摘要
该文应用混沌神经网络求解信道分配问题,给出了信道分配的能量函数表达式和混沌神经网络模型,研究了判别混沌神经网络混沌特性的Lyapunov指数法,讨论了网络模型参数对网络混沌特性的影响,提出了基于混沌神经网络的信道分配算法。仿真结果表明,混沌神经网络具有复杂的瞬态混沌特性,它比Hopfield网络具有更强的搜索全局最优解的能力,和更快的收敛速度。
Chaos Neural Network(CNN)is applied to solve channel assignment problem in this paper. Energy function expression of channel assignment as well as mathematic model of CNN are put forward. Lyapunov exponent is calculated to judge whether there is chaos in CNN, and the influence on chaos characteristic caused by parameters of CNN model is discussed, then channel assignment algorithm based on CNN is proposed. Simulation results indicate that CNN has stronger ability to search global optimal solution and quicker convergence speed than Hopfield network, for its complicated transient chaos characteristic.
出处
《电子与信息学报》
EI
CSCD
北大核心
2005年第9期1429-1432,共4页
Journal of Electronics & Information Technology
基金
船舶工业国防科技应用
基础研究基金(01J3.17)资助课题