针对新工科背景下数据结构课程存在的教学内容滞后、实践脱节及教学模式单一等问题，提出“构、异、阶、跨”混合式教学模式，具体阐述如何通过重构知识体系、差异化任务设计、螺旋式能力培养及跨界资源整合，并结合AI驱动的智能评测、代码大模型辅助实践等创新手段，构建“理论—实践—创新—价值”培养体系。实践表明，该模式能够显著提升学生的算法工程能力、创新思维及社会责任感。

In response to issues such as outdated content,disconnection between theory and practice,and single teaching modes in the “data structures” course under the new engineering education framework,we propose a “construction-differentiation-progression-crossover” hybrid teaching model.This approach reconstructs the knowledge system,designs differentiated tasks,implements spiral competency development,and integrates cross-boundary resources.By incorporating AI-powered intelligent assessment and code-generation models to assist practical training,we establish a “theory-practice-innovation-value” cultivation framework.Practical results demonstrate that this model significantly enhances students’ algorithmic engineering capabilities,innovative thinking,and social responsibility.The course excellence rate has notably increased,with students achieving substantial competition accomplishments.