Application Of Fuzzy PID Control Based On Smith Predictor To Central Air Conditioner

With the development of our country's economic construction and improvement of standard of people's living, Central Air Conditioner, as an important part of intelligent building, is more and more general application in various domains. However, due to the Central Air Conditioner system's feature of higher-order nonlinear, time-varying, large delay, etc, the effect of traditional PID control algorithm, now adopted universally for Central Air Conditioner, is un-ideal, leading to the fact of that the system settling time is long and energy consumption is large. As a result, the study of Central Air Conditioner control parameters'real-time online setting can shorten the settling time of the system, greatly improve the system control precision and stability, effectively reduce energy consumption of air conditioning system.This paper discusses the structure and basic composition of Central Air Conditioner system, Established a plant model of air conditioning system's air handling link on this basis. it commences with three aspects of influence, nonlinear, time-varying and large delay, on the air conditioning control effect, then utilizes the characteristics of that fuzzy control do not depend on the controlled plant accurate model, finally on-line sets PID control parameters by fuzzy algorithm and introduces the theory of variable domain in order to improve control accuracy of fuzzy PID controller. With regard to the fuzzy controller's shortcomings, inadequate control accuracy and likely oscillating at the equilibrium point, this paper introduces the idea, after Hongxing Li's variable domain of the theory was studied, into the design, and then simplify the design of controller algorithm by using variable scaling factor to replace directly changing the domain. As a result, according to changes in the scope of systematic errors, the fuzzy controller can adjust dynamically its domain and achieve the purposes to eliminate system oscillation and improve control accuracy.For the characteristics of large time delay of air-conditioned room system, specifically, this article introduces Smith predictive control algorithms to overcome the delay on the influence of the system dynamic characteristics. Moreover, in order to solve the shortcomings of Smith predictor, which is heavily dependent on the system's accurate model, under the prerequisite condition of system stability, it is to use the Smith predictor's structure improved by C.C.Hang in the interest of the research of the paper's system model. Through the algorithm simulation to find out the problems and shortcomings and pertinently bring forward the improvements, Smith predictor with the C.C.Hang structure will have own adaptive capacity and the system finally gets hold of an ideal dynamic response, which can depend on the dynamic characteristics of the system state to adjust controller parameters.At the end of the article, to compare the proposed algorithm with others three commonly used algorithms by PID, fuzzy PID and fuzzy PID with Smith respectively, so as to get four kinds of simulation characteristic curve of control algorithm under different system conditions. And then according to each item property of the curve in the method of performance list by comparison and analysis and simulation parameters, this proposed algorithm is superior to the other three kinds of algorithms.In summary, simulation verifications show that, for those characteristics of Central Air Conditioner system, such as higher-order nonlinear, time-varying, large delay, etc, the control strategy, designed by the paper, not only possesses better control effect and dynamic characteristic and has certain adaptive ability, but also can overcome the disadvantage factors, time varying, delay, etc., to ensure the stable and efficient operation of the system.