论文部分内容阅读
建立了金属簧片阻尼隔振器有限元参数化模型和数学优化模型。通过优化超拉丁方实验设计,分析设计变量对隔振器抗冲击性能影响的变化规律。对隔振器进行优化,采用基于近似模型的组合优化策略:先用多岛遗传算法探索全局最优解所在区域,再用序列二次规划算法逼近精确最优解。建立RSM和Kriging近似模型,并进行误差分析,其中Kriging模型误差平均值为0.127 53,认为其拟合本隔振器非线性响应可靠有效。经过综合优化策略,承载结构最大加速度响应较初始设计时最大加速度响应下降了56.08%,结果表明此优化策略实用、可行。
The finite element parametric model and mathematical optimization model of metal reed damping vibration isolator are established. The influence of design variables on the impact resistance of vibration isolators was analyzed by optimizing the experimental design of super latin cube. To optimize the vibration isolator, the combinatorial optimization strategy based on the approximate model is adopted: the multi-island genetic algorithm is used to explore the region where the global optimal solution is located, and then the sequence quadratic programming algorithm is used to approximate the exact optimal solution. The RSM and Kriging approximation models are established and the error analysis is performed. The average error of Kriging model is 0.127 53, which is considered to be reliable and effective in fitting the nonlinear response of the vibration isolator. After the comprehensive optimization strategy, the maximum acceleration response of the bearing structure decreased by 56.08% compared with the maximum acceleration response of the initial design. The results show that this optimization strategy is practical and feasible.