在机械优化设计领域针对实体结构的动态分析计算,采用BP神经网络可以提高优化收敛速度和精度,神经网络理论与传统的数值方法相结合具有重要的现实意义。 In the field of mechanical optimization design, the dynamic analysis and calculation of solid structure can improve the speed and accuracy of optimization convergence by adopting BP neural network. The combination of neural network theory and traditional numerical method has important practical significance.