立式振动磨关键结构参数的多目标优化设计
Multi-objective Optimization Design on the Key Structure Parameters of the Vertical Vibration Mill

以获得立式振动磨最佳粉碎效果为目的，建立了立式振动磨参数化仿真模型，采用拉丁超立方方法生成样本数据，构建了立式振动磨响应面近似模型且其精度达到可接受水平。以振动强度、撞击力最大，研磨介质体积最小为目标函数，采用多目标遗传算法NSGA-Ⅱ对响应面近似模型进行多目标优化设计，得到Pareto最优解集和Pareto前沿图，为立式振动磨结构设计和参数选择提供理论指导。最后，通过立式振动磨的优化算例说明了该方法的有效性。近似模型方法能提高优化效率，且该研究也为其他复杂模型的设计优化问题提供了一种方法。
In order to obtain the best pulverization effect, a parameterized simulation model for vertical vibration mill was built in this paper. Latin hypercube method was used to generate sample data, the response surface approximation model (RSM) was established and its accuracy satisfies the accepted level. A multi-objective optimization design on RSM was carried out by multi-optimization genetic algorithm NSGA-Ⅱ, and its objectives were to maximize vibration intensity and impact force and to minimize the volume of mill medium. The Pareto optimal solution set and Pareto frontier were obtained, which provides guidance for mechanical design and parameter selection of the vertical vibration mill. Finally, an optimization example of vertical vibration mill was given to show its effectiveness. Approximate model method can improve optimization efficiency and it also provides a method for design and optimization of the similar complicated models