基于遗传算法的车间设备虚拟布局优化技术研究

Research on optimization technology of facility virtual layout based on genetic algorithms

根据车间布局设计的最小物流费用原则 ,建立车间设备布局优化模型 ,并对该优化模型进行遗传算法求解 .借鉴车间设备布局的块树表示法 ,设计了基本块、分割线和方向的混和染色体 ,并给出解码方法 .根据车间布局优化的特点 ,改进选择算子、交叉算子和变异算子等遗传操作算子 ,设计基于块树的分割线有效性检验算法 ,并融合了选择交叉算法的优点 ,有效地避免了遗传过程中无效个体的大量出现 ,保证了算法的有效性和高效性 .最后 ,给出了一个 1 2台设备车间布局仿真算例 ,结果表明 ,算法快速地在进行到 1 70代左右时达到了收敛 ,并采用虚拟现实技术将计算结果形象地呈现出来 .本文所提出的方法能够很好地解决车间设备虚拟布局问题 .

Following the rule of minimal workflow cost, an optimization model of facility layout is proposed and genetic algorithms is used to solve this model. Hybrid chromosome of basic block, slicing line and orientation is designed using block tree method. Selection, crossover and mutation operators are improved to avoid great emergence of invalid individuals. Slicing line validity verifying algorithm is designed based on block tree and the benefit of selection-crossover algorithm is adopted. An example of 12 facility layout optimization is given and the algorithm is rapidly converged at about 170th generation. The result is presented using virtual reality technology. The technology presented in this paper is proved efficient.