Vol.29 Suppl.2 付冬梅等：基于Varela免疫模型的免疫控制器的设计与仿真 .203. 习能力，具有很好的抗纯滞后及参数自适应性能， 参考文献 (③)T细胞死亡率k红的取值对系统有较大影 [1]Varela F J.Stewart J.Dynamics of a class of immune networks 响，kT增大往往可使系统稳定性增强，但有可能造 Theory Biology.1990,144:93 成过免疫，从而使系统无法跟踪外界控制信号，反之 [2]Varela F J.The immune learning mechanisms:reinforcement. recruitment and their application.Computing with Biological 亦然， Metaphors.1994(1):37 (4)免疫放大系数k对系统的影响与线性系统 [3]李涛.计算机免疫学.北京：电子工业出版社，2004：7 开环放大系数的影响情况一致， [4]Sergei G.Yuri M.A simple non-linear model of immune re- (5)非线性函数M(o)和P(o)中的Km一般 sponse.Chaos Solitons and Fractals.2003,16(1):125 不宜过大，p1、p2总取负值，且p2<P1,仿真研究表 [5]Lydyard P M.Whelan A.Panger M W.Instant Notes in Im- 明其差值取0.1~0.001有利于获得好的控制效果 munology·北京：科学出版社，2001 Design and simulation of immune controller based on Varela immune network model FU Dongmei,ZHENG Deling Department of Automation.Information Engineering School.University Science and Technology Beijing.Beijing 100083.China ABSTRACI The immune network model of Varela is the second generation network model following the Jerne's.The dynamic balance relation of T and B cells without antigen effect is described in Varela immune net- work model.An improved Varela's immune network model considering antigen base on Varela s immune net- work model was founded.Some biologic characteristics can be described by the improved Varela's immune net- work model to a certain extent.Then an immune controller model was constructed on base of the improvement Varela's immune network model that is independent of system object model.With a big lag model as the con- trolled plant,the characteristic of the immune controller was studied with simulation.The result indicates that the controller has the memory,self-learning ability,the good anti-delay and self-adapting performance.The in- fluence of controller parameter change on the control system performance was also studied with simulation. KEY WORDS artificial immunity:Varela immune model;immune controller;simulation习能力‚具有很好的抗纯滞后及参数自适应性能． （3） T 细胞死亡率 kT 的取值对系统有较大影 响‚kT 增大往往可使系统稳定性增强‚但有可能造 成过免疫‚从而使系统无法跟踪外界控制信号‚反之 亦然． （4） 免疫放大系数 k 对系统的影响与线性系统 开环放大系数的影响情况一致． （5） 非线性函数 M（σ）和 P（σ）中的 Km 一般 不宜过大‚p1、p2 总取负值‚且 p2＜ p1‚仿真研究表 明其差值取0∙1～0∙001有利于获得好的控制效果． 参 考 文 献 ［1］ Varela F J‚Stewart J．Dynamics of a class of immune networks． Theory Biology‚1990‚144：93 ［2］ Varela F J．The immune learning mechanisms：reinforcement‚ recruitment and their application．Computing with Biological Metaphors‚1994（1）：37 ［3］ 李涛．计算机免疫学．北京：电子工业出版社‚2004：7 ［4］ Sergei G‚Yuri M．A simple non-linear model of immune re￾sponse．Chaos Solitons and Fractals‚2003‚16（1）：125 ［5］ Lydyard P M‚Whelan A‚Panger M W．Instant Notes in Im￾munology．北京：科学出版社‚2001 Design and simulation of immune controller based on Varela immune network model FU Dongmei‚ZHENG Deling Department of Automation‚Information Engineering School‚University Science and Technology Beijing‚Beijing100083‚China ABSTRACT The immune network model of Varela is the second generation network model following the Jerne’s．The dynamic balance relation of T and B cells without antigen effect is described in Varela immune net￾work model．An improved Varela’s immune network model considering antigen base on Varela’s immune net￾work model was founded．Some biologic characteristics can be described by the improved Varela’s immune net￾work model to a certain extent．Then an immune controller model was constructed on base of the improvement Varela’s immune network model that is independent of system object model．With a big lag model as the con￾trolled plant‚the characteristic of the immune controller was studied with simulation．The result indicates that the controller has the memory‚self-learning ability‚the good ant-i delay and self-adapting performance．The in￾fluence of controller parameter change on the control system performance was also studied with simulation． KEY WORDS artificial immunity；Varela immune model；immune controller；simulation Vol．29Suppl．2 付冬梅等： 基于 Varela 免疫模型的免疫控制器的设计与仿真 ·203·