The Application Of Multivariable Central Air-conditioning Control System

With the development of intelligent building technology, central air-conditioning system has become the important part of intelligent building system, which has the features of larger use and more complex structure. The problem of energy consumption was taken seriously by every country, become a global issue. Air conditioning control system’s performance depends on the control strategy of control system, variable air volume air conditioning has been applied in more and more projects. In order to make the central air-conditioning system keep in the best working condition, to achieve the purpose of saving energy, the central air-conditioning system control idea and strategy has been researched in this paper.The author has done the following work:1. Taking an exhibition hall as the background, the author designed building automation system, which has used the bus type network structure. Air conditioning system used the variable air volume control system structure, which has modular and structural characteristics.2. The author has expounded the concept of the variable air volume air conditioning system, and working principle. And discussed several end control modes and characteristics of VAV central air conditioning in the form of chart. At the same time, according to the different performance requirements, the author has discussed the VAV system of a variety of single heat, cold or simultaneous cooling and heating, which made choices for different regions, different projects, and load calculation and attention have been presented.3. In view of the delay, nonlinear, multivariate characteristics of central air conditioning system, and the shortcomings of traditional PID control and the practical application of the neural network control, the RBF neural network and the combination with predictive control strategy have been put forward. Based on the characteristics of air-conditioning systems, multivariable neural network predictive controller has been used, which has been improved more greatly than the previous two kinds of control algorithm in central air conditioning system, and shorten the transition time, and reduce the overshoot. As a result, the variable air volume air conditioning system rapidly and accurately meets the control requirements, saving a lot of energy.4Air-conditioning system has the characteristics of multivariable, strong decoupling. For that the paper proposed air-conditioning decoupling control strategy. In order to achieve a good comfort, the use of temperature and humidity at the same time control strategy. Temperature and humidity as a two input two output control system, neural network decoupling controller design, the central air-conditioning control system in the comfort and energy efficiency to make further improvements5. Air-conditioning system mathematical model was established, and the simulation of PID control, the neural network control and RBF neural network predictive control have been carried on, of which this three kinds of control strategy were compared and analyzed. The simulation comparison was made for neural network decoupling control between the temperature and humidity, which obtained the good decoupling effect and provided the basis for the application in practical project.