单晶硅电火花线切割过程分析及参数优化

发布时间：2018-06-19 18:52

本文选题：单晶硅 + 电火花线切割　；参考：《西安理工大学》2017年硕士论文

【摘要】：单晶硅具有良好的物理、机械性能,但由于材料的特殊结构导致其硬度高、脆性大,使得其加工过程很困难,目前大多采用线锯切割,但由于切割过程中存在宏观力的作用,使得切片表面的翘曲,划痕无法避免。本文研究采用电火花线切割的加工方法切割单晶硅的过程,目的在于提高单晶硅表面质量与切削效率。在分析电火花线切割单晶硅过程的基础上。将电压、脉宽、脉间、线速度作为影响电火花线切割单晶硅工艺目标的变量参数,设计单因素与多因素单目标试验,研究工艺参数对工艺目标的影响,并对影响规律进行深入分析。进行单因素实验,确定工艺目标最优时单个参数的范围。进行多因素实验由响应曲面法建立关于工艺目标与工艺参数的二阶回归模型并运用最小二乘数法拟合模型方程。对模型方程进行方差分析,确定关于工艺参数与工艺目标的模型方程的有效性。由单目标二元回归方程建立多目优化模型的目标函数,选择适合本文目标函数的多目标优化算法。第二代非支配排序遗传算的精英保留策略使得其能够优化很多种多目标模型,故使用第二代非支配排序遗传算法对本文的目标函数进行优化。通过100代进化得出关于切削效率与表面粗糙度的帕累托解集,将理论解与实验值进行对比,验证理论解的可行性。第二代基于强度帕累托进化算法由于其独有的选择策略与进化策略使得其多目标解集比其他算法的得出的解集更平稳与均匀,故将此算法应用到关于表面粗糙度与切削效率的多目标优化。将第二代基于强度帕累托进化算法得出的帕累托解集与第二代非支配排序遗传算法得出的帕累托解进行对比验证。
[Abstract]:Monocrystalline silicon has good physical and mechanical properties, but because of its high hardness and high brittleness due to the special structure of the material, it is very difficult to process. At present, most of the monocrystalline silicon is cut by wire saw, but because of the macroscopical force in the cutting process, Make slice surface warp, scratches can not be avoided. The process of single crystal silicon cutting by WEDM is studied in this paper. The purpose is to improve the surface quality and cutting efficiency of monocrystalline silicon. Based on the analysis of EDM single crystal silicon process. Voltage, pulse width, interpulse and linear velocity are regarded as variable parameters that affect the target of WEDM single crystal silicon process. Single factor and multi factor single objective test are designed to study the influence of process parameters on the process target. The influence law is analyzed deeply. Single factor experiments were carried out to determine the range of single parameters when the process target was optimal. The second order regression model of process target and process parameters was established by response surface method and the model equation was fitted by least square multiplier method. The variance analysis of the model equation is carried out to determine the validity of the model equation about the process parameters and process objectives. The objective function of the multi-objective optimization model is established from the single-objective binary regression equation, and the multi-objective optimization algorithm suitable for the objective function in this paper is selected. The elite retention strategy of the second generation undominated sorting genetic algorithm makes it possible to optimize many kinds of multi-objective models, so the second generation non-dominated sorting genetic algorithm is used to optimize the objective function in this paper. The Pareto solution set on cutting efficiency and surface roughness is obtained through 100 generations evolution. The theoretical solution is compared with the experimental value to verify the feasibility of the theoretical solution. Because of its unique selection strategy and evolutionary strategy, the second generation Pareto evolutionary algorithm based on strength makes its multi-objective solution set more stable and uniform than the solution set obtained by other algorithms. Therefore, this algorithm is applied to the multi-objective optimization of surface roughness and cutting efficiency. The Pareto solution set of the second generation based on the intensity Pareto evolutionary algorithm is compared with the Pareto solution obtained by the second generation non-dominant sorting genetic algorithm.
【学位授予单位】：西安理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ127.2

【参考文献】