摘要
讨论了Hopfield神经网络算法在优化计算中的应用,提出了一种暂态混沌神经网络模型,把混沌动力学与收敛动力学相结合,使网络逐渐由混沌神经网络向Hopfield网络过渡,达到控制混沌的目的,并且提供一个在全局最优解附近的初值,然后用Hopfield网络得到最优解,有效地解决了Hopfield网络的局部极值问题.仿真结果表明算法对于初始值是稳健的,并且具有很强的克服陷入局部极小能力.
This paper is concerned with the application of Hopfield network in the optimization computation. A transient chaotic neural network is presented, combining with chaotic dynamics and converging dynamics together, makes the neural network trans it gradually to Hopfield network. By introducing converging factor, gets the aim of controlling chaos, provide initial value of Hopfield network that is near to the global optimal solution, the gradient property of HNN is used to reach stable point, and solves the problem of local minimum. Simulation results showed that such an algorithm, which is robust with initial states, can avoid getting stuck in local minima and has better convergence property as well as time property.
出处
《哈尔滨商业大学学报(自然科学版)》
CAS
2006年第1期39-41,45,共4页
Journal of Harbin University of Commerce:Natural Sciences Edition
基金
黑龙江省自然科学基金资助(编号:F2004-04)
黑龙江省青年科学基金资助(编号:QC03C03)