基于神经网络的裂纹扩展过程预测

提出基于神经网络的裂纹扩展过程实时预测方法,其计算效率比近场动力学(peridynamic,PD)模型提高.使用PD算法获取裂纹扩展过程中的损伤云图,构建裂纹扩展数据集.基于数据集构建生成对抗网络(generative adversarial networks,GAN)模型,根据不同加载条件实时生成损伤云图,从而快速预测裂纹的扩展过程.将PD模型计算得到的损伤云图中的RGB值与相应位置处的损伤值结合,构建多层前馈神经网络模型的数据库,并使用多层前馈神经网络模型分析GAN模型产生的损伤云图,得到相应的损伤值.通过数值算例证明该方法的效率和准确性.

Prediction of crack propagation process based on neural network

The calculation time of the peridynamic (PD) model is too long, a method of real-time prediction of the crack propagation process based on the neural network was proposed to solve this problem, which improves the calculation efficiency. Based on the PD model, the damage contours of the crack propagation process were obtained, and a dataset of the generative adversarial network model were constructed and generated. Then, the damage contours were generated in real time according to differential loading conditions, thus the crack propagation process can be predicted quickly. The dataset of the multi-layer feedforward neural network model was constructed based on the R, G, B color values and the damage values from the damage contour generated by the PD model. finally, the damage values were calculated by the R, G, B color values based on the multi-layer feedforward neural network model. The proposed method is in good agreement with the calculation results generated by the PD method, and improves the computational efficiency.