Abstract

As one of the ad hoc majors in urban underground space engineering, the construction and innovation of its practical teaching system is especially necessary. This paper introduces the principle mechanism of simulated annealing algorithm, proposes a K-means clustering method in collaboration with traditional particle swarm algorithm - PSK-means algorithm, overcomes the defect that particle swarm algorithm is easy to fall into local optimum, and optimizes the clustering center to get the optimal clustering result. And the improved algorithm as a teaching evaluation algorithm is combined with the practical teaching system of urban underground space engineering profession to build a practical teaching platform. Using the platform test for teaching evaluation, there are four students with high scores in learning attitude, and their scores are 43.2, 44.1, 41.5, 44.4. In the analysis of learning effect, also these four students with the best learning attitude are in the top, while the one with the worst learning attitude scores 19.2, and their learning effect is also the worst with 69.2. Based on the PSK-means algorithm and the teaching platform based on the combination of PSK-means algorithm and practical teaching system can carry out knowledge assessment statistics according to different knowledge points, thus reflecting the students' ability strengths and weaknesses in each knowledge point and reflecting the practicality of the practical teaching system of urban underground space engineering.