Improved Adaptive Genetic Algorithm for Course Scheduling in Colleges and Universities

Wang Wen-jing


Traditional artificial intelligence and computer-aided course scheduling schemes can no longer meet the increasing demands caused by the informatization of teaching management in colleges and universities. To address this problem, this study designed an improved adaptive genetic algorithm that is based on hard and soft constraints for course scheduling. First, the mathematical model of the genetic algorithm was established. The combination of time, teacher, and course number was regarded as the gene coding. The weekly course schedule of each class was a chromosome, and the course schedule of the entire school was the initial population. The fitness was designed according to the priority of each class, curriculum dispersion, and teacher satisfaction. Local columns between individuals were selected through the roulette principle for a variation of crossover and random columns. Iterative calculation was implemented on the basis of the default mutation and crossover rates to study the optimal course scheduling scheme. Experimental results demonstrate that the improved adaptive genetic algorithm is superior to the original genetic algorithm. When the number of iterations is 150, population evolution is optimal and the fitness does not increase. When the population size is 150 classes, the average scheduling time is the shortest. The basic, adaptive, and improved adaptive genetic algorithms are compared in terms of the number of average iterations required for convergence, maximum individual fitness, and average individual fitness. Comparison results show that the improved adaptive genetic algorithm is superior to the two other algorithms. This study provides references for the model building and evaluation of course scheduling in colleges and universities.


course scheduling; mathematical model; improved adaptive genetic algorithm

Full Text:


Copyright (c) 2018 WANG Wen-jing

International Journal of Emerging Technologies in Learning. ISSN: 1863-0383
Creative Commons License SPARC Europe Seal
Web of Science ESCI logo Engineering Information logo INSPEC logo DBLP logo ELSEVIER Scopus logo EDiTLib logo EBSCO logo Ulrich's logo Google Scholar logo Microsoft® Academic SearchDOAJ logo