机床横梁结构神经网络与算法优化研究
Study on Neural Networks and Optimization Algorithms of Machine Beam Structure

通过正交试验法确定网络训练样本，在MATLAB中利用神经网络对有限元分析得出的样本数据建立了激光切割机横梁结构设计参数与各输出参数的非线性全局映射，利用模糊解法得到多目标优化模型的目标函数，并通过遗传算法对激光切割机横梁进行结构优化。仿真结果表明，采用基于神经网络和遗传算法，并与模糊解法相结合的多目标优化技术，可实现激光切割机横梁结构设计的优化，使得横梁的质量和刚度得到改善。
The orthogonal experiment was used to chose the training sample data, and the sample data was calculated based on the model via finite element method in this paper.With the sample data, a non-linear mapping function from design variables to output structure features of the laser cutting machine beam was established by using the neural networks in MATLAB. As the fuzzy multiobjective optimization methods provided the objective function, the structural optimization of laser cutting machine beam was put forward via genetic algorithms. The computer simulation results showed that the multi-objective optimization technology based on the neural networks and genetic algorithms and combining the fuzzy methods, can achieve the structural optimization of laser cutting machine beam soa s to improve the mass and the rigidity of beam