摘要
文章简要介绍了BP神经网络和模拟退火算法的原理,提出了一种基于VC++环境的BP神经网络建模与模拟退火优化方法,通过模拟退火对BP网络的输入参数进行优化,实现算法接口并进行了软件设计。采用一个焊接算例进行验证,证明了该BP神经网络逼近能力强、收敛速度快,模拟退火能寻找到最优焊接参数,该软件平台具有一定的实用性。
Introducing briefly the principle of BP neural network and simulated annealing algorithm, the paper proposes a method for modelling by BP neural network and simulated annealing optimization which based on VC＋＋. Inputting parameters of BP neural network is optimized by simulated annealing to interface, a software platform is realized. A example about welding is illustrated, which shows that the BP neural network has better features such as excellent approximation and fast converge speed, simulated annealing can search for best inputting parameters, the platform has sufficient practicability.
出处
《电脑与信息技术》
2008年第2期31-33,53,共4页
Computer and Information Technology
关键词
BP神经网络
模拟退火
建模
优化
BP neural network
simulated annealing
modelling
optimization
