基于储备池计算下的螺旋波波头轨迹预测
- 分享到:
摘要：
螺旋波是自组织结构的一种特殊斑图，其形成和维持不需要外部干预，在生物学、化学和物理学的系统中都可以观测到。波头是螺旋波斑图的相奇点,控制住波头就能控制螺旋波。因此，如何通过少量的螺旋波时空状态数据去预测波头轨迹演化行为，是反应扩散系统研究中的一个重要问题。构建了一个混合储备池预测模型，,将FitzHugh-Nagumo系统中单波头模式的时序数据和相应波头位置作为训练集；将上述数据输人，对混合储备池模型进行训练；然后利用训练好的混合模型，对螺旋波系统后续波头的演化轨迹进行预测。结果表明该混合模型可以实现螺旋波波头轨迹的长期精确预测，为反应扩散系统相奇点的预测提供了新的研究视角与方法。
Spiral waves are a special form of self-organized structures that are formed and maintained without external intervention reflecting the complexity of nonlinear system interactions, and have been observed in systems of biology, chemistry, and physics, The tip is the phase singularity of the spiral wave pattern, so controlling the tip can control the spiral wave.Therefore, how to predict the tip trajectory from a small amount of spiral wave spatio-temporal state data is an important problem in the reaction-diffusion systems study. In this paper, a hvbrid reservoir computing prediction model is constructed, in which the time series data of the single tip mode and the corresponding tip positionsin the FitzHugh-Nagumo system are used as the training set; the above data are inputted to train the hybrid reservoir computing model; and then the trained hybrid model is utilized to predict the evolution trajectories of the subsequent tips in the spiral wave system, The results show that the hybrid model can realize the long-term accurate prediction of the traiectories of spiral tip, which provides a new research perspective and method for the prediction of phase singularities of reaction-diffusion systems.
作者:
崔晓华，陈叶羽格，郑才溢
Cui Xiaohua,Chen Yeyuge,Zheng Caiyi
机构地区：
北京师范大学系统科学学院；北京师范大学(珠海)文理学院
引用本文：
崔晓华，陈叶羽格，郑才溢。基于储备池计算下的螺旋波波头轨迹预测[J].学报(自然科学版)，2025,53(6) :8-13.(Cui Xiaohua, Chen Yeyuge,Zheng Caiyi.Prediction of spiral wave tip trajectories under reservoir computing[J].Journal of Henan Normal University( Natural Seience Edition),2025,53(6):8-13.DOI: 10.16366/j.cnki.1000-2367.2024.11.15.0004.)
基金：
国家自然科学基金
关键词：
储备池计算；螺旋波波头；轨迹预测
reservoir computing; spiral wave tip; trajectories prediction
分类号：
O415；N93；TP181
