Wireless Power Transfer (WPT) systems fundamentally translate the way we think about energy transfer, moving from traditional wired connections to a more seamless, wireless approach. Combining history with cutting-edge technology, these systems utilize electromagnetic fields to transmit electrical energy from a power source to a load without the need for physical connectors or wires. This innovative concept, dating back to Nikola Tesla's groundbreaking experiments in the 1890s, has flourished for decades, finding applications on everyday devices such as smartphones, electric toothbrushes and sensor networks that fall below the internet of things.
The WPT technology core has a transmitter coil linked to a power source, converting power source energy into an electromagnetic field. This field is captured by a receiver coil that powers energy to the electronic device. However, one of the main challenges within a WPT system is achieving load-independent (LI) operations. This is an important feature that maintains stable output voltage and zero voltage switching (ZVS) across fluctuating loads. Traditional means of solving this problem often rely on complex analytical equations with idealized assumptions that cannot address the myriad of real-world irregularities.
To tackle these complex challenges, a pioneering research team led by Professor Hirao Sekiya of the Faculty of Information Studies at Chiba University in Japan has made great strides by introducing machine learning-based design methods for LI-WTP systems. Working with electrical engineering and computer science experts, the teams embarked on a journey to increase the efficiency of electricity supply through innovative approaches, including Fukuda, Dr. Komiyamo from Chiba University, Dr. Wenki Zhu from Tokyo University of Science, and Dr. Kojima from Soho University. Their findings were published in Circuits and Systems I.'s IEEE Transactions.
.adsslot_yxg21myqgp {width:728px! Fality; height: 90px! fairity;}
@media (max-width: 1199px) {.adsslot_yxg21myqgp {width: 468px! Fality; height: 60px! fairity;}
}
@media (max-width: 767px) {.adsslot_yxg21myqgp {width: 320px! Faliest; height: 50px! fairity;}
}
advertisement
The proposed new design process employs a completely numerical framework that utilizes differential equations to describe the dynamic behavior of voltages and currents within a WPT system. By embracing a numerical approach, researchers can realistically explain the various properties of physical components, a leap beyond traditional analytical methods. This new approach involves gradually solving equations, which stabilizes as the circuit's performance evolves to steady state.
At the heart of this design methodology is an assessment function that measures the effectiveness of the system by focusing on key parameters such as output voltage stability, power delivery efficiency, and total harmonic distortion. By employing genetic algorithms, the team was able to repeatedly fine-tune system parameters and strengthen the evaluation score until the load-independent operation goals were successfully achieved. This integration of machine learning into design not only demonstrates the practical utility of artificial intelligence, but also means a significant change in how future power electronics research and development will occur.
Professor Sekiya emphasized the transformative implications of this work, “establishing a new design procedure for LI-WPT systems that achieve constant output voltage without control over load fluctuations. We believe that load independence is an important technology for the social implementation of WPT systems.” This innovative thinking depicts a bright future for WPT technology, indicating that load independence can pave the way for wider use.
From a practical application perspective, the researchers applied the method to a specific type of WPT system, i.e., Class-EF WPT system. This design combines the advantages of ClassF inverters and Class D rectifiers to provide a robust solution to the problems faced by traditional systems. Although traditional designs usually lose ZV when the load changes, the Li WPT system developed by the Sekiya team maintained both ZV and stable output voltage, regardless of load variation, and showed significant resilience.
Their assessment revealed a significant contradiction between traditional methods and completely numerical methods. In traditional LI inverter systems, the output voltage can change dramatically as the load changes. In contrast, the newly designed system maintains this variation below 5%, indicating stability that could revolutionize the way WPT technology is used. This enhanced performance extends to lighter loads that allow new systems to better manage the diode parasitic capacitance effect and further enhance its benefits.
A thorough analysis of power losses within the system showed that newly designed transmission coils can dissipate similar levels of power across a variety of load conditions. This efficiency comes from the design of the system and ensures consistent output current, a key component of reliable wireless power distribution. At a rated operating point, the LI Class-EF WPT system achieves an impressive power delivery efficiency of 86.7% at a frequency of 6.78 MHz and is capable of delivering power output of over 23 watts.
From a future perspective, researchers imagine a broader meaning for their discoveries, suggesting that advances in WPT technology could be a step towards a completely wireless society. Professor Sekiya points out that the simplifications made possible by LI operations can lead to reduced costs and size of WPT systems, which can help to promote widespread adoption in everyday applications. The ambition is to normalize WPT technology over the next five to ten years and fundamentally change the interaction between energy transfer and consumption.
Essentially, this research not only reveals important advances in wireless power transfer technology, but also opens an exciting pathway for machine learning integration in the field of power electronics. It highlights the shift towards automated design processes that are poised to redefine how such systems are conceptualized, developed and manufactured, and highlights the possibility that technology can adapt fluidly to actual complexity.
The work carried out by the University of Chiba team encapsulates important milestones in search of efficient and reliable wireless energy transfer. The impact on appliances and wider applications could mark a new era in which power becomes truly wireless and paves the way for innovation that changes everyday life.
As they continue to explore new perspectives in WPT technology, the research team's work is a testament to the synergy between advanced engineering methods and artificial intelligence, demonstrating the power of interdisciplinary collaboration in overcoming years of challenges within electronic systems.
Research subject: Wireless power transmission system
Article Title: Complete ML-based numerical design method for load-independent class-EF WPT systems
News Release Date:18-Jun-2025
Web reference: IEEE Transactions on Circuits and Systems
reference: Not applicable
Image credits: Wikimedia Commons via Creative Commons Search Repository
keyword: Wireless power transfer, load-independent operation, machine learning, differential equations, circuit design, power supply efficiency, Nikola Tesla, University of Chiba, Class EF WPT systems, electronic equipment automation, energy transfer.
Tags: Wireless Power System Challenges WPT SystemseLectRomagnetic Field Energy Propagation Efficiency Infectious Energy Transmission Solutions Power Solution Independent Manipulation Energynikola Tesla Wireless Energy Experiments of Energy Transfising Switteg owores optimizing optimizin
