Masafumi Miyatake
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Masafumi Miyatake has a total of 92 co-authors. Here are the top co-authors by number of shared publications:
Vanxay Phimmasone
Graduate School of Electrical and Electronic Engineering, Sophia University, Tokyo, Japan
6 shared publicationsHideyoshi Ko
Suzuka University of Medical Science
5 shared publicationsHyun Woo Lee
Electric Energy Saving Research Centre, Kyungnam University, Masa, Korea
5 shared publications- Natarajan Rajasekar
Vellore Institute of Technology
5 shared publications Yoshichika Noda
Sophia University Japan
5 shared publicationsYuta Kondo
Graduate School of Electrical and Electronic Engineering, Sophia University, Tokyo, Japan
5 shared publications
Latest Publications
- Open Access
- 14 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 78085-78109
Event-Driven, Condition-Aware MPPT Framework for Robust Tracking Under Partial Shading and Dynamic Irradiance
Partial shading conditions (PSC) introduce multiple peaks in photovoltaic P-V characteristics, causing conventional MPPT algorithms to converge prematurely at local maxima. Existing metaheuristic approaches apply computationally expensive global search regardless of environmental transition type, wasting computation on simple conditions or failing to converge during complex PSC. This paper proposes an Event-Driven, Condition-Aware MPPT Framework where a hierarchical Environmental Monitoring Engine (EME) classifies environmental transitions through complementary detection, PSC pattern recognition via panel irradiance, step change filtering via irradiance rate, and checkpoint tracking via validated state transitions, and selects a suitable optimization strategy accordingly: Confidence Adaptive Sweep for uniform changes, Feedforward-Feedback Incremental Conductance for gradual ramps, or Confidence Adaptive Particle Swarm Optimization (PSO) for complex PSC. The Confidence Adaptive PSO combines gradient seeded initialization with multi-metric confidence termination. Processor in Loop (PIL) validation on STM32-G474RE hardware (170 MHz ARM Cortex-M4) demonstrates 99%+ tracking efficiency across sequential condition transitions from uniform irradiance through gradual ramp to partial shading pattern redistributions, with 62.48 μs worst case execution time corresponding to 6.25% CPU utilization at 1 kHz control frequency. Across all six evaluation conditions spanning uniform irradiance, dynamic transitions, and partial shading, the proposed framework achieves the highest tracking efficiency among all compared methods without exception, with improvements of up to 19.12% over classical hill-climbing methods and up to 1.05% over the best competing metaheuristic under the most demanding partial shading scenario. The framework eliminates the traditional compromise between tracking speed and convergence reliability through condition-aware algorithm selection.
- 14 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 78085-78109
Open AccessEvent-Driven, Condition-Aware MPPT Framework for Robust Tracking Under Partial Shading and Dynamic Irradiance
Partial shading conditions (PSC) introduce multiple peaks in photovoltaic P-V characteristics, causing conventional MPPT algorithms to converge prematurely at local maxima. Existing metaheuristic approaches apply computationally expensive global search regardless of environmental transition type, wasting computation on simple conditions or failing to converge during complex PSC. This paper proposes an Event-Driven, Condition-Aware MPPT Framework where a hierarchical Environmental Monitoring Engine (EME) classifies environmental transitions through complementary detection, PSC pattern recognition via panel irradiance, step change filtering via irradiance rate, and checkpoint tracking via validated state transitions, and selects a suitable optimization strategy accordingly: Confidence Adaptive Sweep for uniform changes, Feedforward-Feedback Incremental Conductance for gradual ramps, or Confidence Adaptive Particle Swarm Optimization (PSO) for complex PSC. The Confidence Adaptive PSO combines gradient seeded initialization with multi-metric confidence termination. Processor in Loop (PIL) validation on STM32-G474RE hardware (170 MHz ARM Cortex-M4) demonstrates 99%+ tracking efficiency across sequential condition transitions from uniform irradiance through gradual ramp to partial shading pattern redistributions, with 62.48 μs worst case execution time corresponding to 6.25% CPU utilization at 1 kHz control frequency. Across all six evaluation conditions spanning uniform irradiance, dynamic transitions, and partial shading, the proposed framework achieves the highest tracking efficiency among all compared methods without exception, with improvements of up to 19.12% over classical hill-climbing methods and up to 1.05% over the best competing metaheuristic under the most demanding partial shading scenario. The framework eliminates the traditional compromise between tracking speed and convergence reliability through condition-aware algorithm selection.
- Open Access
- 25 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 48504-48522
Recurrent Reinforcement Learning for Urban Railway Speed Control
This study presents a recurrent reinforcement learning framework for urban railway speed control, aiming to achieve safe and energy-efficient automatic train operation under partially observable operating conditions. A Recurrent Soft Actor–Critic (RSAC) algorithm is developed to model the sequential decision-making problem of train driving, where temporal dependencies among observations are captured through a recurrent neural network. To improve training stability and accelerate policy convergence, a behavior cloning pretraining phase is introduced, enabling the agent to learn fundamental control patterns from expert trajectories generated by dynamic programming. Furthermore, a safety filter is designed to override policy outputs to ensure compliance with physical and operational constraints. An integrated reward design is proposed to incorporate safety, comfort, energy efficiency, and anticipatory braking behavior, enabling more stable and smooth policy learning. Experimental results demonstrate that the RSAC agent successfully learns smooth and adaptive control behavior while maintaining energy-efficient operation and stable convergence. The findings indicate the potential of recurrent reinforcement learning as a promising direction for next-generation automatic train operation systems.
- 25 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 48504-48522
Open AccessRecurrent Reinforcement Learning for Urban Railway Speed Control
This study presents a recurrent reinforcement learning framework for urban railway speed control, aiming to achieve safe and energy-efficient automatic train operation under partially observable operating conditions. A Recurrent Soft Actor–Critic (RSAC) algorithm is developed to model the sequential decision-making problem of train driving, where temporal dependencies among observations are captured through a recurrent neural network. To improve training stability and accelerate policy convergence, a behavior cloning pretraining phase is introduced, enabling the agent to learn fundamental control patterns from expert trajectories generated by dynamic programming. Furthermore, a safety filter is designed to override policy outputs to ensure compliance with physical and operational constraints. An integrated reward design is proposed to incorporate safety, comfort, energy efficiency, and anticipatory braking behavior, enabling more stable and smooth policy learning. Experimental results demonstrate that the RSAC agent successfully learns smooth and adaptive control behavior while maintaining energy-efficient operation and stable convergence. The findings indicate the potential of recurrent reinforcement learning as a promising direction for next-generation automatic train operation systems.
- Open Access
Cooperative Application of Onboard Energy Storage and Stationary Energy Storage in Rail Transit Based on Genetic Algorithm
The transition towards environmentally friendly transportation solutions has prompted a focused exploration of energy-saving technologies within railway transit systems. Energy Storage Systems (ESS) in railway transit for Regenerative Braking Energy (RBE) recovery has gained prominence in pursuing sustainable transportation solutions. To achieve the dual-objective optimization of energy saving and investment, this paper proposes the collaborative operation of Onboard Energy-Storage Systems (OESS) and Stationary Energy-Storage Systems (SESS). In the meantime, Non-dominated Sorting Genetic Algorithm-II (NSGA-II) is applied to optimize the ESS capacity and reduce its redundancy. The simulation is programmed in MATLAB. The results show that the corporation of OESS and SESS offers superior benefits (70 kWh energy saving within 30 min operation) compared to using SESS alone. Moreover, the OESS plays a significant role, emphasizing its significance in saving energy and investment, therefore presenting a win–win scenario. It is recommended that the capacity of OESS be designed to be two to three times that of SESS. The findings contribute to the ongoing efforts in developing more sustainable and energy-efficient transportation solutions, with implications for the railway industry’s investment and broader initiatives in energy saving for sustainable urban mobility.
Open AccessCooperative Application of Onboard Energy Storage and Stationary Energy Storage in Rail Transit Based on Genetic Algorithm
The transition towards environmentally friendly transportation solutions has prompted a focused exploration of energy-saving technologies within railway transit systems. Energy Storage Systems (ESS) in railway transit for Regenerative Braking Energy (RBE) recovery has gained prominence in pursuing sustainable transportation solutions. To achieve the dual-objective optimization of energy saving and investment, this paper proposes the collaborative operation of Onboard Energy-Storage Systems (OESS) and Stationary Energy-Storage Systems (SESS). In the meantime, Non-dominated Sorting Genetic Algorithm-II (NSGA-II) is applied to optimize the ESS capacity and reduce its redundancy. The simulation is programmed in MATLAB. The results show that the corporation of OESS and SESS offers superior benefits (70 kWh energy saving within 30 min operation) compared to using SESS alone. Moreover, the OESS plays a significant role, emphasizing its significance in saving energy and investment, therefore presenting a win–win scenario. It is recommended that the capacity of OESS be designed to be two to three times that of SESS. The findings contribute to the ongoing efforts in developing more sustainable and energy-efficient transportation solutions, with implications for the railway industry’s investment and broader initiatives in energy saving for sustainable urban mobility.
- 1 September 2023
- Published by Institute of Electrical Engineers of Japan (IEE Japan) in IEEJ Transactions on Industry Applications
- Vol. 143 (9) , 602-610
Optimization of Train Timetables and Placement of Charging Facilities for Battery-Powered LRT
The energy consumption of a battery-powered train in an interstation depends on the running time and state of energy (SOE) at departure. In this paper, we develop an optimization method of train timetables to minimize energy consumption in line with several stations. The variables in this proposed optimization model are running, dwell, and charging times as real numbers and places of charging facilities as binaries. Additionally, we conduct a case study using the real-world light rail transit (LRT) route, vehicle, and onboard battery model to confirm the effectiveness. In the case study, the proposed method can optimize the timetable and placement of charging facilities by considering the track gradient and battery SOE.
- 1 September 2023
- Published by Institute of Electrical Engineers of Japan (IEE Japan) in IEEJ Transactions on Industry Applications
- Vol. 143 (9) , 602-610
Optimization of Train Timetables and Placement of Charging Facilities for Battery-Powered LRT
The energy consumption of a battery-powered train in an interstation depends on the running time and state of energy (SOE) at departure. In this paper, we develop an optimization method of train timetables to minimize energy consumption in line with several stations. The variables in this proposed optimization model are running, dwell, and charging times as real numbers and places of charging facilities as binaries. Additionally, we conduct a case study using the real-world light rail transit (LRT) route, vehicle, and onboard battery model to confirm the effectiveness. In the case study, the proposed method can optimize the timetable and placement of charging facilities by considering the track gradient and battery SOE.
- Open Access
- 1 January 2023
- Published by Japan Society of Mechanical Engineers in Mechanical Engineering Journal
- Vol. 10 (3) , 22-00360
A method for optimizing urban rail transit timetable based on accurate power flow
The timetable of urban rail greatly affects its daily energy consumption. To improve the utilization of renewable energy between trains using timetabling has become an effective way to reduce energy consumption. Previous studies ignore or simplify the modelling of traction power supply network, which failed to accurately describe the flow of energy between trains through the power network. This paper proposed an optimisation method of energy efficiency timetabling considering the power flow of traction power supply network. First, an urban rail transit DC traction network model is established, and the current-vector iterative method is used to characterize the energy consumption. Then, a train timetable optimisation model is proposed to minimize the total energy consumption of the traction network system by adjusting the dwell time and section running time. The genetic algorithm is used to solve the optimisation problem. Finally, simulation result shows that the proposed method can accurately characterize the energy flow and effectively reduce the total energy consumption of the urban rail transit.
- 1 January 2023
- Published by Japan Society of Mechanical Engineers in Mechanical Engineering Journal
- Vol. 10 (3) , 22-00360
Open AccessA method for optimizing urban rail transit timetable based on accurate power flow
The timetable of urban rail greatly affects its daily energy consumption. To improve the utilization of renewable energy between trains using timetabling has become an effective way to reduce energy consumption. Previous studies ignore or simplify the modelling of traction power supply network, which failed to accurately describe the flow of energy between trains through the power network. This paper proposed an optimisation method of energy efficiency timetabling considering the power flow of traction power supply network. First, an urban rail transit DC traction network model is established, and the current-vector iterative method is used to characterize the energy consumption. Then, a train timetable optimisation model is proposed to minimize the total energy consumption of the traction network system by adjusting the dwell time and section running time. The genetic algorithm is used to solve the optimisation problem. Finally, simulation result shows that the proposed method can accurately characterize the energy flow and effectively reduce the total energy consumption of the urban rail transit.
- Open Access
An Attended-Free, All-in-One-Go, Automatic Analysis Assistant Software for E-liked Shape Contactless Inductive Power Transfer Device
This paper presents automatic software for E-liked shaped contactless inductive power transfer (CIPT) device study and design that provides attended-free, multiple-case auto-generating and auto-deploying analysis in one go. It provides visualized and listed results in a design space or for optimizing solutions. To satisfy the demand for static and dynamic charging devices, the software provides specific cores, such as EE-, EI-, IE-, and II-shaped, with or without legs as optional core structures. The software contains three main parts: a user-friendly interface, analytic approaches providing grid analysis that represent the general performance in a designated parameter range, and optimal analysis for multi-objective optimization using a genetic algorithm (GA). The post-analysis processor converts the analysis results to easy-to-read outputs. Users can customize various parameters, such as core type, structural size, circuit configuration, materials, and analysis setting. Automatic functions, such as resistance and compensation calculation, are available for the convenience of the user. By applying one approach, or by combining them in a specific order, the software achieves designs that satisfy the user’s demands within the user-provided range. The software is built in Python and collaborates with a finite element method (FEM) solver, which is JMAG in this paper. Some examples are given to demonstrate the performance of the software.
Open AccessAn Attended-Free, All-in-One-Go, Automatic Analysis Assistant Software for E-liked Shape Contactless Inductive Power Transfer Device
This paper presents automatic software for E-liked shaped contactless inductive power transfer (CIPT) device study and design that provides attended-free, multiple-case auto-generating and auto-deploying analysis in one go. It provides visualized and listed results in a design space or for optimizing solutions. To satisfy the demand for static and dynamic charging devices, the software provides specific cores, such as EE-, EI-, IE-, and II-shaped, with or without legs as optional core structures. The software contains three main parts: a user-friendly interface, analytic approaches providing grid analysis that represent the general performance in a designated parameter range, and optimal analysis for multi-objective optimization using a genetic algorithm (GA). The post-analysis processor converts the analysis results to easy-to-read outputs. Users can customize various parameters, such as core type, structural size, circuit configuration, materials, and analysis setting. Automatic functions, such as resistance and compensation calculation, are available for the convenience of the user. By applying one approach, or by combining them in a specific order, the software achieves designs that satisfy the user’s demands within the user-provided range. The software is built in Python and collaborates with a finite element method (FEM) solver, which is JMAG in this paper. Some examples are given to demonstrate the performance of the software.
- 1 July 2019
- Published by Institute of Electrical Engineers of Japan (IEE Japan) in IEEJ Journal of Industry Applications
- Vol. 8 (4) , 586-591
Energy Efficient Train Trajectory in the Railway System with Moving Block Signaling Scheme
High-frequency train operation with a moving-block signaling system has begun to be introduced mainly in urban rail transit. Under high-frequency operation, however, subsequent trains can repeatedly be forced to slow down according to deceleration pattern generated by the signaling system to avoid collision with preceding trains, which increases the energy consumption of subsequent trains. In this paper, train trajectory optimization with dynamic programming is applied to evaluate the effect of considering preceding trains' trajectory on subsequent trains' energy reduction in a line with moving-block signaling system. The result shows that the method reduces the increase in energy consumption compared to the preceding train 1/3 times more than when not considering the position of the preceding train.
- 1 July 2019
- Published by Institute of Electrical Engineers of Japan (IEE Japan) in IEEJ Journal of Industry Applications
- Vol. 8 (4) , 586-591
Energy Efficient Train Trajectory in the Railway System with Moving Block Signaling Scheme
High-frequency train operation with a moving-block signaling system has begun to be introduced mainly in urban rail transit. Under high-frequency operation, however, subsequent trains can repeatedly be forced to slow down according to deceleration pattern generated by the signaling system to avoid collision with preceding trains, which increases the energy consumption of subsequent trains. In this paper, train trajectory optimization with dynamic programming is applied to evaluate the effect of considering preceding trains' trajectory on subsequent trains' energy reduction in a line with moving-block signaling system. The result shows that the method reduces the increase in energy consumption compared to the preceding train 1/3 times more than when not considering the position of the preceding train.
- Open Access
- 1 January 2019
- Published by Japan Society of Mechanical Engineers in Transactions of the JSME (in Japanese)
- Vol. 85 (880) , 19-00092
A method of generating energy-efficient timetable for catenary free railways with battery trains
Battery train is in development as train which can travel in non-electrified section using power supplied from onboard storage system such as lithium-ion battery. However, as characteristics of this type of train, an energy consumption of battery train depends on the state of charge of the storage system. In consideration of these characteristics, we propose a generation method of the timetable which is the most energy-saving when a single battery train travels on a route section containing multiple stations in this research. Although this optimization can be defined as a nonlinear programming problem, we ease to solve this problem using linear approximation to the energy consumption characteristic. In the end, we carried out a simulation in a simple case and achieved 4.7% reduction in energy consumption by the proposed method.
- 1 January 2019
- Published by Japan Society of Mechanical Engineers in Transactions of the JSME (in Japanese)
- Vol. 85 (880) , 19-00092
Open AccessA method of generating energy-efficient timetable for catenary free railways with battery trains
Battery train is in development as train which can travel in non-electrified section using power supplied from onboard storage system such as lithium-ion battery. However, as characteristics of this type of train, an energy consumption of battery train depends on the state of charge of the storage system. In consideration of these characteristics, we propose a generation method of the timetable which is the most energy-saving when a single battery train travels on a route section containing multiple stations in this research. Although this optimization can be defined as a nonlinear programming problem, we ease to solve this problem using linear approximation to the energy consumption characteristic. In the end, we carried out a simulation in a simple case and achieved 4.7% reduction in energy consumption by the proposed method.
- 1 January 2019
- Published by Japan Society of Mechanical Engineers in Transactions of the JSME (in Japanese)
- Vol. 85 (878) , 19-00079
A study on energy-efficient train driving in a line with short headway
Energy-efficient driving is one of the major effective methods to reduce energy consumption of railway system. It has substantial advantage that need much less investment cost than other countermeasures. However, that has not been investigated in detail at short headway situation in rush hours where two trains interact with each other through signaling systems to prevent train collision. In this situation, some trains make unexpected accelerations and decelerations much often to track the preceding train and increase energy loss. Here we show theoretically what driving technique is much better from energy consumption point of view. In the simulation, we examine a technique to drive a train in an energy-efficient way when the preceding train is delayed departing station by simulations. The dynamic programming (DP) is employed to optimize the energy-efficient train speed profile and modified to implement the behavior of fixed-block signaling systems. The sensitivity analysis to change the delay of first train is made to clarify the relation between energy consumption and train delay propagation to the second train. The results show that it can suppress energy consumption less than or equal to regular running time situation.
- 1 January 2019
- Published by Japan Society of Mechanical Engineers in Transactions of the JSME (in Japanese)
- Vol. 85 (878) , 19-00079
A study on energy-efficient train driving in a line with short headway
Energy-efficient driving is one of the major effective methods to reduce energy consumption of railway system. It has substantial advantage that need much less investment cost than other countermeasures. However, that has not been investigated in detail at short headway situation in rush hours where two trains interact with each other through signaling systems to prevent train collision. In this situation, some trains make unexpected accelerations and decelerations much often to track the preceding train and increase energy loss. Here we show theoretically what driving technique is much better from energy consumption point of view. In the simulation, we examine a technique to drive a train in an energy-efficient way when the preceding train is delayed departing station by simulations. The dynamic programming (DP) is employed to optimize the energy-efficient train speed profile and modified to implement the behavior of fixed-block signaling systems. The sensitivity analysis to change the delay of first train is made to clarify the relation between energy consumption and train delay propagation to the second train. The results show that it can suppress energy consumption less than or equal to regular running time situation.
- Open Access
Increase in Robustness against Effects of Coil Misalignment on Electrical Parameters Using Magnetic Material Layer in Planar Coils of Wireless Power Transfer Transformer
Utilization of wireless power transfer in light rail transits is seen as one solution for electrification of lines. The main advantage of this supply system is the reduction of installation; moreover, the alignment between the transmitter coil in the track and the receiver coil in the train should be perfect in order not to affect the power transfer. To reduce the effects of misalignment on the input and output electric parameters of the system, a new planar core and coil design, called hybrid intercore coil, is proposed. The proposed design uses a magnetic material layer between the windings in the inner half of the coil to create a non-uniform magnetic field distribution, which makes the system more robust against the effects of coil misalignment on the system current and voltage. Simulations with finite element method software were conducted to compare designs. The results show that the proposed design is less susceptible to the effects of misalignment and is more efficient. Prototype cores were constructed to verify the simulation results. Measurements show a smaller input overcurrent and output overvoltage when operating in resonance mode. The proposed design reduced the effects of coil misalignment on electrical parameters.
Open AccessIncrease in Robustness against Effects of Coil Misalignment on Electrical Parameters Using Magnetic Material Layer in Planar Coils of Wireless Power Transfer Transformer
Utilization of wireless power transfer in light rail transits is seen as one solution for electrification of lines. The main advantage of this supply system is the reduction of installation; moreover, the alignment between the transmitter coil in the track and the receiver coil in the train should be perfect in order not to affect the power transfer. To reduce the effects of misalignment on the input and output electric parameters of the system, a new planar core and coil design, called hybrid intercore coil, is proposed. The proposed design uses a magnetic material layer between the windings in the inner half of the coil to create a non-uniform magnetic field distribution, which makes the system more robust against the effects of coil misalignment on the system current and voltage. Simulations with finite element method software were conducted to compare designs. The results show that the proposed design is less susceptible to the effects of misalignment and is more efficient. Prototype cores were constructed to verify the simulation results. Measurements show a smaller input overcurrent and output overvoltage when operating in resonance mode. The proposed design reduced the effects of coil misalignment on electrical parameters.