A GLNPSO for multi-level capacitated lot-sizing and scheduling problem in the poultry industry

This paper presents a computation tool for the multi-level capacitated lot-sizing and scheduling problem in hen egg production planning with the aim of minimizing the total cost. A mixed-integer programming model was developed to solve small-size problems. For large-size problems, particle swarm optimization (PSO) was firstly applied. However, the component of traditional PSO for social learning behavior includes only personal and global best positions. Therefore, a variant of PSO such as the particle swarm optimization with combined gbest, lbest and nbest social structures (GLNPSO) which considers multiple social learning terms was proposed. The local search procedure was applied to decide the new sequence of chick and pullet allocation to rapidly converge to a better solution. Moreover, the re-initialization and the re-order strategy were used to improve the possibility of finding an optimal solution in the search space. To test the performance of the algorithm, the two criteria used to measure and evaluate the effectiveness of the proposed algorithm were the performance of the heuristic algorithm (P) obtained by comparing their solutions to optimal solutions, and the relative improvement of the solution (RI) obtained by the firm's current practice with respect to those of traditional PSO and the GLNPSO algorithms. The results demonstrate that the GLNPSO is not only useful for reducing cost compared to the traditional PSO, but also for efficient management of the poultry production system. Furthermore, the method used in this research should prove beneficial to other similar agro-food industries in Thailand and around the world.