Research On Hot-rolled Steel Job Scheduling System Based On Improved Genetic Algorithm

The production process of steel industry is a typical hybrid system; it has continuous andintermittent characteristics. In the actual production, developing a reasonable productionscheduling plan to coordinate well the production process of materials, equipment, maintenancetime, delivery date and other factors can directly affect the core competitiveness of enterprise.The layout of rolling planning and scheduling directly affect the production of the productsquality and production cost, whether the contract period in accordance with the contract tocomplete the production, it has deeply impact on the development of enterprises, so construct areasonable and efficient production scheduling plan is of great significance to improve theproductivity of the enterprise as well as economic benefits.This paper takes the actual hot-rolled production line of hebei iron and steel group as thebackground. In the paper, the iron and steel production process flow, the basis of rolling plan andconstraints are discussed. Designing improved genetic algorithm to obtain the optimal rollingscheduling plan can solve many problem, such as more complex production process, small batchorders for products of hot-rolled steel production process, and so on. the genetic algorithm hasthe premature convergence, in order to avoids this disadvantages, the author proposed improvedthe genetic algorithm with tabu search algorithm and effectively improve the global searchability of the algorithm. However, the improved genetic algorithm with tabu search accuratelydepending on the initialization parameters, further design the strategy that adjusts the algorithmparameters adaptively. According to the average fitness in the population of individuals, thealgorithm parameters adaptive strategy tries to dynamically adjust the crossover probability, thelength of the taboo list, et al. The improved algorithm effectively reduces the dependence of thealgorithm initialization parameters and improves the efficiency of the algorithm search. Thehierarchical three group parallel strategy based on self-adaptive tabu search genetic algorithmensures the ability of global search algorithm, meanwhile effectively speeds up the overall convergence rate directly. In addition, the simulation experiments demonstrate that the proposedimproved genetic algorithm has high reliability and good performance.Based on improved Genetic algorithm in this paper, a job-shop of hot-rolled steelscheduling system model is designed and implemented into existing manufacturing executionsystem, a lot of production data further shows that the optimal management system improvedproduction efficiency and reduced production costs.