Key Technology Research Based On Internet Of Things Of Forest Fire Warning

At present, because of the characteristics of wide spread, high frequency and risk and the like, forest fires, once occurring, will not only bring a great loss of money, but will also destroy forest resources, disturb the normal order and cause serious pollution to our environment. Therefore, governments around the world pay more attention to the forest fires. As to the researchers as well as the scholars home and abroad nowadays, how to use Internet of Things, remote sensing and other high-tech means to warn and monitor forest fires has become a hot research topic. On the basis of existing intelligent security monitoring system of forest fires and under the premise of detailed analysis of the system, this paper has made a research into the realization of the key technology of forest fire early warning system, by the way of some modern high and new technologies such as Internet of Things and computer technology.With the application of video, graphics and image, GIS and some other technologies, forest fire intelligent security monitoring system monitors fires and raises the alarm in time. The system focuses on the alarm and decision support when the fire occurs, while forest fire early warning system puts emphasis on the prevention of fires, which is realized by real-monitoring the data in the forest environment (fire fuel state index, weather index, forest fire danger division level index and fuel type index), which, mainly realize early warning before the occurrence of fires and auxiliary realize the prediction of fire spread after the occurrence of fires. Therefore, how to collect the data information, how to guarantee the accuracy of the data information, how to realize the reliability and real-time of data transmission and how to use information to predict risk level precisely are technical difficulties to be solved. As an important part of Internet of Things, wireless sensor networks are comprehensive data information processing platform which set computer technology, network communication technology, sensor technology as one. This paper uses the technology of wireless sensor network to collect data information timely and puts particular emphasis on researching on the shortcomings like low accuracy of node location, high energy-consuming of node and short network lifetime. This paper explores location technology and communication network technology of wireless sensor networks in depth, dedicates to overcome the location error of non-line-of-sight environment and its cumulative error in the course of wireless sensor networks localization and provides routing protocol algorithm to realize the longest life cycle of sensor node according to specific environment, which can predict forest fires timely and reliably. Main work summarized below:I. According to the reason for low location accuracy of traditional DV-HOP algorithm distance value between hop distance Hospi and actual per-hop distance between nodes. This paper proposes improved DV-HOP algorithm which mainly improves the way to get average single hop distance in the algorithm. First, adopting weighted mean:using corresponding hop hi as weight and formula to find mean; second, using strategy to narrow the estimation region of unknown nodes and basing the principle of centroid algorithm, the coordinates of unknown nodes is the centroid of narrowed estimation region.II. Improved centroid algorithm divides the whole network according to the number of beacon nodes covered by the area and forms closed region. The algorithm unites the RSSI value that unknown nodes collect and further minimize the estimation region of unknown nodes in the closed region.The centroid of estimation region is estimation coordinates of unknown nodes, experiment shows that this improved centroid algorithm has higher location accuracy than traditional centroid algorithm and higher accuracy than common improved centroid algorithm.III. On research of the data processing problem in sensor network, the new algorithm proposed consists of four steps:formation of clusters, choice of cluster heads, processing data in clusters, data-gathering route of mobile agent. The algorithm uses fuzzy C-means clustering algorithm to generate the number of clusters; in the central part of clusters, choses multiple nodes round-robin to be cluster heads to achieve the propose of balancing energy dissipation of cluster heads; the communication between cluster heads and data point using the transmission method between cluster heads and cluster heads or applying mobile agent acquisition mode to transfer data, the two strategies both aim at avoiding long distance transmission between cluster heads and data point which will lead energies of nodes to deplete too fast.IV. Constructing experiment system:building simulation platform according to experiment target, establishing simulation platform which conforms to actual application situation and obtains experimental results of related technology under the software circumstance.