| In the market there are many wired circuit connection mode of fire alarm devices which is difficult wiring, circuit ageing, easily influenced by the environment in a actual and complex environment. And its cost is higher. And the alarm devices is composed of a sensor selection for a product after the fire alarm equipment composition. These signal sensor has a high false alarm rate. Therefore, this paper uses multiple sensor circuit for data acquisition and uses wireless network to transmit the data and uses algorithm to process it for classification which can effectively reduce the false alarm problem.In the first chapter, it introduces the origin of this research, the development trend of purpose and significance research ideas and methods at home and abroad as well as the fire alarm. Then in the second chapter, it introduces the ZigBee protocol stack and the concept of the fuzzy algorithm. It provides the later system design with the theory supporting. In the third chapter, it mainly produces the details of the system design. Firstly, the system is built in the base of CC2430. For the detection node, it contains data acquisition module design, power module design, sound and light module design. The gateway node design contains LCD design, key module design, power module design. Secondly, the software design is based on ZigBee. This section begins with the brief introduction of the nodes in the Z-STACK and then it introduces the basic initialization protocol stack and system initialization. It is the first that it gives the processing user which is defined at the system initialization. After the whole initialization, it starts the system.Then the paper mainly introduce the flow of which setting the network or joining the network and some basic configurations of different nodes.Then it design the communication format according to the requirements of the system. The communicatin format size must meet to accommodate the whole bytes which contains command type,upload/send the data command, change upload cycle aommand,change the sensor four threshold command,open the electromagnetic valve command. Then it complete the flow of node according to the need of gateway node and detection node and complete the specific soft design of the node. The detection node mainly complete the data acquisition, alarm, wireless transceiver design. The gateway node must complete design of serial communication, keyboard, LCD display, wireless transceiver and grade evaluation aspects. The fourth chapter mainly do the research about fuzzy algorithm of early warning mechanism. This chapter introduces the fuzzy C means algorithm and use the algorithm for simulation about classification for the fire warning mechanism. Then on the basis, it gives the trapezoidal algorithm and uses this algorithm to do the research about fire early warning mechanism. In the fifth chapter, the system start the test after completing the whole design. We simulate the fire scene for the system test to complete some functions which contains communication, topology, alarm, command operation test. Finally, I use C#to complete software for whole test.Experimental tests are done in the laboratory. The nodes distribution must convert the entire laboratory space. Through the test, the fire alarm system basically meet design requirements. If it has a further improvement, it should have a broad application prospects. |