铝合金铸造工艺的Metropolis准则蜂群算法优化

Aluminum Alloy Casting Technology Optimization Based on Bee Colony Algorithm Fused Metropolis Criterion

为了提高铝合金铸件质量和铸造效率,提出了基于Metropolis准则蜂群算法的参数智能优化方法。分析了铝合金低压铸造原理,使用Anycasting有限元分析法模拟了铸造过程,获得了不同实验条件下的仿真值。使用单隐藏层神经网络分别拟合凝固时间、缩松面积与工艺参数间的非线性关系,得到了拟合精度较高的网络模型。以减少铸件的凝固时间和缩松面积为目标,建立了工艺参数优化模型。以传统蜂群算法为基础,提出了观察蜂对蜜源的动态选择策略和蜜源的Metropolis准则评价策略,使蜜蜂不仅选择最优蜜源,同时保留有潜力蜜源,从而给出了基于Metropolis准则蜂群算法的参数优化方法和步骤。经验证,Metropolis准则蜂群算法优化的铸件缩松面积比传统蜂群算法优化结果减少了24.16%,凝固时间缩短了3.32%,说明铸件质量和铸造效率得到了提高。经实际加工验证,Metropolis准则蜂群算法优化后的铸件显微组织分布均匀,硬度满足性能要求,可以应用于实际生产。

In order to improve aluminum alloy casting quality and efficiency,parameters optimization based on Metropolis criterion bee colony algorithm is proposed.Principle of low pressure casting is analyzed.Anycasting finite element is used to simulate casting process,and quality parameters can be gotten under different experiment condition.Nonlinear relationship between solidification time,shrinkage area and technique parameters are fitted by single hidden layer neutral network respectively,so that high fitting accuracy network model is acquired.Aimed at decreasing solidification time and shrinkage area,technique optimization model is built.On the basis of traditional bee colony algorithm,dynamic choosing strategy of observer bee and honey source evaluating method based on Metropolis criterion are put forward,so that parameters optimization method and procedure based on Metropolis criterion bee colony algorithm is provided.Clarified by trial,shrinkage area of casting optimized by Metropolis criterion bee colony algorithm is 24.16%less than it optimized by bee colony algorithm,and solidification time is 3.32%decreased,the data above indicates casting quality and efficiency is improved by optimization.Through actual casting process,micro-structure of casting optimized by Metropolis criterion bee colony algorithm is even,and casting hardness satisfies property requirement,which means it can be used to actual casting process.