【作者】 贾少青；

【导师】 陈平；

【作者基本信息】 山东理工大学 ， 计算机应用技术， 2007， 硕士

【摘要】 利用计算机技术对输液过程进行监控实现输液过程的自动化和信息化管理具有无可比拟的优势。由于输液系统是一个非线性、时变、滞后的复杂控制对象,采用传统的控制策略难以取得良好的控制和实时监控效果,需要采用新的智能控制算法来实现迅速准确的实时监控。模糊系统和人工神经网络都是智能控制理论的重要发展方向,虽然在概念内涵上有着明显的不同,但都是为了处理实际系统中的不确定性、不精确性等引起的用传统控制方法难以控制的问题。神经网络是从生理结构上模仿人脑神经元的功能,在认知处理上比较擅长,而模糊逻辑则是模仿人脑的逻辑思维机理在技能处理上比较擅长。因此神经网络和模糊逻辑各有所长存在着互补性,将两者有机的结合融合各自的优点将可得到性能更好的控制器-模糊神经网络控制器。应用它研究复杂非线性系统具有重要的理论意义和实际应用价值,它已广泛的应用于自动控制领域。本文提出了一种基于模糊神经网络模型的控制方法,该方法结合了模糊逻辑控制和神经网络的优点,利用神经网络记忆模糊规则,实现模糊控制,并利用误差反向传播算法实现当控制对象参数变化时控制器的在线自学习。从而能快速、准确的对实测数据进行综合、分析和处理,达到智能实时控制的效果。本系统的下位机以MSP430单片机为控制核心,结合输液传感器、电动执行器、液晶显示器、键盘和通信接口等,实现了对输液过程的自动监控。更多还原

【Abstract】 The using of computer technology has matchless superiority in transfusion monitor to realize automation and infomationization manage. Because transfusion system is nonlinear, time-varying and stagnant, conventional control method can hardly obtain benign real time monitor effect. New intelligent control method should be adopted to achieve real time monitor quickly and correctly.Fuzzy system and artificial neural network are the important developing directions of the intelligent control theory. Though they are different from conception and connotation, they can both be used to solve practical system problems caused by unaccuracy which can not be solved by traditional control methods. Neural Network can simulate the capacity of human brain nerve cell from the physiological structure. It is good at learning process. While Fuzzy logic can simulate the logic thought mechanism of human brain.It is good at skill process.They both have their advantages and disadvantages. We will get better controller if their advantages can be combined together. The combination of the fuzzy logic and artificial neural network is the fuzzy neural system. It has the important theory meaning and the actual application value to be used in the research of complicated nonlinear system. It has been widely used in the automatic control field.This article introduces a control method according to fuzzy neural network, which integrates the excellence of fuzzy logical control and neural network. It uses neural network to remember fuzzy rule and achieve fuzzy control, and uses error back propagation to realize online self-learning when control object parameter has changed. This method can colligate, analysis and dispose data quickly and exactly to achieve intelligent control.The monitor of this system has several pieces of main module, including the MSP430 MCU, the transfusion sensor, the electric apparatus, the LCD screen, the keyboard and the serial communication, etc. It realizes automatic monitor in the transfusion process.更多还原