The digital transformation of engineering education is leading to a deeper application of mathematical analysis tools and virtual simulation platforms in teaching. The integration of artificial intelligence into teaching and learning systems is a gradual process, but one that is conducive to the development of effective teaching and learning systems ( Sinha T. Beyond good AI: The need for sound learning theories in AIED.:Technology, Knowledge and Learning, 2025: pp. 1–15.). Despite the efficacy of automated assessment technology in domains such as assignment correction, a notable assessment blind spot persists within the Simulink simulation assessment framework in control engineering. This hindrance gives rise to challenges in teaching practice, characterized by low grading efficiency and high subjectivity. By utilizing the MATLAB data comparison function for two Simulink models, dynamic capture and analysis of simulation variables can be achieved. This is complemented by programming algorithms to create multi-dimensional scoring rules. An interactive scoring interface is developed using APP Designer, allowing the system to automatically evaluate scores and visualize data analysis through network diagrams. This approach greatly enhances assessment efficiency and objectivity compared to traditional manual scoring methods, offering valuable technical support for control engineering education and enhancing the overall teaching experience.