基于无线传感器网络森林火灾模型参数估计.pdfVIP

第30 卷第1期 东 北 大 学 学 报 ( 自 然 科 学 版 ) Vol30, No. 1 2 00 9 年 1 月 Journal of Northeastern U ni ersity( Natural Science) Jan. 2 0 0 9 吴成东, 孟 伟, 纪 鹏, 陈 飞 ( , 110004) : , , , , ( EM) ( MR) , EMMR , , EMMR , : ; ; ; : TP 242 : A : 2009) 01002 105 Parameter Estimation Based on Wireless Sensor Network for Forest Fire Model W U Chengdong, MEN G Wei, JI Peng , CH EN Fei ( School of Information Science Engineering , Northeastern Uni ersity, Shenyang 110004, China. Correspondent: MENG Wei, Email: mengwei2499 @ yahoo. com. cn) Abstract : Since forest fire is too complicated to search, extinguish and rescue, the w ireless sensor network ( WSN) is applied to forest fire detection and an exact model is de eloped to estimate the location and intensity of fire. The approach to adapti e sampling by use of sensor nodes is discussed w ith a multipeak Gaussian model proposed. Combining the expectation max imum ( EM) w ith multiple resolution ( MR) algorithm, an ad anced EMMR algorithm is proposed to estimate the parameters in the forest fire model. Based on the results of comparati e simulation, this model can depict the relati e real site of forest fire and pro ide the satisfactory estimation accuracy w ith computation complex ity significantly reduced. Key words: wireless sensor network; forest fire; Gaussian model; parameter estimation , , , ,