Study Of Fuzzy Neural Network On The Target-controlled Infusion Control Algorithm In Anesthesia

Anesthesia targeted control infusion is based on pharmacokinetic and pharmacodynamicand it’s significant to ensure and control the patienrs’anethesia depth by adjusting the targetdrug concentration.Due to the large anesthesia surgery individual specificity, it’s difficult tocontrol physical model parameters,although anesthesia target controlled infusion hasdeveloped, but it’s very lowly.The most improtant part of target contorl infusion is theresearch of anesthesia target control infusion control algorithm.Regarding to the problemof big individual difference and imperfect automatic control method,this thesis mainly studiesthe method of anesthesia target controlled infusion control system based on BIS index andfuzzy neural network through sorting the experience of anesthesiologists and using the abiltyof fault-tolerant,self-learning in neural network to adjust PID parameters.Set-ting up thecontrol system platform and realize the intelligent control of anesthesia target controlledinfusion.Dissuse the development of depth of anesthesia monitoring methods and monitoringinstrument form pharmacology and electrophysiological basis.In the part of pharmacologybasis of anesthesia monitoring,introduces several paremeters and instrument which commonlyused in the clinic,such as BIS (bispectral idex),Narcotrend, AEP(Auditory ElectricaPotiential),CSI(cerebral state index), Entropy;reveal the reliability and limitations of theseindex and intrusment and provide the basis of the choice of anesthesia monitoring index inthis thesis.Studies the method of anesthesia target controlled infusion control system based on BISindex and fuzzy neural network:collect BIS of patients through the BIS anesthesia depthmonitor,adjust the paremeters of PID intelligently using the experience of the fuzzy controlrule base and the fault-tolerant and self-learning ability of Neural network,realize theintelligent control of anesthesia target controlled infusion and develop the relevant controllerdesign in the matlab.The step response is quickly in the simulation of this controller,the BIScurve is relatively stable,can reach the target set.With80C52as the core,set up the control platform of target controlinfusion,includes:system architecture design,fuction planning,hardware platform structures and software program developing.The key is to discuss the design of hardware circuit and thealgorithm program,the development of interface program and instruction system.In theend,this thesis carry out related situmlation experiment to prove the stability and reliability ofthe system and realize the intelligent control of anesthesia target controlled infusion.