基于PID型迭代学习控制算法的皮层网络节律调制  

Cortical Network Rhythm Modulation Based on the PID-Type Iterative Learning Control Algorithm

作　　者：王佶宣 邓斌 王江 伊国胜 Wang Jixuan;Deng Bin;Wang Jiang;Yi Guosheng(School of Electrical and Information Engineering,Tianjin University,Tianjin 300072,China)

机构地区：[1]天津大学电气自动化与信息工程学院,天津300072

出　　处：《天津大学学报（自然科学与工程技术版）》2022年第10期1071-1081,共11页Journal of Tianjin University：Science and Technology

基　　金：国家自然科学基金资助项目(62071324);天津市自然科学基金资助项目(19JCQNJC01200).

摘　　要：皮层网络节律是大脑动力学的普遍特征,它与认知和记忆功能密切相关.异常大脑节律通常伴随着癫痫、帕金森等疾病的产生.通过电刺激能有效调制异常脑节律,进而减轻疾病症状.目前常用的是开环电刺激调制方法,刺激能耗高且无反馈信息,难以实现个性化有效调制节律.为了实现自适应调制皮层节律,构建皮层网络的平均场模型,产生不同节律的自发振荡活动.首先,以通过PID型学习律处理的实际和期望主频的差值作为反馈变量,采用PID型迭代学习控制算法实时自适应调制40 Hz电刺激的幅值,实现了将主频从theta节律调制到alpha节律.其次,利用开环控制、PID控制与PID型迭代学习控制方法,对不同个体和相同个体的节律分别进行调制,证明PID型迭代学习控制具有最优的个性化调制结果.然后,利用数值仿真实验验证了PID型迭代学习控制能将delta、theta和gamma节律的任意主频调制到alpha节律.最后,从理论推导方面对PID型迭代学习控制算法进行了收敛性分析,证明该方法在理论上的有效性.本文研究从数值仿真和理论推导两方面,证明了PID型迭代学习控制算法能通过控制电刺激振幅,进而有效调制大脑皮层节律至alpha节律,具有较好的个性化调制结果,进一步为临床通过闭环控制治疗节律异常疾病提供了方法指导.The cortical network rhythm is a ubiquitous feature of brain dynamics closely related to cognitive and memory function.Abnormal brain rhythms cause certain disorders,such as epilepsy and Parkinson’s disease.Electrical stimulation could effectively modulate abnormal brain rhythms and further relieve the symptoms of diseases.The open-loop electrical modulation is a common method with high energy consumption,which has no feedback resulting in individual modulation difficulty.To realize the adaptive modulation of cortical rhythms,a mean-field model of the cortical network rhythm is proposed to generate spontaneous oscillation activities with different rhythms.First,the difference of the actual and desired dominant frequencies processed by the proportional-integral-derivative(PID)-type learning control law is taken as the feedback variables.The PID-type iterative learning control(ILC)algorithm is adopted to adaptively modulate the amplitude of 40 Hz electrical stimulation.Modulating the dominant frequency from the theta rhythm to the alpha rhythm is realized.Next,the rhythms of the same and different individuals are modulated using the open-loop control,PID control,and PID-type ILC.The results prove that the PID-type ILC algorithm has the best individual modulation effect.The numerical simulation demonstrates that the algorithm could modulate arbitrary dominant frequencies in the delta,theta,and gamma rhythms to the alpha rhythm.Finally,the convergence of the algorithm is theoretically analyzed,which proves its effectiveness.Both theoretical derivation and numerical simulation have proven that the PID-type ILC algorithm could modulate the cortical rhythm to the alpha rhythm effectively by controlling the electrical stimulation amplitude,which presents a good individual modulation result.These results provide methodological guidance for the clinical treatment of abnormal rhythm diseases through closed-loop control.

关 键 词：节律调制 皮层网络 迭代控制 闭环控制 

分 类 号：Q612[生物学—生物物理学]