⚡ electrical engineering

Co-Optimization of a Lightweight and Misalignment-Tolerant Concentric Multi-Ring Coupler for Wireless Power Transfer

This paper proposes a surrogate-assisted discrete-parameter optimization method to co-design a lightweight and misalignment-tolerant concentric multi-ring coupler for wireless power transfer, achieving a validated trade-off between high rated mutual inductance, a 110.87 mm misalignment tolerance, and a reduced receiver mass of 446.39 g.

Original authors: xudong pan, Peng Yang, Mingxuan Zhang, Ruibing Sun, Dechun Yuan

Published 2026-07-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: xudong pan, Peng Yang, Mingxuan Zhang, Ruibing Sun, Dechun Yuan

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where your electric car, your phone, or even a medical implant inside your body could recharge without ever needing to plug in a cable. This isn't magic; it's a technology called Wireless Power Transfer (WPT). Think of it like a magical invisible rope that carries electricity through the air. But here's the catch: just like trying to catch a specific radio station, this "invisible rope" only works perfectly if the sender (the charging pad) and the receiver (the device) are lined up exactly right. If you shift them even a little bit to the side, the connection gets weak, the charging slows down, or stops completely. This is called misalignment, and it's the biggest headache for engineers trying to make wireless charging practical for moving things like cars.

To fix this, engineers use special magnetic coils and blocks of a material called ferrite (which acts like a funnel for magnetic fields) to guide the energy. However, there's a tricky balancing act. You want the system to be super tolerant of being out of alignment so you don't have to park perfectly, but you also don't want the receiver part to be a heavy, clumsy brick. If the receiver is too heavy, it's hard to install and wastes energy just by existing. The challenge is finding the perfect recipe of coil turns and ferrite blocks that makes the system light and forgiving, without having to test millions of impossible combinations.

This is exactly the puzzle a team of researchers from Northeast Forestry University tackled in their new study. They looked at a specific design called a Concentric Multi-Ring Coupler (CMRC). Imagine this coupler as a set of nested rings, like a target board, where a main ring and several smaller helper rings work together to create a wide, stable magnetic field. The problem is that the design has many "knobs" to turn—like how many times the wire is wrapped around the coil or how many ferrite segments are used. These aren't smooth knobs you can turn slightly; they are discrete steps, like counting integers (1, 2, 3...).

Trying to find the best combination by testing every single possibility is like trying to find a specific grain of sand on a beach by digging up every inch of the sand first. The researchers calculated that checking every possible combination would require over 162,288 complex computer simulations. That would take forever and cost a fortune in computing power.

Instead of brute force, the authors invented a clever shortcut. They used a "surrogate model," which is like a super-smart guesser. First, they ran a small number of precise simulations (only about 68 total) to teach the computer the rules of the game. Then, they used a strategy called beam search to rapidly scan through the millions of possibilities, keeping only the most promising "lightweight" candidates and discarding the rest. It's like a detective who, instead of interviewing every person in a city, quickly narrows down the suspect list to just a few people who fit the description, and then investigates those few deeply.

The result of this smart search was a specific design configuration that turned out to be the "Goldilocks" solution. The researchers found that using 6 ferrite segments (a specific number of the magnetic blocks) on the receiver side offered the best trade-off. This design allowed the receiver to handle a misalignment of up to 110.87 mm (about 43 inches) while keeping the receiver's mass down to 446.39 grams (about 0.98 pounds).

To prove this wasn't just a computer fantasy, the team built a real-life prototype. They set up a wireless charging station with a large transmitter and a smaller receiver. When they physically moved the receiver around, the system held up remarkably well. The real-world test showed an effective charging radius of 105 mm (about 41 inches) before the connection dropped below the safe limit. The difference between the computer prediction and the real-world result was small, only about 5.3%, which confirms their method works.

The paper explicitly rules out the idea that you just need to add more ferrite to get better performance. While adding more segments (up to 10) did slightly improve the alignment tolerance, it made the receiver significantly heavier without giving enough extra benefit to be worth it. The study suggests that the "heavier is better" approach is a trap; the sweet spot is a balanced design where you don't over-engineer the weight.

In short, this research doesn't just say "wireless charging is hard." It provides a practical, mathematically proven roadmap for building a system that is both light enough to be useful and robust enough to handle the bumps and shifts of real life. By using a smart, step-by-step search method instead of a brute-force one, they managed to solve a complex engineering problem efficiently, bringing us one step closer to truly hassle-free wireless power.

Technical Summary: Co-Optimization of a Lightweight and Misalignment-Tolerant Concentric Multi-Ring Coupler for Wireless Power Transfer

Problem Statement
Wireless Power Transfer (WPT) systems face significant challenges in balancing rated mutual inductance matching, misalignment stability, and receiver lightweighting. While Concentric Multi-Ring Couplers (CMRC) offer strong potential for misalignment tolerance through synergistic main and auxiliary coils, their design involves strongly discrete and coupled variables (e.g., coil turns and ferrite segment counts). Traditional full-factorial Finite Element Analysis (FEA) traversal of these multidimensional discrete spaces is computationally prohibitive. Furthermore, existing surrogate-based optimization methods often focus on continuous variables, lacking dedicated strategies for strongly discrete engineering problems where co-optimizing mass and misalignment tolerance under strict inductance constraints remains a gap.

Methodology
The paper proposes a surrogate-assisted discrete-parameter optimization framework to address these challenges. The methodology consists of four main stages:

Adaptive Kriging Surrogate Modeling:
- Sampling: Optimal Latin Hypercube Sampling (OLHS) is combined with equiprobable binning to map continuous samples to discrete design variables (Transmitter main coil turns $N_{Ta}$ , auxiliary coil turns $N_{Tb}$ , receiver coil turns $N_R$ , and ferrite segments $FC$).
- Model Construction: A Kriging model using a Gaussian exponential correlation function is trained on initial FEA data.
- Adaptive Infill: A Maximum Mean Squared Error (Max-MSE) criterion is employed to iteratively add samples to the training set, specifically targeting regions of high predictive uncertainty to improve global accuracy with a limited sample budget.
Discrete Constrained Optimization Strategy:
- Beam Search-Local Refinement (BSLR): To navigate the discrete space efficiently, a beam search strategy is used. The search space is grouped by the number of ferrite segments ($FC$). Within each group, the surrogate model rapidly predicts mutual inductance and mass. Candidates are ranked and pruned based on satisfying the rated mutual inductance constraint ( $29.8 \mu H \le M(0,0) \le 30.2 \mu H$ ) while minimizing mass.
- Local Refinement: A local refinement step applies small-step neighborhood perturbations to the surviving candidates to ensure the search does not miss optimal discrete solutions near constraint boundaries.
Validation and Selection:
- Final candidate solutions undergo high-fidelity FEA verification to evaluate misalignment performance (normalized mutual inductance curves, one-sided tolerance, and effective charging area ratio).
- A two-stage decision strategy selects the final design: first screening for lightweight candidates under inductance constraints, then selecting the dominant candidate based on the trade-off between misalignment tolerance and mass.

Key Results

Surrogate Model Accuracy: The adaptive Kriging model achieved high accuracy with a small sample size (50 initial + 18 infill points). It demonstrated a Root Mean Square Error (RMSE) of 0.4 $\mu H$ and a Mean Absolute Percentage Error (MAPE) of 0.49%, outperforming a reference neural network model. The average relative error on test samples was controlled within 1.35%.
Optimization Efficiency: The proposed method reduced the required simulation count to less than 1% of the 162,288 combinations required for a full-factorial traversal.
Optimal Design Configuration: The optimization identified the configuration with $FC=6$ as the globally optimal compromise.
- Misalignment Tolerance: Achieved a one-sided misalignment tolerance of 110.87 mm.
- Receiver Mass: Reduced the receiver mass to 446.39 g.
- Trade-off Analysis: While $FC=10$ offered slightly higher tolerance (114.46 mm), it resulted in a 10% increase in mass. The $FC=6$ design provided near-optimal tolerance with significantly lower mass, representing the engineering optimum.
Experimental Validation: A prototype was fabricated and tested on a WPT platform (85 kHz, 100 V DC input, 80 mm air gap).
- The experimental effective misalignment radius was 105 mm (under a $\pm 5\%$ mutual inductance tolerance), confirming the validity of the simulation trends.
- The measured receiver mass was approximately 504 g (including auxiliary materials), with a simulation-experiment error of roughly 5.3%, attributed to geometric assembly errors and material parameter variations.

Significance and Claims
The paper claims to establish a coordinated optimization framework specifically for strongly discrete CMRC designs, bridging the gap between adaptive surrogate modeling and discrete search strategies. The primary significance lies in demonstrating that a systematic, data-driven approach can effectively co-optimize conflicting objectives (lightweighting vs. misalignment tolerance) under strict physical constraints without resorting to computationally expensive exhaustive searches. The study validates that the proposed method not only accelerates the design process but also yields a practical, balanced engineering solution that has been verified through both simulation and physical experimentation. The authors note that future work will extend this framework to include multiphysics constraints such as thermal and core loss considerations.

More like this