基于群智能算法的设备布局离散优化研究

Discrete optimization problem of machine layout based on swarm intelligence algorithm

针对单向环形设备布局设计问题,建立了新的数学模型。利用多维实数编码及映射方法,将连续粒子群优化算法应用于求解设备单向环形布局问题,提供了求解离散优化问题的新思路。利用遗传算法中的杂交策略扩展了粒子群优化算法,提高了粒子群优化算法性能。建立了问题的图结构描述,以引入蚁群系统算法搜索优化解。给出了两种方法的求解步骤。通过实例计算和结果比较,说明该算法能有效地解决此类离散优化问题,降低成本,提高效率,所得解质量较高,有很好的实用价值。

To deal with the layout design problem of machines in a unidirectional loop manufacturing system, a new mathematical model was constructed. By adopting the specific multidimensional-real-coding and mapping method,the continuous Particle Swarm Optimization （PSO） algorithm was applied in solving layout design in a unidirectional loop. A novel particle presentation for the discrete optimization problem was proposed. Hybrid strategy of Genetic Algorithm （GA） was used to extend PSO to improve its performance. A particular graphic structure was established to describe the problem, and Ant Colony System （ACS） algorithm was introduced to search optimization solutions. Detailed steps of these two solutions were specified. Simulation results demonstrated that this method could effectively solve discrete optimization problem with lower cost.