Abstract: To address the limitations of existing human-machine collaborative product design methods, which lack systematic processes and sufficient system integration capabilities in complex operating conditions, and to resolve issues such as low efficiency and inadequate utilization of data in traditional product design meth-ods, this paper proposes a human-machine collaborative product design strategy and form generation method. By employing the GBDT algorithm to analyze target product design data and user requirements, this paper co-structs a product design strategy generation method based on Llama3. Combined with the Midjourney image generation platform and Fusion 360's parametric modeling technology, a complete intel-ligent design workflow is established. Finally, a low-carbon heavy-duty transport autonomous vehicle was selected as the target product to validate the feasibility of the proposed design methodology. Compared with traditional design processes, the proposed method reduces the time required for single-round solution generation by 70.6%, achieves a relative improvement of 41.6% in product functionality, and decreases the average number of design iterations by one. The results demonstrate that the proposed approach effectively supports product concept generation, in-depth design exploration, and complex design tasks, thereby im-proving overall design efficiency and further expanding the practical applicability of human–machine co-creation in product design.