A Vision‐Guided Automatic Alignment System for Improving Wireless Charging Efficiency in Electric Vehicles

ABSTRACT

Wireless power transfer (WPT) provides a convenient and contactless charging solution for electric vehicles (EVs); however, spatial misalignment between the transmitting and receiving coils can significantly degrade power transfer performance. This study proposes a lightweight vision‐guided automatic alignment system for EV wireless charging that performs post‐parking three‐dimensional micro‐alignment without modifying the WPT power electronics. The proposed system integrates a single embedded camera with a motorised XYZ positioning platform. Otsu adaptive binarisation, Canny edge detection and principal component analysis‐based geometric verification are employed to identify a square marker and estimate its spatial displacement. A proportional‐derivative (PD) closed‐loop controller then converts the detected positional errors into incremental motion commands for automatic alignment along the X, Y and Z axes. MATLAB simulation and embedded hardware experiments are conducted to validate the proposed approach. Experimental validation using a double‐D magnetic coupler shows that lateral misalignment drops from up to 8.31 cm to under 0.2 cm, while vertical clearance stabilises at 6 cm. This alignment boosts power transfer from 1443.1–1654.4 VA to ∼1923.7 VA, improving efficiency from 71.2%–81.7% to 95.0%. Robust against low light and partial occlusion with a US$100 hardware cost, the proposed system provides a cost‐effective add‐on for automated EV charging.