基于粒子群算法的切削参数优化及其约束处理

An Algorithm for Cutting Parameter Optimization and Constraint Handling based on Particle Swarm Optimization

切削参数优化问题通常是多约束、非线性的,通过对其目标函数进行分析,发现这类问题的最优解通常位于可行域边界上。针对该问题的求解,在约束处理方法上引入了非固定多段映射罚函数法和半可行域概念,并考虑到绝对半可行域宽度导致的不同约束条件难以同步得到满足问题,提出了相对半可行域设置方法,即将半可行域宽度与各约束许用值的相对误差相对应,应用于粒子群算法实现了切削参数优化,并通过实例计算对所提出的方法进行了验证。

Cutting parameter optimization problems are usually multi -constrained and nonlinear. Through the analysis of the objective function,the optimal solution is found often located on the boundary of constraints. Accordingly,the non-stationary penalty function and the concept of semi-feasible region were introduced to deal with the constraint conditions. Since absolute width of semi-feasible region leads to that different constraints conditions cannot be satisfied instantaneously,the concept of Relative Semi-feasible Region was proposed to deal with this problem,which corresponded the widths of the semi-feasible region with the relative errors of the constraint permissible value. Furthermore,a new algorithm based on Particle Swarm Optimization (PSO) is presented for optimization of cutting parameters. Finally,validation calculation was successfully implemented through a cutting parameter optimization example.