基于仿真实验的固液混合搅拌器桨叶结构多目标优化  

Multi-objective optimization of a solid–liquid mixer blade structure based on simulation experiments

作　　者：易力力[1] 李佳[1] 李子昂 杨波[1] 何彦[1] YI Lili;LI Jia;LI Ziang;YANG Bo;HE Yan(State Key Laboratory of Mechanical Transmission for Advanced Equipment,Chongqing 400044,China)

机构地区：[1]重庆大学高端装备机械传动全国重点实验室,重庆400044

出　　处：《实验技术与管理》2024年第11期109-113,共5页Experimental Technology and Management

基　　金：国家自然科学基金项目(52375482)。

摘　　要：熔混阶段的固液混合过程作为熔铸装药工艺的重要环节,决定了固液两相多组分物料混合的均匀性,进而影响着最终的装药质量。而螺带-冲孔四斜叶搅拌桨作为该阶段的核心部件,其合理的结构参数对于提升固液混合性能具有显著影响。该文采用数值模拟、最优拉丁超立方采样法、响应面法和多目标优化算法相结合的方式对该桨叶的关键结构参数进行优化以进一步提高该桨叶的混合效率,实现熔混过程物料的高效混合。具体来说,首先,选择固液混合搅拌器的桨叶离底高度、桨叶层间距、桨叶倾角以及螺带宽度4个关键参数作为优化变量,以提高悬浮均匀度和降低功率消耗为优化目标;然后,采用最优拉丁超立方采样方法进行试验设计,并基于CFD仿真模拟获得50组样本数据;接着,根据样本数据构建桨叶结构参数与混合均匀度和功率消耗的二阶响应面代理模型,并结合优化变量约束范围建立桨叶结构参数多目标优化数学模型;最后,为了克服传统天鹰优化算法在求解后期由于种群多样性减少易陷入局部最优的问题,引入精英混沌反向学习策略和柯西-高斯变异策略,提出一种改进的多目标天鹰优化(improved multi-objective aquila optimization,IMOAO)算法,并采用IMOAO算法对桨叶结构参数多目标优化问题进行求解,以获取最优搅拌桨结构参数组合。结果表明:相比于初始设计,优化后的桨叶在功率消耗基本不变的情况下,混合均匀度提升了11.96%,同时,结合桨叶优化前后槽内的固相浓度分布情况可以看出,优化后搅拌槽内的固相浓度分布均匀性明显优于初始设计,基本实现了均匀混合,进一步证明了桨叶结构参数多目标优化方法的有效性。[Objective]As an integral aspect of the charging process of fusion casting,the solid–liquid mixing process establishes the mixing uniformity of solid–liquid two-phase multicomponent materials,subsequently influencing the final charge quality.Vital components at this stage,reasonable structural parameters of the ribbon-punch quadrangle impeller will notably improve the performance of solid–liquid mixing.However,due to the absence of efficient blade structure optimization methodologies,the design of the blade structural parameters primarily depends on prior experience at present.Therefore,obtaining the optimal structural parameters of the blade is difficult.[Methods]This study employs numerical simulation,the optimal Latin hypercube design(OLHD)method,the response surface method,and a multi-objective optimization algorithm to optimize the crucial structural parameters of the blade.First,four critical parameters of the solid–liquid mixing agitator(the blade height from the bottom,blade layer spacing,blade angle,and ribbon width)are chosen as optimization variables to enhance the mixing uniformity of solid and liquid and minimize power consumption.Subsequently,the OLHD method is adopted to design the simulation experiment,and 50 groups of sample data based on computational fluid dynamics(CFD)simulation are obtained.Then,a second-order response surface proxy model of the blade structure parameters is constructed based on the sample data.In addition,a multi-objective optimization mathematical model of the blade structure parameters is established considering the optimization variable constraint range.Finally,to overcome the issue that the traditional Aquila optimization algorithm is susceptible to falling into a local optimum due to a decrease in population diversity,this study proposes an improved multi-objective Aquila optimization(IMOAO)algorithm that introduces an elite chaos reverse learning strategy and the Cauchy–Gaussian mutation strategy.The IMOAO algorithm is utilized to solve the multi-objective op

关 键 词：固液混合 数值模拟 天鹰优化算法 多目标优化 

分 类 号：TQ051.72[化学工程] TP18[自动化与计算机技术—控制理论与控制工程]