一阶线性psi系统预测 控制及其应用研究-predictive control of first order linear psi system and its application.docxVIP

摘要系统运行于外部环境之中并受其制约和作用是普遍的现实，其中预期性强干预（PSI）会对系统的状态和性能产生显著的影响甚至导致系统的崩溃。本文论述了PSI问题的研究背景及现状，在此基础上本文给出了累积型PSI和阈值型PSI的定义，提出了两种类型PSI的初始时刻和终止时刻的估计方法。本文以一阶线性离散时间控制系统模型为主要研究对象，分别给出了理想环境中（系统可以提供任意需要的能量）累积型PSI和阈值型PSI的最优控制策略的估计定理，并分别给出了在有约束的场合中累积型PSI和阈值型PSI的最优控制策略的估计方法。通过对强干预发生机理、机制的分析针对强干预方式的分类分别进行了应用性研究。本文用一阶线性离散时间控制系统模型刻画灾害的预测和过程控制问题，将持续性将造成大的灾害的气候（降水量）变化作为预期强干预，并以此为出发点，讨论了旱灾或因长期阴雨形成的涝灾、由强降水导致的灾害等问题，给出了旱涝灾害的初始时刻和终止时刻的估计方法。本文对旱涝灾害的控制进行了量化，并给出了相应的最优控制序列估计的基本思路和方法。关键词：预期性强干预；预测和控制；强干预的初始时刻；强干预的终止时刻；最优控制AbstractTheoperationofthesystemintheexternalenvironmentbeingrestrictedandaffectedisacommonreality,prospectivestrongintervention(PSI)tothestateofthesystemandtheperformanceofthesystemmaybesignificantinfluencedthesystemorevencausedthecollapseofthesystem.Inthispaper,thebackgroundandcurrentsituationofthePSIhavebeenexpounded.ThedefinitionofcumulativetypePSIandthresholdtypePSIhasbeengivenbasedonthat.Theforecastingofinitialtimeandterminationaltimeofthetwotypeshavebeengiven,respectively.Inthispaper,thefirst-orderlineardiscrete-timecontrolsystemmodelhasbeentakenasthemeanstudyobject.Intheidealenvironment(thesystemcanprovideanyenergywhileitisneeded),theestimatetheoremoftheoptimalcontrolofcumulativetypePSIandthresholdtypePSIhasbeengiven,respectively.Andinconstraintsoccasions,theestimatemethodoftheoptimalcontrolofcumulativetypePSIandthresholdtypePSIhasbeengiven,respectively.Throughanalysisofthemechanismtoinducethestrongintervention,appliedresearchoftheclassificationofstronginterventionshavebeendone,respectively.Inthispaper,first-orderlineardiscrete-timecontrolsystemmodelhasbeenusedtodepictedtheforecastofdisasterandprocesscontrolproblem.Thecontinuousandinducinggreatdisastersclimate(rainfall)changeshavebeenusedastheprospectivestrongintervention,andtheyhavebeenusedasastartingpointindiscussingthedroughtorfloodoflong-termovercastandrainy,floodofheavyrainfallandotherissues.Inthispaper,theinitialtimeandtheterminationaltimeofthedisastersoffloodanddroughthavebeenestimated.Thecontrolofdisastersarequantifi