AnadaptedSLABmodelusingsensordatafortheprediction.docVIP

An adapted SLAB model using sensor data for the prediction on the dispersion of hazardous gas releases Won Soa, Dongil Shinb, En Sup Yoona aSeoul National University, Seoul 151-742, Korea bMyongji University, Yongin, Kyunggido 449-728, Korea Abstract In this research, we propose an emergency response system against hazardous gas releases. We use a sensor network already set up and operational in an area where hazardous gas releases can happen. From the real-time sensor data, we detect and analyze harmful materials releases and their concentrations. Based on that, the release rate of hazardous materials is estimated using the neural network. The proposed dispersion model, a modified SLAB model, performs the simulation of the expected dispersion consequence by combining the sensor data, GIS data and the diagnostic result of the source term. The result of this study will improve the confidence of residents living next to storage facilities of hazardous materials, by supplying the improved emergency response plan and monitoring system for toxic gas releases. Keywords: hazardous gas, dispersion model, neural network, sensor data Introduction Economic development and ever-increasing adoption of technology for everyday life have unavoidably introduced many dangerous facilities inside and nearby city. Accidental discharges of dangerous gases, either flammable or toxic, more likely to occur during manufacture, storage or transport could be considered as one of those cases. In this research, we propose an emergency response system against toxic gas releases [1, 2, 9], including chlorine releases. We set up a ubiquitous sensor network in a hazardous area where chlorine gas releases can happen. Fixed or mobile sensors detect and analyze harmful materials and their release concentrations. Based on this pseudo real-time information, the amount of release and its position on the GIS (Geographic Information System) map are estimated and warnings are generated if necessary. This