摘要
基于神经网络对时序问题的预测能力 ,本文提出了将混沌和神经网络相结合 ,应用神经网络来训练混沌序列的预测模型及方法 ,实现了将混沌系统快速地稳定到期望点上。理论分析和仿真结果均表明了该方法的有效性 ,且算法弹性大 ,可扩充性好 ,稍作修改后 。
Based on prediction models effection to the marginal ability of time series,using prediction model to train the chaos time series image model and methods was advanced.It made the chaos system miser to the expectable spot rapidly and stably.Both the analysis and the result of simulation make clear that,this train algorithm has resilience and the expand is large.If little amended,it will suit to the different of chaos reflex.
出处
《计算机应用与软件》
CSCD
北大核心
2004年第4期81-83,共3页
Computer Applications and Software