Application Research Of Data Fusion Technology In The Supervision And Control System Of Central Air-condition System

Current world has entered an information age with developing of science and technology of modern society, and rapid and drastic development is taking place in the field of intelligent building. People are trying to create a safe, healthy, comfortable and pleasant, and energy-efficient office and living environment. Therefore, high-quality HVAC systems are also being taken more and more attention. There has been not only satisfied with the conventional heating and cooling regulatory function, but pay more attention to the indoor thermal comfort, health, high efficiency and energy saving. So the design and control of air-conditioning system is put forward higher requirements. As for the existing supervision and contro systems, in which include tens of thousands of sensors/actuators, and data processing and data transmission capabilities of sensors are extremely limited. To meet comfort, health and energy conservation requirements of modern people on HVAC system, the subject introduces the multi-sensor data fusion technology into the HVAC control system, and graphical programming language (G language) LabVIEW is used to develop the supervision and contro system.This paper analyzes the structure, basic functions and working principle of various parts in the central air conditioning system, and studies structure, functions and methods of multi-sensor data fusion. A variety of information fusion methods are compared from several aspects such as the level of fusion, operating environments, information types, presentation of information and uncertainty etc., then select appropriate fusion algorithms and apply to the HVAC system. Several modules of the supervision and contro system are designed based on prior analysis of the system structure, functional requirements, such as information acquisition module, signal display module, data storage module, and alarm module. Then graphical programming language LabVIEW is applied to realize functions of each module. Finally, develop human-computer interface of the supervision and contro system.After a comparative analysis of the fusion methods, the paper selects self-adaptive weighed algorithm for primary data fusion to remove invalid data and provide more reliable and stable data for secondary data fusion. And select fuzzy optimal decision-making algorithm as secondary data fusion method to optimize target parameters, then do the last decision-making based on PMV criteria. After the decision-making, decision results as the input of fuzzy controller are conducted fuzzy logic reasoning with actuator status information feedback to make the system more stable and more accurate.In this paper, a three-step data fusion scheme that includes self-adaptive weighed algorithm, fuzzy optimal decision-making algorithm and fuzzy logic reasoning is designed based on analysis of the characteristics and demand of HVAC supervision and controsystem as well as the principles of data fusion. Then various fusion algorithms are simulated on the LabVIEW and Matlab software platform. The results show that data fusion technology is applied in the HVAC supervision and contro system is effective and feasible, as well as can meet the comfort, energy-saving requirements.