This paper combines the necessary functional requirements for teaching system generated by teaching activities in the context of mobile Internet, designs the general framework of the system, users and their rights management, and constructs a set of teaching system. Subsequently, the traditional PSO algorithm is introduced, and the processing scheme of the scheduling problem is defined as particles to form an initial particle swarm, while the particle swarm position in the algorithm is updated by drawing on the crossover idea of the genetic algorithm, so as to optimize and obtain the scheduling algorithm based on DPSO. Then we test the teaching system of this paper from three levels of pressure bearing, response delay and stability performance to ensure the operating environment of the scheduling algorithm of this paper. The courses of three colleges of a university are used as experimental data to analyze the performance of the scheduling algorithm in this paper. In the comparison of course arrangement in different colleges, the adaptability of this paper’s scheduling algorithm is above 0.900, while the highest adaptability of manual scheduling is only 0.8147, which indicates that compared with manual scheduling, this paper’s scheduling algorithm is able to make a more reasonable course arrangement.