Research And Design Of Class Scheduling System Based On Improved Genetic Algorithm

The scheduling problem has been proven to be an "NP-complete problem" since a long time ago,and the scheduling problem is also a combinatorial optimization problem.After studying the scheduling problem and combining it with practical business scenarios,this paper attempts to solve the actual scheduling problem using genetic algorithms.Compared with manual scheduling,when the staff manually schedules classes,they rely on the random results generated by the brain and do not have a rigorous workflow to do whatever they consider,so they often do not consider the situation well.Computers do not operate in the same way as the human brain,and do not think in a diffuse manner as the human brain does,but rather convert all information into a kind of data,which is then processed by specific rules.If a computer is used for scheduling,it is fast and capable of processing,and can quickly find solutions that meet the constraints,so that a more scientific,accurate,and practical scheduling plan can be developed,which can greatly reduce the workload of academic logistics.This paper first summarizes the common scheduling algorithms,and designs the class schedule scheduling system according to the steps of requirement analysis,outline design,and detailed design based on software engineering methods and relevant specifications.Their characteristics are analyzed,with emphasis on genetic algorithms,and the problems of traditional genetic algorithms are introduced,adjusted to meet specific requirements,exponentially limited population initialization to ensure diversity,self-adjusting variation probabilities are used,hard and soft constraints are modeled,and fitness functions are designed,while experiments are conducted on the scheme.Then the specific requirements of the scheduling system are described,the scheduling algorithm is applied to it,and the optimal teaching schedule plan is finally derived through processing steps such as selection,crossover,and variation.The resulting schedule can be better ensured to be reasonable when constructing the fitness function.The modules and overall structure of the system are designed in relation to the actual requirements,and the modules of the system are implemented using Dubbo-based microservices and finally tested.After introducing the theoretical basis,a more complete,efficient,very versatile and flexible scheduling system is implemented using a front and back-end separation technique.The use of a database to manage information solves the problem of manual entry,while effectively avoiding scheduling conflicts.