Toyota Motor Corporation (Japan)

Federated learning (FL) in vehicular networks faces significant challenges due to client nodes’ limited resources and highly dynamic network conditions, which hider active participation and degrade model performance. This paper proposes a novel incentive mechanism that explicitly addresses these challenges by combining a fuzzy logic-based contribution estimation algorithm with a reinforcement learning-based rewarding module. The estimation algorithm evaluates each client’s potential contribution by considering mobility patterns, link quality, and computational capacity, while the rewarding module dynamically adjusts incentives based on both estimated potential and actual performance. The estimation algorithm evaluates each client’s potential contribution by considering mobility patterns, link quality, and computational capacity, while the rewarding module dynamically adjusts incentives based on both estimated potential and actual performance. Extensive simulations on realistic vehicular network scenarios show that our scheme increases client participation by approximately 37% and shortens the average convergence time by about 30 global epoches compared with traditional incentive approaches. These results demonstrate that our method effectively enhances cooperation and model performance in resource-constrained, dynamic vehicular FL environments.

A general formulation of the energy dissipation in soft magnetic materials was established, and the contributions from both eddy currents and magnetostriction were compared using this formulation. The energy dissipation of soft magnetic materials has been discussed based on the Steinmetz equation, considering contributions from coercivity and eddy current. However, recently, it was reported that magnetostriction can also generate energy dissipation due to changes in strain corresponding to the domain-wall motion. Hence, a modification of the Steinmetz equation is indispensable for better understanding the mechanism of energy dissipation in soft magnetic materials. Our general formulation elucidates the origin of magnetic energy dissipation across various soft magnetic materials. The results provide a novel explanation for the mechanisms underlying the damping of the domain-wall motion. • The general equation of the energy loss of the soft magnetic materials was formulated considering the effects of coercivity, eddy current, and magnetostriction. • Our general formulation reproduced experimental energy losses. • Prediction of the dominant loss mechanisms in practical materials can develop efficient magnetic cores. • Excess losses are significantly decreased by the increase in the number of domain walls in the soft magnetic materials with the small magnetostriction. • The results provide design criteria for soft magnetic materials considering the effects of magnetostriction and eddy current.

Abstract In this study, scalable synthesis of highly uniform Y(OH)3 nanoparticles, serving as Y2O2S precursors, via a solvothermal method was demonstrated. Subsequently, Y2O2S nanoparticles were successfully synthesized via postsulfurization under a carbon disulfide (CS2) atmosphere. The strong sulfurization ability of CS2 enabled processing at relatively low temperatures and within short durations, preserving the nanoscale dimensions of the product. The small particle size of the precursor also played a key role in promoting rapid sulfurization. Furthermore, the degree of sulfurization could be controlled by adjusting the temperature, allowing selective formation of oxysulfides or sulfides.

In recent years, the importance of sustainable energy sources in space exploration has been increasing, with regenerative fuel cells (RFCs) gaining attention. This study examines the design and performance evaluation of an RFC system with a focus on its application to lunar rovers. RFCs store energy in the form of hydrogen and oxygen and can efficiently generate electrical energy and water from the stored hydrogen and oxygen, making them particularly useful for long-duration missions.The Artemis program plans to explore the lunar south pole, with the possibility of future missions targeting the equatorial region. The lunar surface has a long day-night cycle, necessitating the storage of electricity generated by solar cells during the day for use during the long night. In this regard, RFCs, when combined with solar power generation, become a promising candidate for nighttime energy supply.Although RFCs are less efficient compared to lithium-ion batteries, they offer significant advantages in terms of energy storage density and mass per unit of stored energy. Additionally, the fact that the oxidizer and fuel can be stored independently from the cell enhances safety during large energy storage.At Toyota Motor Corporation, we have expertise and experience gained from our commercially available fuel cell electric vehicle (FCEV), the MIRAI, and we have developed water electrolysis technology in collaboration with Mitsubishi Heavy Industries (MHI) and the Japan Agency for Marine-Earth Science and Technology (JAMSTEC). Based on these technologies, we have constructed an RFC system and conducted demonstration experiments.We will present the performance test results of the laboratory prototype and use these results to predict the performance of the RFC system for lunar rovers. Additionally, we will introduce several safety evaluation results, demonstrating that RFCs hold potential as a solution that meets the functional and safety requirements for lunar rovers.This research provides a new approach to energy storage and regenerative energy cycling for space vehicles and lays the foundation for future lunar exploration, long-duration missions, and the operation of lunar rovers.

Recently, local 5G is expected to spread in earnest due to lower installation costs. There are various use cases of Local 5G in factories. However, there can be poor communication conditions due to a lot of movable metal equipment and partition walls. Therefore, we consider metasurface reflector that diffusely reflect radio waves to expand the coverage area. In this paper, we designed and manufactured metasurface reflector, and measured RSSI and BLER. Moreover, experimental trial of L5G coverage area construction using metasurface reflector in automobile factory was carried out. As a result, BLER was improved by using a metasurface reflector even in an actual environment, and effectiveness for coverage area construction was verified.

A multiphase induction motor (IM) can change the pole number by controlling it as a superposition of 3phase IMs with different pole numbers through vector space decomposition (VSD). However, during pole change, excessive current is generated, which can trigger the current limiter, causing torque fluctuations. In this paper, to address this, a new pole change method applying maximum torque per ampere (MTPA) control between different pole numbers, minimizing the current during pole change, is proposed. This method prevents current from reaching its limiter and expands the region where pole change can occur at a constant torque. The efficacy of the method is demonstrated through simulations using a 400kW IM.