Research On Application Of MAS Technology For Energy-Efficient Control In Central Air Condition System

Automatic control of central air condition is the main component in the buildingautomation system. The control quality of central air conditioning system not only determines comfort degree of air condition places, but also has a direct effect on the level of energy consumption of building. The central air condition's energy consumes occupy about 50%-70% of whole building; it is the point of the building automation economy energy for central air condition automatic control. Because the central air condition system is very huge, the respond speed slower, the fall behind phenomenon is more serious, a lot of difficulty increased of the stability of machine set circulation. The current central air condition automatic control system is still adopted the traditional control technique almost, the adaptability between work and environment variety for traditional control technique is bad, controling inertia bigger,the result of economy energy ignores to think. On the other hand, the energy consumption of central air condition is the sum total of all the plant, it is possible to implement whole energy-efficiency of central air condition through take the teamwork of the plant into account from aspect of system. Then it has important theories value and the actual meaning very much for increasing stability of the machine set'circulating, dependable, economy energy and push the application development of artificial intelligent technology.Multi-Agent system embodys well the cooperativity of human being by virtualizing human rational behaviour to describe interaction between Agents and is specially applied to central air condition system. This dissertation is under the application demand of energy-efficiency of central air condition, introduces MAS to automatic control system of central air condition, and establishes a MAS based Model for Intelligence Control System of Central Air Condition (MAS-ICSCAC). MAS-ICSCAC implements the autonomy of intelligent sub-systems and cooperation, coordination among them, by realizing autonomy of agents and cooperation, coordination among agents. MAS-ICSCAC acts upon FIPA specifications, adpots point to point method as communication mode, and the coordination mechanism combining centralizing coordination and distributing coordination, MAS-ICSCAC fulfils the cooperation among agents by the model of using"request-service"task allocation. The platform of developing MAM-ICSCAC is JADE. The dissertation realizes the prototype by using Agent-oriented programming method, which has the characteristics of simplicity, flexibility, scalability, and can be provided with good interoperability.Through the test of MAS-ICSCAC prototype, the dissertation proves the superiority of MAS technological that this subject study and the correction of application technology route, tools and methods. Meanwhile, MAS-ICSCAC prototype is simulated preliminarily with EnergyPlus, combined with graphic user interface and the third-party tools, analyzes the results of simulation. The simulation results shows that MAS-ICSCAC with new architecture, can adapt well to the dynamic characteristis of central air condition, overcome drawbacks of traditional control methods, and implements the globally energy-efficiencyt of central air condition, compared to traditional automatic control system of central air condition with better energy-efficiency.