Study on the integration of generative artificial intelligence in the modeling design method for preschool children's toy cars

This study aims to explore the innovative role of Generative Artificial Intelligence (GAI) in reshaping traditional product design processes and to propose a GAI-based design method specifically for children’s toy car modeling. Using the modeling design of toy cars for preschool children as an example, the study first extracts a large volume of form-related image demands through a design research system. Then, the CRITIC method is applied to objectively determine the relative weight of each demand, and high-weight demands are converted into specific prompts via ChatGPT. Next, Midjourney is used to generate a reference library of toy car form images. Through a combination of the Grey Relational Analysis (GRA) method and focus group evaluations, the most appropriate design schemes that align with the intended imagery demands are selected. These schemes are further refined in terms of human-machine interaction using Cinema 4D. Finally, high-quality 3D renderings are generated using Stable Diffusion based on the selected schemes, thereby realizing a full-process design transformation from conceptual imagery to form output. Design evaluation results indicate that this method effectively generates toy car models that meet the imagery expectations of child users. The proposed GAI-based design method demonstrates significant advantages in prompt semantic transformation, product form construction, and detail optimization through multi-model collaboration. It improves both design efficiency and creativity, expands the intelligent pathways of product design processes, and offers a new paradigmatic reference for designers.