基于GA-ANN的高速烟支切割支撑装置结构优化

Structural Optimization of High-Speed Cigarette-Cutting Support System Based on GA and ANN

提出了一种基于人工神经网络技术和遗传算法的结构优化设计方法,在ADAMS软件环境中建立高速烟支切割装置的刚柔耦合虚拟样机模型进行仿真,得到64组实验数据构成样本,在MATLAB软件环境中,用BP神经网络构建支撑装置的数学模型,为遗传算法提供适应度函数,通过遗传算法完成最小值优化,得到的结果为原仿真样本中最小值的1/5．

A method of structure parameter optimization, based on the artificial neural network（ANN） and genetic algorithm（GA） was proposed. Based on the analysis result of the coupling model method of rigid and flexible bodies, a mathematical model of system was built with the backpropagation（BP） neural network in the MATLAB environment. And then, optimization was carried out by using genetic algorithm （GA）. The result of the optimization was satisfactory.