Study And Implementation Of Intelligent Building Energy Management System Based On Wireless Sensor Networks

Known to all, energy is one of the four survival problems that humanbeings face. Currently, energy situation is severe, so it's imperative to save energy and reduce consumption. The buildings energy consumption makes a considerable proportion in the total society energy consumption, most of which is unreasonable, so there is a big space for energy saving. Nowadays, most buildings adopt the Building Automation System (BAS) based on wired networks to save energy by controlling the HVAC system and lighting system. However, because of the mutually independence between the sub-systems of the BAS, it's hard to detect, analyze and control the overall building consumption. The wired network has the disadvantages of complex cabling, high-cost, small capacity, high-power, and so on. So it's difficult to realize the large-scale deployment and the complete coverage, and it's hard to comprehensively detect the energy requirement and the energy efficiency state and to achieve the practical and effective energy-saving goal. Especially for the energy-saving reconstruction of the existing buildings, it's especially unreasonable to adopt the wired networks. According to these problems, a building Energy Management System based on Wireless Sensor Networks is researched and developed in the thesis.In the thesis, the architecture of the Intelligent Building Energy Management System is researched, the disadvantages of the IB-EMS based on wired networks is analyzed, the channel fading and the penetration capability of the electromagnetic wave are tested, and the nodes deployment and network planning are reasonably researched. Then an Intelligent Building Energy Management System based on Wireless Sensor Networks is designed, includes the hardware and software design of the Detecting and Controlling Nodes and the Sink Nodes and the design of the network communication protocols. In the thesis a kind of "Two Parallel-lines Zonal Mesh Network" is designed, a strategy of combining moving statistical average PLR with RSSI to find its neighbors dynamically is put forward, a new routing algorithm called "Generation Routing" is put forward, which is based on local information, and a new data structure called "Cross Binary-tree" for the algorithm is defined. At last, the experiments and tests on the node power consumption in different work state, the establishment time of the Wireless Sensor Network, the network delay, the routing searching complexity, and the energy-saving effect of the system are carried out.The results show that, firstly, the designed network has low power consumption compared with the traditional RS-485 networks. The average power consumption of the network with 20 nodes is less than 1 watt. Secondly, the network performance meets the application demands. For a network with 10 generations at 2.4kbps transmission rate, the self-organization time is less than 11 seconds, the network delay is less than 1.7 seconds, and the routing searching complexity is O(N). Thirdly, the building energy-saving effect is obvious by using the system, which can save 30% energy on average. Fourthly, the system is low-cost, free of cabling, easy to extend, and convenient to maintain. To sum up, the designed system is suitable for the building energy-saving, especially for the energy-saving reconstruction of the existing buildings.