【正文】 ss,the controlled objects usually own the time delay,nonlinear,timechange characteristic and exist the category of ,if the traditional PID control method is only used,the static and dynamic output capability is not very fuzzy control as a branch of intelligence control field could be good at controlling those plicated,nonlinear systems with the characteristic of the parameter drift,the inaccurate model and time delay because the essence of fuzzy control is nonlinear control and adaptive control makes the systems more stable and more paper describes some causes of degrading the general fuzzy control quality and presents a kind of fuzzysmith control strategies for a large deadtime simulation results show that the fuzzysmith control method has some excellent features,such as high accuracy and robustnesses,which is suitable for a system with deadtime2varying parameters. Aiming at the long time delay, the paper bines the Smith predictive control with self adaptive fuzzy control, and utilizes self adaptive method to tune parameters of fuzzy controller on line. Namely, adjusting the scale and quantity factors of fuzzy controllers according to different characteristic of system presenting in different phases. Consequently, making the controller fit the diversification of object characteristic. Simulation shows that the presented method can effectively overe limitations of mon fuzzy control algorithm not fitting long time delay system and general Smith algorithm depending on precision model excessively. At the same time, the self adaptive fuzzy Smith control algorithm has strong robustness and better control performance. Although the fuzzy control theory has made considerable progress, but pared with the conventional control theory is still immature. Analysis and design of fuzzy control system has not yet established effective method, in many cases still need to rely on experience and trial and error. Attempts have been made in recent years, many people will be the concept of conventional control theory and method is extended to fuzzy control system, and the bination of fuzzy control and neural network method has bee a hot research, the bination of the two effectively promote the development of the selflearning fuzzy control.Fuzzy control is easy to obtain by the expert knowledge of language expression, that is difficult to establish accurate model can be effectively controlled and thumb can control system, and the neural network is due to the bionic characteristics more effectively use information system itself, and be able to map arbitrary function relations, has the parallel processing and selflearning ability, fault tolerance is also very strong. In the big system of integration, the neural network can be used to deal with lowlevel perception data, fuzzy logic can be used to describe the highlevel logical framework. The bination of fuzzy logic and neural network has two situations: one is the fuzzy technology for the neural network to form a fuzzy neural network, a neural network to realize fuzzy control. Both can be found in a large number of research literature.Conventional fuzzy control of the two main problem is: to improve the steady state control precision and improve the level of intelligence and ability to adapt. Can be seen from a large number of literature, in practice, to live is to the thought of fuzzy control and fuzzy reasoning, and other relatively mature control theory or method together, play to their strengths, to get ideal control effect. Such as: using the theory of Fuzzy pound control step control, PI or PID control strategy is introduced into the Fuzzy controller, a Fuzzy PI or Fuzzy PID pound control。s ability to learn and adapt to controller, the fuzzy control is more intelligent. Selftuning fuzzy controller, the parameter selftuning fuzzy control and other control methods were also significantly enhanced the ability to adapt to the environment changes. Fuzzy control with other intelligent control method of the bination of fuzzy control, such as expert fuzzy control can express and using the heuristic knowledge necessary to control the plex process and object, attach importance to the needs of multilevel and classification of knowledge, make up the rules of fuzzy controller structure is too simple, to pare a single defect, gives the higher intelligent fuzzy control. The bination of the two also can have a plex process control knowledge, and be able to more plex cases to make effective use of the knowledge. Fuzzy control based on neural network can achieve partial or all of the fuzzy logic control function. Fuzzy controller is adaptive, selforganizing, selflearning direction, makes fuzzy control parameters, the rules in the process of control to automatically adjust, modify and perfect, so as to improve the control performance of the system, to achieve better control effect, and expert system, neural network and other intelligent control technology bines bee its development trend.Since 1965, the United States automatic control expert at the university of California professor Zadeh puts forward the concept of fuzzy, since has attracted many scholars to study, the theory and method is perfected. After years of development, the fuzzy theory has bee a fuzzy set, fuzzy logic and fuzzy measure as the core of the new branch of mathematics, the fuzzy mathematics. Based on fuzzy theory is based on fuzzy logic, description and theory of human language characteristic of the fuzzy information processing. The fuzzy mathematics theory is applied to the field of automatic control and the control method called fuzzy control. The birth of the fuzzy control is the development of science and technology and society and need to be inseparable. Traditional control method when executive control, often need to obtain a mathematical model of the object, but in practice, many of the mathematical model of the controlled object is diffic