To improve the algorithm design and analysis abilities of computer students, it was necessary to enhance traditional learning models. By studying students' needs for project-based blended learning(PJBBL) models, a 5-dimensional student needs questionnaire was designed, and a PJBBL model was developed based on the analysis of student needs. Firstly, 100 students were randomly selected for a student needs questionnaire survey. From the survey results, it was found that the mean value of each dimension was greater than 4, indicating that students had a strong willingness to participate in the learning of the PJBBL model in algorithm design and analysis courses. Secondly, based on the requirements results, a PJBBL model was developed, which was divided into six parts: build project, scenario construction, divisions of student roles, project plan, project implementation, and project evaluation and improvement. Finally, an experimental group consisting of 36 students and a control group consisting of 33 students were used for course learning using PJBBL and traditional learning models, respectively. After completing the course, the algorithm design and analysis abilities of two groups of students were assessed using achievement to obtain test scores. After t-test analysis, it was observed that the mean value of the experimental group was 72.73, while the mean value of the control group was 64.17, and the significance value (sig) was 0.003, which was less than 0.05. Therefore, it was statistically confirmed that the PJBBL model significantly improved students' algorithm design and analysis abilities.