Control Method Research Of Supply Air Temperature In Central Air-Conditioning System

Energy impact of the national economy is an important factor of sustainable development,energy consumption of center system commonly occupies about 50 percent of energy consumption of the whole building. But in fact, a majority of center systems inefficiently runs,and wastes a mass of energy. The main reason for this phenomenon: First, center systems are generally in accordance with the most loads calculated adopting invariable working point, and the time of center systems running under the most loads is very short; Second,central air-conditioning controller using fixed-parameter design in general.For air supply system with time-varying,nonlinear,large Time-Delay,large inertia characteristics are unable to obtain more accurate mathematical model or models are very rough. This paper analyzes the technology of the supply air system operation and control features. The intelligent methods of self-learning ability is adopted.On this basis,in the process of air-conditioning running, set the air temperature, adopt critical sensitivity method to determine the parameter of PID parameter self-tuning. Applying negative feedback control loop regulation to air supply system, achieve the rapid system response, stability and accuracy of the contradiction between the three, allowing the system to have a more rapid rate of stability to set a target position, and can achieve prolonged stability. Using the modifier based on fuzzy neural network does not need to set up the air-conditioning system model, from the two variable values which could represent the state of air-conditioning system working to produce a compensation volume of air supply temperature, while ensuring the normal operation of the system on the basis of the air supply temperature so that reduced. The introduction of air conditioning energy efficiency ratio as the evaluation function of the system to assess the efficiency of energy use. Using BP learning algorithm of gradient descent algorithm for fuzzy neural network's parameters as amended regulation, reduce the air conditioning running and the ideal current state of the gap between running to try to achieve the desired energy efficiency standards. During testing and simulation, in the course of the work, through negative feedback control loop PID regulator, air-conditioning run smooth, has fast response and high accuracy. The modifier based on fuzzy neural network compensate air supply temperature in order to decrease it. When the system ran under the most loads, the temperature unchanged, air conditioning has been running on low power state ,at last.