基于量子进化算法的批量生产问题

在现代制造业的供应链中,生产批量计划(Lot-sizing)问题是企业经济效益最大化的关键因素之一,其主要研究在给定批量产品的需求下,确定最佳的生产方案,使得制造成本、库存成本和调整成本的总和最小化或者利润最大化。近年来的群智算法如遗传算法和粒子群算法等为解决复杂的Lot-sizing问题提供了新途径,但是这些算法易陷入局部最优。为了获得全局,将量子算法融入经典进化遗传算法中,首先,运用量子理论中独特的概率幅和量子比特对计划产量的决策变量进行编码;然后在迭代过程中,通过动态调整量子旋转角度来控制基因的变异速度,保持最优个体的基因信息,以免陷入局部最优的陷阱。Lot-sizing问题的案例实证表明,与上述常见的群智粒子群算法相比,量子进化算法的求解精度更高、收敛速度更快,可以有效解决复杂多约束的Lot-sizing问题,提高企业的生产效率。

Research on Lot-sizing Problem of Supply Chain Based on Quantum Evolutionary Algorithm

In the supply chain of modern manufacturing industry, the problem of lot-sizing is one of the key factors for enterprises to maximize their economic benefits. It mainly studies to determine the best production scheme for a given batch of products to minimize the sum of manufacturing costs, inventory costs and adjustment costs or to maximize profits. In recent years, swarm intelligence algorithms such as genetic algorithm and particle swarm algorithm provide a new way to solve the complex lot-sizing problem, but these algorithms are prone to fall into local optimal. In order to obtain the overall situation, this paper integrates the quantum algorithm into the classical evolutionary genetic algorithm. First, the decision variables of planned output are encoded with the unique probability amplitude and quantum bit in the quantum theory. Then, in the iterative process, the mutation rate of the gene is controlled by dynamically adjusting the quantum rotation Angle to maintain the genetic information of the optimal individual, so as to avoid falling into the trap of local optimization. The empirical case of lot-sizing problem shows that, compared with the common swarm intelligence particle swarm optimization algorithm mentioned above, the quantum evolutionary algorithm has higher accuracy and faster convergence speed, which can effectively solve the complex and multi-constrained lot-sizing problem and improve the production efficiency of enterprises.