The traditional evaluation mechanism of ideological education in universities faces many challenges. How to improve the quality of knowledge service work in the university and enhance instructors' skill level and effectiveness is the main problem facing the sustainable development of higher education in the new century. This article studies the innovation and expansion of the evaluation and management mechanism for university teachers based on the neural system in order to achieve effective scientific and systematic teaching methods. The experimental results show that the improvement rate of higher education academic performance under neural system learning is 75.12%, which is 17.34% higher than traditional Blended Open Learning(BOP) learning methods. Therefore, the improvement rate using the neural system learning is higher. This article targets different groups responsible for university-level teaching, thereby greatly improving the learning of college students.