配气凸轮优化设计的惩罚函数法和增广拉格朗日乘子法

Sequential Unconstrained Minimization Technique and Augmented Lagrange Method for Optimal Design of Engine Inlet Valve Cam

将约束最优化问题的间接解法中有代表性的惩罚函数法和增广拉格朗日乘子法应用到配气机构五项式凸轮型线的最优化设计中 ,优化设计结果表明混合法 (内点法 )的求解精度最高 ,外点法的收敛速度最快 ,而增广拉格朗日乘子法的求解精度较高且收敛速度也较快 ,更适于用来求解此类问题 .

SUMT (Sequential Unconstrained Minimization Technique) and ALM (Augmented Lagrange Method) methods, which are the representative methods among the indirect solutions of constrained optization problems, are applied in the shape optimial design of engine valve cams. The optimized results show that the optimal result of blending method (inner point method) is best, the convergence speed of outer point method is fastest whereas the optimal result of ALM is better and its convergence speed is faster, therefore ALM is more fit for solving this kind of problems