Resource
2026 EN
Khang Le · Daniel Monagle · Steven B. Leeb
Current transformer magnetic energy harvesters (CTMEHs) harvest and convert magnetic energy from ac conductors into electrical energy for low-power sensors. The CTMEH magnetic core is a critical design component as its material properties significantly influence harvested power. Previous works have mainly focused on modeling CTMEH behavior or improving power harvest given a specific core material. Some existing research has explored the effects of different core materials on power harvest in limited fashion; but a comprehensive, comparative study of high permeability, high saturation flux density CTMEHs had yet to be explored. This paper presents a thorough comparison of the effects of CTMEH core material on power harvest. State-of-the-art magnetic cores are evaluated at several ac transmission line currents and loading conditions. Results demonstrate that high effective initial permeability materials excel at lower-current applications, while high saturation flux density cores offer better performance at higher-current applications. Meanwhile, both high effective initial permeability and high saturation flux density are desirable for a magnetic energy harvester core. The comparative results of this paper offer energy harvesting sensor designers targeted guidance in CTMEH material choices.
Resource
2026 EN
Yuansheng Li · Yixing Du · Junhu Ma
This paper presents a novel loopy sum-product algorithm (LSPA) designed specifically for simultaneous tracking of point and extended targets—scenarios in which both target types may appear concurrently. Built upon the principles of belief propagation, the proposed approach iteratively updates target state beliefs using a generalized measurement model that jointly accommodates both target types. These beliefs constitute principled approximations to the marginal posterior densities of individual targets. To ensure computational feasibility, we introduce a structured factor graph and message-passing-based dimensionality reduction target tracking method. This method strictly "stretching" the factor nodes and variable nodes in the underlying factor graph to obtain a compact node representation with reduced dimensions, and then performs the LSPA algorithm on this factor graph. For closed-form tractability and robust deployment, each target's belief is modeled as a statistically coherent mixture density: Gaussian components for the kinematic state of point targets, and Gamma-Gaussian-Inverse-Wishart (GGIW) components for the kinematic and extended states of extended targets. We provide a complete mathematical derivation of the resulting algorithm and empirically validate its superiority through extensive Monte Carlo simulations. Compared to state-of-the-art benchmark methods, the proposed approach achieves statistically significant improvements in tracking accuracy, convergence speed, and robustness.
Resource
2026 EN
Jan Stake · Steven Durant · Lucy Williamson
+1 more
Calorimetric sensors that incorporate a rectangular waveguide interface facilitate absolute and precise power measurements. Moreover, by using a waveguide taper that transitions to a smaller, single-mode waveguide feed, the bandwidth limitation of a standard waveguide band can be extended to higher frequencies, enabling accurate power measurements across a wide portion of the electromagnetic spectrum. For aligned waveguide apertures without discontinuities, the incident field remains in the fundamental mode, thereby enabling broadband power measurements. In this work, we present a systematic study of intentionally misaligned waveguide junctions, specifically symmetrical H -plane or E -plane displacements at the WR-10 input flange of a PM5B power sensor, and their overall impact on the mismatch. The return loss was measured and simulated over a frequency span of 75 GHz to 1100 GHz, exceeding a decade. Electromagnetic simulations agree well with the experimental results, showing an excellent return loss for the aligned junction and a degradation attributed to the onset of higher-order parasitic modes in the misaligned case. In particular, it is found that the onset of a propagating TM $_{11}$ mode due to E -plane displacement is the primary concern, while H -plane displacements have a negligible effect. However, a return loss better than 20 dB can be maintained if this offset is kept below 4% of the waveguide height. Hence, broadband, precise, and absolute power measurements are feasible provided that waveguide feed junctions are aligned and fabricated with high tolerances.
Resource
2026 EN
Sofia Mvokany · Jack A. Molles · Steven Verhaverbeke
+4 more
This paper presents a variety of broadband, low-loss interconnects designed on an experimental process that includes 6 organic resin-filled dielectric layers laminated on both sides of a silicon interposer. The process allows metallization of each dielectric layer, making 6 metallic layers available for design. This paper highlights 3 interconnects using 3 metal layers, as well as a coaxial through-silicon via (TSV) interconnect using all 6 layers. Interconnects on both high-resistivity and low-resistivity interposer wafers are characterized from 2 to 50 GHz, showing a return loss better than 10 dB and insertion loss below 0.5 dB across the frequency range for organic-layer interconnects. The TSV transitions demonstrate 0.11 dB/mm and 0.16 dB/mm insertion loss at 10 GHz when fabricated on high and low resistivity interposers, respectively. Broadside couplers are designed and characterized to evaluate cross-coupling between lines on different metal layers. A second generation wafer including multilayer passive circuits is fabricated and measured. Results for a 6-dB grounded coplanar waveguide coupler and a substrate-integrated waveguide multi-layer filter are presented.
Resource
2026 EN
Vatan Mehar · Nicholas Frooninckx · Isaac Abram
+4 more
This paper presents the integration of a 200-kW inter-leaved buck–boost dc–dc converter with a Class-8 electric heavyduty vehicle and its evaluation on a dynamic wireless power transfer roadway. The converter has been rigorously validated through simulation, hardware-in-the-loop testing, and full-scale on-road deployment at vehicle speeds up to 105 km/h. A detailed description of the receiver-side control architecture is provided, addressing key challenges such as coil misalignment and pad-to-pad transitions to enable stable and regulated power delivery under variable operating conditions. The control system dynamically responds to fluctuations in receiver voltage to maintain alignment with vehicle power demand and enhance performance under imperfect transmitter–receiver coupling. Comprehensive hardware results from vehicle-level testing confirm the converter’s ability to regulate peak power across a highly variable input profile, demonstrating robust and efficient operation under realistic driving scenarios.
Resource
2026 EN
Roark Chao · Alexander W. West · Joshua B. Wolfgram
+5 more
In this work, we experimentally demonstrate InGaN/GaN $\mu $ LEDs that maintain stable emission characteristics up to $250 circ }$ C, representing one of the highest reported operating temperatures for stable visible $\mu $ LEDs. Devices preserve their spectral integrity with minimal wavelength drift and sustain measurable optical output across the full temperature range. The $\mu $ LEDs also exhibit reliable current and voltage characteristics under high-temperature biasing, indicating robust junction behavior and negligible degradation of carrier transport. This represents one of the highest demonstrated operating temperatures for visible $\mu $ LEDs with preserved spectral integrity. These results highlight the intrinsic thermal resilience of III-nitride emitters and underscore their suitability for photonic links in harsh-environment conditions, paving the way for compact, energy-efficient, and thermally tolerant VLC transmitters integrated for next-generation data-center architectures and other thermally challenging environments.
Resource
2026 EN
Steven Matthew Cheng · Andres Fontana · Dimitra Psychogiou
In this letter, a reflectionless low-noise amplifier (LNA) filter (RLNAF) is presented, incorporating four distinct functions: 1) filtering; 2) reflection cancellation; 3) low-noise amplification; and 4) RF switching within a single component. It is comprised of a transmissive path and two absorptive paths. Specifically, a complex-termination methodology is applied to the transmissive path incorporating a bandpass filter (BPF) and LNA functionalities, removing the need for additional interstage matching networks (MNs). The absorptive paths are composed of a bandstop filter terminated with a loading resistor in one port to selectively dissipate the out-of-band reflected RF signals. Moreover, switch-off functionality is demonstrated by reconfiguring the bias of the transistor. For proof-of-concept validation purposes, two RLNAF prototypes are designed, manufactured, and measured at C-band. The RF characteristics are given as follows: RLNAF I: $f_{cen} = 4.65$ GHz, gain = 15.8 dB, noise figure (NF) = 2.07 dB, and quasi-reflectionless characteristics better than 10 dB from DC to 4.27 and 4.34 GHz to 6.42 GHz and at least 9.8 dB from 4.27 to 4.34 GHz; RLNAF II: $f_{cen} = 3.7$ GHz, gain = 15.3 dB, NF = 1.09 dB, and quasi-reflectionless characteristics better than 10 dB from DC to 4.75 and 5.75 GHz to 7.03 GHz and at least 6 dB from 4.75 to 5.75 GHz.
Resource
2026 EN
Steven Rice · Sherif Saad · Ahmed Azab
+1 more
Inverse Kinematics (IK) is an integral part of robot manipulation. IK can be challenging to solve, and many computer-aided approaches have been proposed but each has its limitations. The emergence of Large Language Models (LLMs) has seen them applied to solving complex tasks including mathematical problems. This work proposes “LLM-IK” to utilize LLMs to solve IK problems. Relevant serial manipulator information is extracted from descriptor files, prompt engineered, and then provided to the LLMs with methods and feedback to interact with and learn about the kinematic chain. Multiple methods of breaking down kinematic chains into distinct sub-problems are implemented allowing for incremental problem solving. Experiments showed LLM-IK solves up to six Degrees-of-Freedom (DOF) and outperforms IKFast, TRAC-IK, and IKPy in both accuracy and solving time, highlighting this methodology can produce highly efficient and human-readable solutions.
Resource
2026 EN
Aref Amiri · Steven M. LaValle
Feedback motion planning over cell decompositions provides a robust method for generating collision-free robot motion with formal guarantees. However, existing algorithms often produce paths with unnecessary bending, leading to slower motion and higher control effort. This paper presents a computationally efficient method to mitigate this issue for a given simplicial decomposition. A heuristic is introduced that systematically aligns and assigns local vector fields to produce more direct trajectories, complemented by a novel geometric algorithm that constructs a maximal star-shaped chain of simplexes around the goal. This creates a large “funnel” in which an optimal, direct-to-goal control law can be safely applied. Simulations demonstrate that our method generates measurably more direct paths, reducing total bending by an average of 91.40% and LQR control effort by an average of 45.47%. Furthermore, comparative analysis against sampling-based and optimization-based planners confirms the time efficacy and robustness of our approach. While the proposed algorithms work over any finite-dimensional simplicial complex embedded in the collision-free subset of the configuration space, the practical application focuses on low-dimensional ( $d\le 3$ ) configuration spaces, where simplicial decomposition is computationally tractable.
Resource
2026 EN
Michael Starke · Benjamin Dean · Namwon Kim
+4 more
Complex modern energy systems with multiple converters, distributed energy resources, and dynamic control modes have posed many challenges, particularly in terms of system coordination, reliability, and scalability. Among them, start-up is one of the most challenging, as the activation of one device often depends on others. In such a scenario, traditional pre-configured start-up methods become impractical and inflexible. In this context, this paper proposes a resilient, optimization-based framework for the start-up of networks populated with power electronic systems. To address this challenge, a linear programming optimization methodology is proposed to determine the sequential activation of devices based on system topology, available control modes (e.g., bus-forming or grid-following), and the presence of faults. The framework supports systems with shared buses and integrates converter-level information via resource integration controllers and a centralized resource management controller. Device start-up is modeled through time-step-based formulations that reflect bus energization constraints, converter capabilities, and interdependencies between subsystems. The proposed solution is implemented and validated on a real-time controller hardware-in-the-loop platform. To validate the proposed framework, four use cases are evaluated: i) grid-based activation using AC/DC converters, ii) energy storage-initiated start-up with bus-forming capability, iii) a faulted converter case that triggers re-optimization, iv) a fault occurs in a partially started system considering the worst case impact. Results show that the framework can dynamically adapt to changing conditions, accommodate new converter capabilities, and maintain reliable start-up even with failed devices. This approach enhances the flexibility and resiliency of PES-integrated systems and offers a scalable path forward for autonomous system activation in complex electrical networks.