Scheduling Optimization of Experimental Courses Based on Improved Genetic Algorithm

To address complex scheduling issues in high-concurrency laboratory reservation scenarios, an online experiment scheduling optimization method is proposed. This approach utilizes an adaptive genetic algorithm, incorporating the results of an improved FCFS (First-Come, First-Served) strategy as the initial solution. In this method, the improved FCFS strategy considers both the shortest service time for experiments and the remaining available time for equipment, aiming to balance the utilization rates of the equipment. The genetic algorithm employs integer encoding, where each gene represents an experimental task. Each chromosome is a combination of these genes, representing a set of scheduling schemes for the experimental tasks. The fitness function of the genetic algorithm includes penalties for time preference, overtime, and conflicts, taking into account scheduling task priority, device service time, and experimental time willingness. Additionally, an adaptive mechanism is introduced to dynamically adjust the mutation and crossover rates of the genetic algorithm, this effectively prevents the common issues of traditional genetic algorithms, such as getting trapped in local optima and lengthy iteration times.