基于响应面法的柔性铰链优化设计

Optimization Design of Cross-spring Flexural Pivots Based on Response Surface Method

为提高交叉簧片式柔性铰链(简称“交叉铰链”)的第2阶固有频率,通过基于遗传算法优化的BP网络响应面映射交叉铰链的结构参数与其第2阶固有频率之间的关系,建立了交叉铰链的结构参数优化模型,并在MATLAB中编程实现。最后,对最优化计算结果进行了动静态仿真分析和试验验证。结果表明,优化模型计算值、有限元仿真值以及测试值的一致性较好,误差均在工程允许的范围内,与目前在用某型号交叉铰链相比其第2阶固有频率提高了21.7%。可见该优化设计方法是可行的,能够有效提高交叉铰链的第2阶固有频率。

To improve the 2-order inherent frequency of cross-spring flexural pivots, the BP neural network response surface optimized with genetic algorithms is used to map the relationship between its structure parameters and the 2-order inherent frequency. And a mathematic model for the structure optimization is established based on MATLAB. Then, the result of the optimization calculation is used to make the static and dynamic analysis, simulation and experiment. The results show that the uniformity of optimization model computation results, simulation results and test results is good, and the errors are within the allowable range. Compared with the cross-spring flexural pivot currently used, the 2-order inherent frequency is increased by 21.7%. So, the optimization method in this article is available for use. It can effectively enhance the 2-order inherent frequency of the cross-spring flexural pivots.