论文部分内容阅读
针对进行随机分析时采用蒙特卡罗计算法效率低,未能考虑参数之间相关性,导致在分析参变量对结构力学性能的影响时得到错误的灵敏度系数,以及原有灵敏度计算方法只能考虑局部梯度等问题,采用改进的混沌粒子群算法优化网络寻址结构,利用混合神经网络构建复杂结构响应的近似模型,通过相关参数与独立正态参数之间的等效变换建立符合参数相关性的随机序列对结构进行随机性分析,并根据文中提出的灵敏度度量方法计算随机变量的全局灵敏度系数。通过算例验证所提方法的可行性,且考虑参数之间相关关系得到的结构随机响应更符合工程实际情况。同时,利用所提出的随机灵敏度计算方法可以更好地反映各随机变量对结构响应的相关性和敏感性。
For the random analysis using Monte Carlo calculation method is inefficient, failed to consider the correlation between parameters, resulting in the analysis of the parameters of the structural mechanics properties of the error sensitivity coefficient, and the original sensitivity calculation method can only be considered Local gradients, and so on. The improved chaotic particle swarm optimization algorithm is used to optimize the network addressing structure, and the hybrid neural network is used to construct the approximate model of complex structure response. Through the equivalent transformation between the correlation parameters and the independent normal parameters, Random sequence analysis of the structure of random, and according to the proposed sensitivity measurement method to calculate the global sensitivity coefficient of random variables. An example is given to verify the feasibility of the proposed method. The random response of the structure obtained by considering the correlation between the parameters is more in line with the actual situation of the project. At the same time, the correlation and sensitivity of each random variable to the structural response can be better reflected by the proposed random sensitivity calculation method.