The Coke Oven Pushing Plan Optimization Scheduling Research Based On The Ant Genetic Fusion Algorithm

In the metallurgical production, coke is one of the essential rawmaterials. Its quality directly affects the quality of metallurgicalproducts. Pushing coke is a vital part of coke production, reasonablearrangements of coke-pushing plan plays an important role in thestable production of coke oven. Coke-pushing plans in the face ofunusual circumstances are all artificial arrangement, the artificialarrangement for the coke-pushing plan is likely to cause a waste ofresources and equipment damage. It is bad for the quality of coke andthe stable operation of coke oven. Therefore, the key of ensuring andimproving the quality of coke is to make a reasonable coke ovenoperations plan to achieve its optimal operation.All kinds of intelligent algorithm was gradually applied in solvingthe optimization scheduling problems, the most widely used is AntColony Algorithm and Genetic Algorithm. Therefore, this paperpresents a kind of fusion algorithm based on the Ant ColonyAlgorithm and Genetic Algorithm. It used the Genetic Algorithm toobtain the optimum solution as the initial pheromone of Ant ColonyAlgorithm in the early stage, which overcomes the problem of slowaccumulation rate of initial pheromone of the Ant Colony Algorithm.Then, this paper uses the Ant Colony Algorithm to obtain the optimalsolution which is selected, crossover and mutation. It improves theglobal optimization capability of Ant Colony Algorithm to overcomethe problem of easy convergence to the local optimal solution. At the same time, the integration algorithm is tested on the TSP problems andcompared with traditional Ant Colony Algorithm and GeneticAlgorithm. The result shows that the initial pheromone distribution ofAnt Colony Algorithm has been changed, the algorithm searchesstability and global optimization capability have been improved. Thealgorithm has better solution performance at the aspects of time andsolution efficiency.Aiming at the abnormal conditions that affects coke ovenoperation in the process of the coke-pushing, the optimal schedulingmodel is proposed for minimize the total cost. And the fusionalgorithm was applied in this question. In order to verify theeffectiveness of the algorithm，we take a steel plant coke oven as anexample to simulate. Compared with the Ant Colony Algorithm, thisalgorithm has better convergence to global optimization. And theoptimal solutions in every generation have gradually approaching toglobally optimal solution. Meanwhile, this algorithm has less iterationto get global optimal solution and realizes the optimization of cokeoven operation plan.