摘要
在分析了传统优化方法不足的基础上,提出了一种基于神经网络模拟退火优化的新方法.用阶梯折算法,给出了目标函数、几何约束条件关于设计变量的显式表达式,对均布载荷作用下一次等强度超静定梁进行了优化设计.
Based on the analysis of disadvantages of traditional design method, a new optimal method based on neural networks is proposed. By making use of the method of reduction by steps, an equi-strength static-indeterminant beam under distributed loads is studied. The governing equation can be explicitly expressed by the design variables as well as geometric constraints.
出处
《甘肃科学学报》
2005年第3期16-19,共4页
Journal of Gansu Sciences
基金
兰州理工大学科研发展基金(SB10200420)
关键词
神经网络
模拟退火法
超静定梁
结构优化
neural network
simulated annealing
static-indeterminant beam
optimization of structure
