Research On Data Fusion Technology In Wireless Sensor Network For Forest Fire Prevention

Forest fires pose a serious threat to the safety of people’s lives and property.Forest fire monitoring is an important means to prevent fire.However,traditional forest fire monitoring technology has the disadvantages of long monitoring period,low sensitivity,and difficulty in implementation,which can not meet people’s requirements for forest fire monitoring technology.Wireless sensor network technology has the advantages of low cost,good flexibility and high degree of automation.It has Good application prospect in forest fire prevention.Data fusion technology has obvious advantages in dealing with WSN data intelligent processing.Therefore,this paper will focus on the application of WSN in forest fire monitoring.On the basis of the analysis of the characteristics of the fire development process,the paper selects temperature,smoke concentration and CO concentration as the reference index of the forest fire monitoring,and proposes a two level data fusion method based on WSN to improve the timeliness and accuracy of forest fire monitoring.The main work of this paper is as follows:1.The main purpose of the first level of data fusion is to complete the fire prediction and data deduplication.Based on the analysis of fire process characteristics,this paper proposes the concept of fire support degree,and on this basis,a data fusion algorithm based on fire support degree weighting is proposed.he traditional data redundancy algorithm is carried out on the same kind of sensors.It can not make full use of the characteristics of the same trend of monitoring indicators when the forest fire occurs and is insensitive to the identification of initial fires.The fire support weighted fusion algorithm proposed in this paper is based on heterogeneous sensor data,and takes account of the change characteristics of monitoring indicators.The algorithm can effectively improve the sensitivity of initial fire detection and improve the timeliness of fire identification.2.In the second level data fusion,this paper analyzes the deficiency of single D-S evidence theory or rough set theory applied to forest fire monitoring,and proposes a data fusion algorithm that combines D-S evidence theory and rough set theory.This paper proves the feasibility of the algorithm from a theoretical point of view,and it also provides a solution to the problem of evidence conflict in the D-S evidence theory andthe data discretization problem in the rough set theory.The new algorithm can make up for the vulnerability of subjectiveness of initial function valuation of D-S evidence theory and the flaw of rough set theory to solve uncertainties,and get more objective and convincing decision results.3.Through the simulation of Chinese standard test fire data,it is proved that the proposed data fusion algorithm has certain improvement in the timeliness and accuracy of forest fire monitoring.