An Improved Genetic Algorithm for College Course Scheduling

The course scheduling is a NP-complete problem. At present, various intelligent optimization algorithms have provided many feasible solutions to the course scheduling problem in colleges and universities, with differentiated advantages and disadvantages. This work tries to design a general and efficient algorithm to solve the large-scale course scheduling. In particular, we analyze and model the problem of course scheduling in colleges and universities, and put forward the improved genetic algorithm to solve the problem of large course scheduling under various constraints. First, the teaching task number is stored in a two-dimensional time-class matrix, which represents the information of teachers, and a two-dimensional classroom matrix is established to store the classroom information of the class. Next, we apply the improved genetic algorithm to cross and mutate the time-class matrix to obtain new individuals, thereby adjusting the corresponding classroom matrix. Then, the excellent individual is selected from the parent and child generation, and iterated until the optimal individual is produced. Finally, experimental results show that the convergence speed of the proposed algorithm is faster and higher fitness values can be obtained.