The Realization of Improved Genetic Algorithm on University Class Division Problem

摘要

The class division problem in university is a NP optimization problem with multiple constraints. At present, by using traditional methods, most of the solutions cannot balance the time and the multiple expected targets. In the context of university class division, the goal is to accomplish massive students class division in a highly efficient and intelligent way, also the original targets need to be met. Considering this background, the strategy of class division does not have to reach the optimal value, but needs to find a relatively good result under the premise of ensuring the implementation speed. This paper proposes a creative class division method based on an improved genetic algorithm. Based on the speed requirement, it optimizes traditional genetic algorithm's difficult to find the optimal solution problem. At the same time, based on the algorithm proposed by this paper, a real system is developed and delivered online. Under actual circumstances, the system achieved the classification of 400 students in only 4.37 seconds.