- 31 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Power Electronics
- Vol. 6, 1896-1909
Indirect Adaptive Polynomial Wavelet-Based Neuro-Fuzzy Controller for STATCOM-Equipped Power Systems
Modern power systems face growing stability challenges due to rising network complexity and dynamic operating conditions. Traditional control mechanisms often struggle to effectively mitigate Low-Frequency Oscillations (LFOs), underscoring the need for more advanced and adaptive damping strategies. Flexible AC Transmission Systems (FACTS), especially Static Synchronous Compensators (STATCOMs), have shown considerable promise in strengthening system stability under such challenging conditions. However, their performance is highly dependent on the quality of the Supplementary Damping Controller (SDC) strategy, and conventional methods may fall short under nonlinear and dynamic conditions. To tackle these issues, this paper presents a novel Indirect Adaptive Polynomial Wavelet-based Neuro-Fuzzy Control (ANFWC) framework designed to damp LFOs in STATCOM applications. The ANFWC includes three controllers, each employing a distinct Orthogonal Polynomial Wavelet-based Neural Network (PWNN) within an Adaptive Neuro-Fuzzy Inference System (ANFIS)-based Takagi-Sugeno-Kang (TSK) controller: the Legendre Wavelet-based Controller (ANFLWC), the Hermite Wavelet-based Controller (ANFHWC), and the Chebyshev Wavelet-based Controller (ANFCWC). These controllers enhance ANFIS learning and nonlinear mapping by leveraging PWNNs in the consequent layer. The performance of these controllers is evaluated through MATLAB simulations on the Single-Machine Infinite Bus (SMIB) and IEEE 9-bus Western System Coordinating Council (WSCC) test systems under various fault and disturbance conditions. Comparative analyses show that ANFLWC achieves the best performance, followed by ANFCWC and ANFHWC. All proposed controllers significantly outperform the conventional ANFIS-based TSK controller (ANFTSKC) and Lead-Lag Control (LLC), demonstrating the effectiveness of the ANFWC approach in improving power system damping and stability.
- 31 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Power Electronics
- Vol. 6, 1896-1909
Open AccessIndirect Adaptive Polynomial Wavelet-Based Neuro-Fuzzy Controller for STATCOM-Equipped Power Systems
Modern power systems face growing stability challenges due to rising network complexity and dynamic operating conditions. Traditional control mechanisms often struggle to effectively mitigate Low-Frequency Oscillations (LFOs), underscoring the need for more advanced and adaptive damping strategies. Flexible AC Transmission Systems (FACTS), especially Static Synchronous Compensators (STATCOMs), have shown considerable promise in strengthening system stability under such challenging conditions. However, their performance is highly dependent on the quality of the Supplementary Damping Controller (SDC) strategy, and conventional methods may fall short under nonlinear and dynamic conditions. To tackle these issues, this paper presents a novel Indirect Adaptive Polynomial Wavelet-based Neuro-Fuzzy Control (ANFWC) framework designed to damp LFOs in STATCOM applications. The ANFWC includes three controllers, each employing a distinct Orthogonal Polynomial Wavelet-based Neural Network (PWNN) within an Adaptive Neuro-Fuzzy Inference System (ANFIS)-based Takagi-Sugeno-Kang (TSK) controller: the Legendre Wavelet-based Controller (ANFLWC), the Hermite Wavelet-based Controller (ANFHWC), and the Chebyshev Wavelet-based Controller (ANFCWC). These controllers enhance ANFIS learning and nonlinear mapping by leveraging PWNNs in the consequent layer. The performance of these controllers is evaluated through MATLAB simulations on the Single-Machine Infinite Bus (SMIB) and IEEE 9-bus Western System Coordinating Council (WSCC) test systems under various fault and disturbance conditions. Comparative analyses show that ANFLWC achieves the best performance, followed by ANFCWC and ANFHWC. All proposed controllers significantly outperform the conventional ANFIS-based TSK controller (ANFTSKC) and Lead-Lag Control (LLC), demonstrating the effectiveness of the ANFWC approach in improving power system damping and stability.
- Open Access
- 19 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 167045-167060
Soft Computing-Based Damping Controllers With Online Parameters Tuning for Stability Enhancement of Power Systems
Power system stability continues to be a major challenge as modern grids grow more complex, uncertain, and increasingly reliant on renewable energy sources. This paper presents two new Neuro-Fuzzy controllers for Static Synchronous Compensators (STATCOMs): the Direct Chebyshev Wavelet-Based Neuro-Fuzzy Controller (DNF-CW), which adapts parameters online using fixed-structure rules, and the Indirect Chebyshev Wavelet-Based Neuro-Fuzzy Controller (IDNF-CW), which uses an online identifier to measure plant sensitivity. Chebyshev wavelet-based neural networks are utilized in the consequent part of both controllers to enable accurate local modeling and improved damping performance. The proposed methods are evaluated using a Single Machine Infinite Bus (SMIB) system and the IEEE 9-bus multi-machine system under a variety of fault and loading conditions. Benchmark comparisons include a conventional Indirect Adaptive Neuro-Fuzzy Takagi–Sugeno–Kang (IDNF-TSK) based controller, a configuration without STATCOM (No STATCOM), and a configuration with STATCOM but without auxiliary control (No Control). In the SMIB scenario, the IDNF-CW achieves a 40% reduction in settling time compared to the IDNF-TSK. In the more demanding multi-machine setup, the IDNF-CW restores stability within 3 seconds after a sequence of faults, outperforming DNF-CW and IDNF-TSK. Additionally, reductions of over 53% in the Integral of Time-weighted Absolute Error (ITAE) and 36% in the Integral of Absolute Error (IAE) are observed. These tests under multiple fault conditions and 10% measurement noise confirm stable operation, with overshoot limited to 3.57%–4.7% and minimal impact on settling time. These findings highlight the effectiveness of combining Chebyshev wavelets, adaptive control, and indirect architectures for enhancing power system stability.
- 19 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 167045-167060
Open AccessSoft Computing-Based Damping Controllers With Online Parameters Tuning for Stability Enhancement of Power Systems
Power system stability continues to be a major challenge as modern grids grow more complex, uncertain, and increasingly reliant on renewable energy sources. This paper presents two new Neuro-Fuzzy controllers for Static Synchronous Compensators (STATCOMs): the Direct Chebyshev Wavelet-Based Neuro-Fuzzy Controller (DNF-CW), which adapts parameters online using fixed-structure rules, and the Indirect Chebyshev Wavelet-Based Neuro-Fuzzy Controller (IDNF-CW), which uses an online identifier to measure plant sensitivity. Chebyshev wavelet-based neural networks are utilized in the consequent part of both controllers to enable accurate local modeling and improved damping performance. The proposed methods are evaluated using a Single Machine Infinite Bus (SMIB) system and the IEEE 9-bus multi-machine system under a variety of fault and loading conditions. Benchmark comparisons include a conventional Indirect Adaptive Neuro-Fuzzy Takagi–Sugeno–Kang (IDNF-TSK) based controller, a configuration without STATCOM (No STATCOM), and a configuration with STATCOM but without auxiliary control (No Control). In the SMIB scenario, the IDNF-CW achieves a 40% reduction in settling time compared to the IDNF-TSK. In the more demanding multi-machine setup, the IDNF-CW restores stability within 3 seconds after a sequence of faults, outperforming DNF-CW and IDNF-TSK. Additionally, reductions of over 53% in the Integral of Time-weighted Absolute Error (ITAE) and 36% in the Integral of Absolute Error (IAE) are observed. These tests under multiple fault conditions and 10% measurement noise confirm stable operation, with overshoot limited to 3.57%–4.7% and minimal impact on settling time. These findings highlight the effectiveness of combining Chebyshev wavelets, adaptive control, and indirect architectures for enhancing power system stability.
- Open Access
- 29 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 152669-152681
Multi-Objective Optimization of On-Grid Park and Ride EV Parking Station With Renewable Energy
In response to the growing need to reduce greenhouse gas emissions and achieve a decarbonized society, this study focuses on the optimal design of electric vehicle (EV) charging stations powered by renewable energy sources. Battery electric vehicles (BEVs), which emit no CO2 during operation, are particularly effective when combined with Park and Ride (P&R) systems that mitigate urban traffic congestion. This paper investigates the optimal operation and configuration of P&R BEV charging stations in Okinawa, Japan, comparing two power supply methods: on-grid and off-grid systems using photovoltaic (PV), wind power generation (WG), and battery energy storage systems (BESS). A multi-objective optimization model is developed to minimize the annual operating cost while meeting EV charging demand. To identify the optimal configuration, an exhaustive search of equipment unit combinations is conducted. The resulting pareto front reveals trade-offs between total annual EV charging energy and parking operator profit. In addition, three representative charging levels are analyzed in detail. Results show that on-grid systems achieve higher profit margins under moderate charging demand, leveraging grid power to maintain stable operations. In contrast, off-grid systems, while requiring higher initial investment, exhibit greater self-sufficiency and perform better in high demand scenarios or in remote areas with limited grid access. Thus, the optimal power supply method should be selected based on site-specific conditions, such as EV charging scale, regional infrastructure, and resilience requirements. Among the options, the on-grid system is identified as the most practical and economically viable solution under current technological and institutional conditions.
- 29 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 152669-152681
Open AccessMulti-Objective Optimization of On-Grid Park and Ride EV Parking Station With Renewable Energy
In response to the growing need to reduce greenhouse gas emissions and achieve a decarbonized society, this study focuses on the optimal design of electric vehicle (EV) charging stations powered by renewable energy sources. Battery electric vehicles (BEVs), which emit no CO2 during operation, are particularly effective when combined with Park and Ride (P&R) systems that mitigate urban traffic congestion. This paper investigates the optimal operation and configuration of P&R BEV charging stations in Okinawa, Japan, comparing two power supply methods: on-grid and off-grid systems using photovoltaic (PV), wind power generation (WG), and battery energy storage systems (BESS). A multi-objective optimization model is developed to minimize the annual operating cost while meeting EV charging demand. To identify the optimal configuration, an exhaustive search of equipment unit combinations is conducted. The resulting pareto front reveals trade-offs between total annual EV charging energy and parking operator profit. In addition, three representative charging levels are analyzed in detail. Results show that on-grid systems achieve higher profit margins under moderate charging demand, leveraging grid power to maintain stable operations. In contrast, off-grid systems, while requiring higher initial investment, exhibit greater self-sufficiency and perform better in high demand scenarios or in remote areas with limited grid access. Thus, the optimal power supply method should be selected based on site-specific conditions, such as EV charging scale, regional infrastructure, and resilience requirements. Among the options, the on-grid system is identified as the most practical and economically viable solution under current technological and institutional conditions.
- 25 July 2025
- Published by Elsevier in Energy Conversion and Management
- Vol. 344
Harnessing buoyancy for scalable and sustainable energy storage
- 25 July 2025
- Published by Elsevier in Energy Conversion and Management
- Vol. 344
Harnessing buoyancy for scalable and sustainable energy storage
- Open Access
- 29 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 80796-80820
Enhancing Smart Microgrid Resilience and Virtual Power Plant Profitability Through Hybrid IGWO-PSO Optimization With a Three-Phase Bidding Strategy
The increasing energy demand and rising fossil fuel prices are accelerating the transition to renewable energy, supported by government initiatives due to their environmental and economic advantages. However, challenges such as limited capacity and stability constraints hinder the widespread adoption of distributed energy resources (DERs). Virtual Power Plants (VPPs) enhance market participation by aggregating DERs, while electric vehicles (EVs) contribute to environmental sustainability by reducing emissions. Additionally, integrating distribution static compensators (DSTATCOMs) within VPPs improves microgrid stability and reactive power support. This study proposes a two-stage optimization approach to enhance network resilience and VPP profitability in a radial distribution network (RDN). The first stage focuses on minimizing resilience-related costs and energy not supplied (ENS) during natural disasters, while the second stage optimizes VPP profit using a three-phase bidding strategy, which includes the day-ahead market, real-time market, and overall market. A hybrid improved grey wolf optimization-particle swarm optimization (IGWO-PSO) algorithm is developed to solve this complex optimization problem. To demonstrate the effectiveness of the proposed approach, IGWO-PSO is compared with other hybrid optimization algorithms. Validation on a modified IEEE 33-bus RDN confirms that the proposed model enhances VPP placement and sizing, leading to improved economic, operational, and resilience metrics. Furthermore, the model accounts for uncertainties in load demand, renewable generation, energy prices, and equipment availability, ensuring a robust and adaptable energy management strategy.
- 29 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 80796-80820
Open AccessEnhancing Smart Microgrid Resilience and Virtual Power Plant Profitability Through Hybrid IGWO-PSO Optimization With a Three-Phase Bidding Strategy
The increasing energy demand and rising fossil fuel prices are accelerating the transition to renewable energy, supported by government initiatives due to their environmental and economic advantages. However, challenges such as limited capacity and stability constraints hinder the widespread adoption of distributed energy resources (DERs). Virtual Power Plants (VPPs) enhance market participation by aggregating DERs, while electric vehicles (EVs) contribute to environmental sustainability by reducing emissions. Additionally, integrating distribution static compensators (DSTATCOMs) within VPPs improves microgrid stability and reactive power support. This study proposes a two-stage optimization approach to enhance network resilience and VPP profitability in a radial distribution network (RDN). The first stage focuses on minimizing resilience-related costs and energy not supplied (ENS) during natural disasters, while the second stage optimizes VPP profit using a three-phase bidding strategy, which includes the day-ahead market, real-time market, and overall market. A hybrid improved grey wolf optimization-particle swarm optimization (IGWO-PSO) algorithm is developed to solve this complex optimization problem. To demonstrate the effectiveness of the proposed approach, IGWO-PSO is compared with other hybrid optimization algorithms. Validation on a modified IEEE 33-bus RDN confirms that the proposed model enhances VPP placement and sizing, leading to improved economic, operational, and resilience metrics. Furthermore, the model accounts for uncertainties in load demand, renewable generation, energy prices, and equipment availability, ensuring a robust and adaptable energy management strategy.