Application Of Heuristic Algorithms In Production And Operation Management Problems

Production and operation management problems, such as all kinds of facility location problems and job scheduling problem represented by transmission line maintenance plan scheduling problem, aim at rationally allocating enterprises production factors in order to efficiently provide products and services. Most of them are complex and nonlinear optimization problems, and they are difficult to solve by traditional optimization methods. Therefore, application of artificial intelligence algorithms in production and operation management problems has become a research focus for scientific researchers. The main works of this dissertation are listed as follows:Directing at the ignorance of coverage problem, economic benefit problem and insufficient development condition problem in existing models, a new mathematical model has established based on the research of the capacitate facility location problem. A new improved ant colony algorithm combined with greedy strategy was proposed to get satisfactory and optimal solutions for solving the capacitated facility location problem which contained a complex decision-making process. Experiment results verified the effectiveness of new mathematical model, and the improved ant colony algorithm improved optimizing quality and overcame the weakness that solutions are easily stuck in local optimal areas in original ones.A hybrid harmony search algorithm has proposed for competitive location problems. A greedy strategy was incorporated to the initialization process to decrease the probability of infeasible solutions. The preying behavior of artificial fish-warm was introduced into the optimizing process to improve the ability of jumping out of local optimal solutions and accelerate the convergence speed. Current global optimal solution was fully considered to guide the improvisation to generate better solutions and the search capacity was strengthened with a new gene adjustment method. Experiment results conducted on the competitive location problems demonstrated the effectiveness of mentioned algorithm.A novel competitive location model based on Analytic Hierarchy Process was presented to compensate the deficiency that single consumer group was considered only in traditional competitive, location models. The proposed model concerned the effectiveness of retail store scale, retail store price, product quality, fitment degree and traffic condition on the facilities and the influence differed from the consumer groups. Experiments on specific cases were conducted with the genetic algorithm for solving the model, and the results indicated that the proposed model and algorithm are efficiently and effectively applied to competitive location problems.To guarantee the safety and reliability of power system, a time constrainted TSP model was established based on single-group maintenance scheduling of transmission lines. Under the condition that the whole time spent on line maintenance was in the range of its maintenance time-choice during the search, this model took the importance of line into account and made the minimal economic loss based on failure rate as the target in searching. A discrete firefly algorithm, in which perturbation mechanism was designed based on variable neighborhood search algorithm, was proposed to achieve the best maintenance scheduling to avoid too quick convergence to the local optimal solutions. Results showed that the presented model and algorithm are suitable for solving transmission lines maintenance problem.In order to improve the efficiency of making maintenance scheduling of transmission lines, a new model was designed based on multiple repair groups for transmission lines maintenance scheduling. Based on single-group maintenance scheduling of transmission lines, this model considered the condition that parallel lines can not be repaired at the same time and maintenance task would be uniform distributed to make the allocation of maintenance staff more reasonable and efficient. A novel discrete artificial bee colony algorithm was presented to tackle transmission lines maintenance problem. Based on the characteristics of the problem at hand and discrete variable, transform mechanism and search strategy of bees were redefined. Results indicated that the proposed algorithm can get good results, and it is suitable for solving transmission lines maintenance problem.