Localization Of Toxic Gas Boundary Area Based On Wireless Sensor Networks In Large-scale Petrochemical Plant

Wireless sensor network is consist of a large number of cheap micro sensor nodes,which are deployed in the monitoring region. By using wireless communication, the sensor nodes forms a multi-hop Ad Hoc network system. Sense, collect and process the information of object collaboratively in the network coverage region. And then they send the information to the observer.In the large-scale petrochemical plant, it can realize intelligent factory by using wireless sensor network. And it is very helpful to monitor temperature, monitor hydraulic pipeline, diagnose equipment fault, monitor pollution emissions and monitor toxic gas leakage. In industry,production equipment may be corroded with long-term expose of high temperature and strong acid chemicals, which can result in equipment fault, toxic gas leakage and explosion. When the explosion occurs and the toxic gas leaks, it will pose a great threat to workers’ lives and cause significant economic losses to the plant if we don’t take measure in a timely manner.Toxic gas, as a continuous object, is different from the common continuous objects, such as liquid water, oil, etc., because it is colorless, tasteless and it has no clear boundary. And it is hard to be found by naked eyes. In the plant, people care about the internal boundary and internal region of the liquid, such as oil, because people can see the situation of oil leakage by naked eyes. So they just want sensor nodes to collect and report the information of internal boundary. However, it is different to the toxic gas. Gas will diffuse and the naked eye cannot see it. So it is not only to report the internal region and internal boundary of the toxic gas, but also to explore the scope of the spread of toxic gases in order to evacuate the first-line workers in the factory. So in this paper, we try to explore the scope of the rough boundary of the toxic gas instead of the internal boundary by using wireless sensor networks.In order to locate the boundary region of the toxic gas better, we import the planarization algorithm. The planarization algorithm can effectively cut a complex network graph into a planar graph composed of each clear Face, which is easy to calculate the boundary area of toxic gas. To explore the influence of planarization algorithms about the area of toxic gas boundary, we compare the planarization algorithm of GG with the RNG. Meanwhile, we explore the influence of different node densities and different network topologies, such as random topology, grid topology, etc. And we can know from the simulation results that different planarization algorithms form different boundary area of toxic gas and the GG planarization algorithm is more accurate than the RNG. And with the increase of the node density, the area of toxic gas boundary is more accurate. But when the density reach a certain value, the effect of node density is very small. And different network topologies form toxic gas boundary region. Experimental results show that the grid topology is the ideal, but it is hard to be deployed in the plant.