论文部分内容阅读
模态试验结果中包含试件不同状态不同阶次的频率和振型信息,对动特性模型修正时需要建立多个目标函数。提出一种基于Pareto最优的Timoshenko梁模型修正方法。通过全局交叉和变异操作,求解群体的非劣解集,定义新的个体适应度评价方法,并根据拥挤距离对非劣解集进行排序,对动特性模型进行修正。仿真结果表明,采用该方法能够获得较高的精度,Pareto最优前沿收敛,且形状为非凸,与试验结果相比,修正后的模型一阶频率偏差不超过1%,二阶频率偏差不超过6%,振型与试验结果一致。
The modal test results contain information about the frequencies and mode shapes of different orders of different states of the specimen, and multiple objective functions need to be established when the dynamic model is modified. A modified Pareto-based Timoshenko beam model is proposed. Through global crossover and mutation operation, the non-inferior solution set of population is solved, and a new evaluation method of individual fitness is defined. According to the congestion distance, the non-inferior solution set is sorted and the dynamic characteristic model is modified. The simulation results show that the proposed method can obtain high precision, Pareto optimal front convergence, and non-convex shape. Compared with the experimental results, the first-order frequency deviation of the modified model does not exceed 1%, and the second-order frequency deviation is not More than 6%, mode shapes consistent with the test results.