An AI-Enhanced Interactive Simulation Platform for Electromagnetics in University Physics

<p>Understanding electromagnetic concepts in university physics is often challenging due to their abstract nature. This paper presents an interactive simulation platform, developed with HTML5 and AI-assisted coding, that aligns with the electromagnetics chapters of the widely used textbook by Mao and Gu. The platform comprises 22 simulation modules and 22 corresponding theory note pages, all operable offline without plugins. A key visualization innovation is the use of plasma-coloured heatmaps for continuous field-strength mapping, overlaid with electric field lines, equipotential surfaces, and particle-flow animations. A dirty-flag caching and off-screen double buffering technique reduces heatmap recomputation time from ~200 ms to under 5 ms (a 40-fold speedup), enabling smooth real-time interaction on standard laptops. The note pages integrate 66 self-test questions, 76 inquiry-guided prompts, and a five-step report wizard with AI suggestions, providing scaffolding for self-directed learning. A post-use questionnaire (n=31) indicated that 93.5% of students found the simulations helpful for grasping abstract concepts, and 90.3% considered them more intuitive than static diagrams. In a controlled experiment, the platform group achieved a higher mean gain (12.22%) compared to the control group (7.78%), with a moderate effect size (Cohen's d = 0.719), suggesting practical pedagogical value. The platform is publicly accessible and offers a zero-cost, scalable tool for enhancing electromagnetics education.</p>