Abstract: Rapid advances in artificial intelligence, together with changes in the urban-rural planning profession and the restructuring of its knowledge system, have created new demands for educational reform. Traditional planning and design education - dominated by physical-space design, expert experience, and apprenticeship-style instruction - is increasingly unable to respond to the diversification of stakeholder interests in the era of urban regeneration or to accelerating technological change. Taking the reform of residential area planning and design education as an example, this paper explores a dual-driven model of urban-rural planning and design education based on 'professional knowledge + artificial intelligence'. It develops intelligent tools, including a large-language-model-based multi-agent interaction system and a lowcode automatic scheme generator, to support a complete three-stage teaching process: preliminary planning, scheme generation, and post-evaluation. The proposed approach updates teaching content and methods while strengthening students' interdisciplinary learning, scientific reasoning, and design innovation.