基于多种群遗传算法的废旧手机拆卸序列规划

为解决废旧智能手机对环境的污染问题,高效率求解废旧智能手机完全拆卸序列最优方案,文章针对废旧手机完全拆卸提出一种多种群遗传算法。根据废旧智能手机的连接特点,建立废旧智能手机拆卸混合图模型,并利用该模型描述约束关系,同时考虑紧固件的连接关系;提出一种拆卸可行性评估方法,建立拆卸时间、拆卸利润、拆卸能耗等多目标优化数学模型,并利用多种群遗传算法进行优化分析;以“小米5”手机为例验证提出算法的可行性,并分析多目标不同权重系数对生成结果的影响。结果表明:与人工经验拆卸相比,优化后拆卸手机成本利润提高0.3035元/台,拆卸时间缩短30s/台,拆卸人工能耗减少0.272J;多种群比单种群收敛速度更快;相较于并行拆解,废旧智能手机等小型精密电子产品更适合于串行序列拆解。

In order to solve the problem of the environmental pollution caused by used smartphones, and efficiently solve the complete disassembly sequences of used smartphones, a multi-population genetic algorithm is proposed for the complete disassembly of used mobile phones. According to the connection characteristics of used smartphones, the disassembly mixed graph model of used smartphones is established to describe the constraint relations while considering the fastener connection relations. A feasibility assessment method for disassembly is proposed, the mathematical model of multi-objective optimization of disassembly time, disassembly profit and disassembly energy consumption is established, and the multi-population genetic algorithm is used for optimization analysis. Taking Xiaomi Mi 5 mobile phone as an example, the three weight coefficients are used to prove that compared with manual experience disassembly, the disassembly profit of each phone after optimization is increased by 0.3035 yuan, the disassembly time is shortened by 30 s, and the labor energy consumption of disassembly is reduced by 0.272 J. In the case of multiple populations, the convergence speed is faster than that of a single population. Meanwhile, compared with parallel disassembly, small precision electronic products such as used smartphones are more suitable for serial disassembly.