Emergencies Forecasting Based On Wave-let Method

This paper is based on the project Emergency Response Command InformationSystem (ERCIS) of the Tianjin Science and Technology Committee. ERCIS is aninformation system which integrates information collection, transfers, district securitycontrol, emergencies forecast, Command and disposition into one system. Optimizingresources deployment and improving the disposition efficiency are the targets ofERCIS. One precondition of implementing the targets is estimating the intendingdistributing state of emergencies by the rule and line. In the prophase work, this paperanalysed the emergency time-series by using the traditional method, but theacceptable result was not reached because there was too much error. On the base ofprophase work, this paper hang the wave-let analysis and high-level Auto-Regressivemethod together and puts forward Wavelet-AR model which has high forecastprecision.This paper designs an emulational experiment with the data of Tianjin 110 Center.This paper analyses , modeling and simulatively forecasts the emergencies time-series.The results show us the Wavelet-AR model can improve the precision obviouslyhich can satisfy the request of actual combat.At last this paper applies the Wavelet-AR model into ERCIS and try to constructa Emergency Stream Forecast System(ESFS) which has a open construction. Thesystem is a subsystem of ERCIS and provides the Assistant Decision service toERCIS.