A Research Of Air Handling Unit Of The Semi-central Air Conditioning Control

Nowadays, the requirement of traditional energy becomes more and more numerous while new energy, which is still under development, needs to be popularized. Though the amount of traditional energy reserves in China is large, the per-capita amount is below world-wide level. With China’s economy rapidly developing, comfortable environment for living and working becomes a new goal which people pursue. To meet the demand of modernization, people widely set central air conditionings in modern buildings. As a result, people’s living and working environment has been significantly improved. However, the high energy consumption of central air conditioning and how to reduce the high consumption have sparked off hot problems all over the world.By reading a large amount of literature, analyzing the structure and operation principle of central air conditioning and summarizing researches and development direction of energy saving, it is found that the running efficiency of air handling unit has a significant impact on the total energy consumption of central air conditioning. Central air conditioning unit has features like multi input, multi output, nonlinear and strong coupling. It is difficult to creat an accurate model for such kind of object by traditional method. Therefore, BP neural network is applied in this thesis to creat the model of the air handling unit model to provide a reliable method to the research of control strategy. Since the air handing unit is a complex system with multi-input and multi-output, there is a complex relationship among all the input and all kinds of interference having an impact on the output, it is hard to have an ideal control by means of common control methods. Because of above reasons, fuzzy-PID control strategy is finally chosed. The theory of fuzzy-PID control, which is more adaptive and more outstanding on anti-interference ability than classic PID control strategy, is applied to adjust the parameters online.MATLAB/SIMULINK simulation environment is adopted to establish the experimental system. By multi groups of simulation experiments, it is confirmed that the neural network model of the central air conditioning of air handling unit can reflect the real running state of the system and can be used as a controlled object in simulation experiments. Meanwhile, it is also confirmed that air handling unit with multiple input, multiple output and multi types of interference can be controlled effeciently. Fuzzy-PID control strategy can meet the control requirement.