Research On No-stable Gas Source Location Using Wireless Sensor Network

Wireless sensor network (WSN) are widely used in military, environment, industry and other fields for its low cost, low power, multi-function, self-organization and other characteristics. Location is one of the key technologies of WSN. It can be used to estimate the gas leak position through the gas concentration measured by the monitoring node and gas diffusion model. Wireless sensor network is used for the gas source localization, timely access to gas leakage location and source information, contributing to the early detection of a fire and hazardous gas leak alterm.This paper introduces the current research status of gas source localization and the characteristics of wireless sensor networks. The localization technology in wireless sensor networks is studied further. The existing gas source localization algorithms only consider steady state of gas diffusion, but this paper focuses on the research of non-steady gas diffusion problem. The main work of this paper is as follows:The gas diffusion theory is studied further and the formula of unsteady gas diffusion model is derived. On this basis, the gas concentration measurement sensing model is established. Combined with the characteristics of wireless sensor network, the centralized localization algorithm of gas source location is proposed. With the measurements of sensor nodes, the problem is transformed into a global optimization problem. The proposed algorithm employs multi-populations with different evolutionary mechanism. It has advantages of strong global searching ability and fast convergence speed and it can effectively calculate parameters. As the selection operation destroys population diversity of the algorithm, the cyclic shift crossover is introduced. It increases the diversity of population and improves the effectiveness of the algorithm. Simulation results show that that proposed algorithm is feasible for two-dimensional non-stationary gas source localization. It also shows that the proposed algorithm has higher accuracy and grater robustness with respect to the centroid localization algorithm.Most of the current gas source localization algorithms in wireless sensor networks consider the scene of two-dimension, but the actual applications are all three-dimensional cases. Therefore, the two-dimensional gas source localization algorithm is extended to three-dimensional cases in this paper. The diffusion law is used to derive the gas concentration measurement model in three-dimensional case. The optimization objective function is constructed and the parameters of the gas source are solved via multi-population genetic algorithm. Simulation results show that the proposed algorithm has excellent location performance in three-dimensional space.