Abstract: A control model for a variable frequency air conditioning system was established in the Simulink environment, and simulation and analysis were performed using a parameter self-adjustable fuzzy controller. Keywords: Variable frequency air-conditioner; fuzzy control; parameter self-adjustable; quantitative factor; scaling factor 0 Introduction In recent years , variable frequency air conditioners, with their energy-saving and comfort features, have gradually replaced fixed-speed air conditioners and entered thousands of households. Their control technology is also constantly being improved and perfected with the development of control theory and the deepening of understanding. Because air conditioning systems have time delay, time variation and nonlinearity, it is difficult to implement classical control and modern control methods based on accurate models; while fuzzy control is widely used because it does not require the establishment of a mathematical model of the controlled object, has strong robustness and anti-interference, and can well adapt to the control requirements of variable frequency air conditioning systems. However, conventional fuzzy control strategies based on lookup tables still have shortcomings such as unsatisfactory system rise characteristics, large overshoot, long adjustment time and large steady-state error. Therefore, people have conducted a lot of research[1]. This paper uses the Fuzzy Logic Toolbox in the Simulink environment to conduct an in-depth discussion and analysis of the parameter self-adjustment fuzzy control strategy. 1 Schematic diagram of variable frequency air conditioning system control Household variable frequency air conditioners are mostly split-type air conditioners. The control system consists of two parts: indoor control unit and outdoor control unit. The indoor control unit mainly completes the control functions such as receiving remote control signals, sampling, displaying, sending indoor temperature signals and implementing control algorithms. The outdoor control unit needs to complete the control tasks such as receiving indoor unit signals and variable frequency speed regulation of the compressor. Variable frequency air conditioners change the speed of the compressor motor proportionally according to the size of the indoor air conditioning load, and at the same time control the opening degree of the electronic expansion valve to reasonably control the flow of refrigerant, so that the cooling capacity increases proportionally. [2] The system control principle diagram of the parameter self-adjusting fuzzy control strategy is shown in Figure 1. The controller selects the error and the error change rate as input quantities and selects the output of the controller as the control quantity of the system. However, it adds two functional modules compared with the conventional fuzzy controller: one is the system performance measurement module, which calculates the data representing the system performance index from the system error e, the error change rate ec, etc.; the other is the quantization factor and proportional factor adjustment module, which is used to adjust the values ​​of each factor appropriately according to the information of the system performance measurement module. [3] [align=center] Fig.1 Control Principle Chart[/align] A large number of studies and simulation experiments show that the proportional factors Ke, Kec and Ku have a great influence on the dynamic and static characteristics of the fuzzy control system. The conclusions are as follows [4][5]: (1) The larger Ke is, the larger the overshoot of the system, which increases the adjustment time and the longer the transition process; the smaller Ke is, the slower the system changes and the lower the steady-state accuracy. (2) The larger Kec is, the greater the ability to suppress changes in system state, the smaller the overshoot of the system, and the greater the stability of the system, but the slower the system changes. (3) Increasing Ku is equivalent to increasing the total amplification factor of the system, and the system response speed is faster. However, if it is too large, it will lead to an excessively large output rise rate of the system, resulting in excessive overshoot and increasing the system settling time. When Ku decreases, the overshoot of the system decreases, but the response becomes slower and the steady-state accuracy becomes lower. When Ku takes different values ​​from large to small, the unit step response curves of the system are shown in Figure 2 a, b, c respectively. [align=center] Figure 2 Unit Step Response of Different Ku[/align] In fact, the parameter self-adjusting fuzzy control, like conventional control, has certain contradictions between its dynamic and static characteristics. It is difficult to obtain satisfactory dynamic and static characteristics at the same time using fixed parameters. Therefore, in order to improve the performance of the fuzzy controller, it is often necessary to correct the controller parameters online based on information such as the system error and error changes. However, adjusting all three parameters at the same time would make the control algorithm too complex, with a large amount of computation and high requirements for system hardware. In practical applications, the method of offline tuning of Ke and Kec and online adjustment of Ku is often used. [5] 2 Requirements of proportional factor for system response process For the typical step response curve of common control system, as shown in Figure 3, in order to analyze the different requirements of proportional factor for different response stages of the system, the response curve is divided into several segments such as oa, ab, bc, cd, de as shown in Figure 2, and explained respectively. oa segment: When E·EC<0, the indoor temperature tends to the set value, the system response speed should be as fast as possible, and Ku should take a larger value; when it is close to point a, the output is close to the steady state value, and Ku should be reduced in order to reduce the overshoot. ab segment: When E·EC>0, the actual room temperature has exceeded the set value and changes in the direction of increasing deviation, and Ku should be increased appropriately to reduce the overshoot. bc segment: When E·EC<0, the deviation begins to decrease, and the system shows a steady state change trend under control, and Ku should be reduced to avoid the pullback. In segment cd, when E·EC>0, the system experiences a callback, and the value of Ku is basically the same as in segment ab. In subsequent segments, if the deviation is not large, Ku can maintain a small stable value. From the above analysis, it can be seen that the determination of Ku is closely related to the deviation and the rate of change of the deviation. The relationship between the corrected proportional factor Ku* and Ku can be expressed by the expression Ku*=Ku(1+q). The correction coefficient q can be determined by parameter optimization or fuzzy correction table [4]. In order to reduce the amount of calculation, a fixed q value can be determined segment by segment according to the optimization result or the q value can be determined as a function of the deviation to simplify the programming and calculation of the control system. [align=center] Fig.3 Typical Unit Step Response[/align] 3 Control System Model and Response Curve The control system model established in the Simulink environment using the Fuzzy Logic Toolbox provided by MATLAB is shown in Fig.4. The proportional factor adjustment module in the model is composed of if-else control flow module [6]. The delay time of the delay element is 20 seconds. The response of the system model under the excitation of a unit step signal is shown in Fig.5 (the horizontal axis in the figure is in seconds). [align=center] Fig.4 Control System Model Fig.5 Unit Step Response of System[/align] 4 Conclusion The author's innovation: Based on the actual situation of household inverter air conditioners, this paper proposes a fuzzy control algorithm suitable for small control systems such as single-chip microcomputers by simplifying and analyzing the parameter self-adjusting fuzzy control algorithm, which is only applicable to complex system control. Compared with conventional fuzzy controllers, its computational load does not change much, so it has little impact on hardware. However, its unit step response curve shows that the system exhibits excellent characteristics in terms of rise time, settling time, and overshoot, which can better meet the control requirements. References: [1] Gao Hongyan, Wang Jianhui. Online self-adjusting correction factor fuzzy control method and application [J]. Microcomputer Information, Vol. 22, No. 5-1, 2006 [2] Liu Weihua. New technologies and progress in refrigeration and air conditioning [M]. Beijing: Machinery Industry Press, 2004. [3] Li Guoyong. Intelligent control and its MATLAB implementation [M]. Beijing: Electronic Industry Press, 2005. [4] Ma Bingchang, Ni Guozong. Application of parameter self-adjusting fuzzy controller in central air conditioning control system [J]. Automation Technology and Application, Vol. 22, No. 2, 2003 [5] Tian Guoguang. Online self-learning of fuzzy controller rules [D]. Master's thesis of Beijing University of Chemical Technology, 2003 [6] Li Ying, Zhu Boli, Zhang Wei. Simulink dynamic system modeling and simulation basics [M]. Xi'an: Xi'an University of Electronic Science and Technology Press, 2005