基于GIS的BP神经网络洪涝灾害评估模型研究.pdfVIP

31 4 V o.l 31, No. 4 2009 8 A cta A gr icu lturae Un iversitatis J iangx iensis A ug. , 2009 : 1000 - 2286 2009) 04 - 0777 - 04 GI S BP 1 2 1 3 单九生 , 徐星生 , 樊建勇 , 管 珉 1. , 33004 6; 2. , 330046; 3. , 2 10044) : , G IS, G IS , , BP , m atlab , G IS BP , , , 12% : G IS; m atlab; ; BP : S422 : A A Study on the N euralN etworks A ssessm entM odel of F lood andW aterlog D isaster Based on GI S 1 2 1 3 SHAN Jiu - sheng , XU X ing- sheng , FAN J ian- yong , GUAN M in 1. M eteoro log ica l Sc ience Inst itute of J iangx i Prov ince, N anchang 330046, China; 2. M eteo ro log ical Ob servato ry of J iangx i P rov ince, N anchang 33004 6, China; 3. N an jing Un iversity o f Info rm ation Science and T ech nology, N anjing 210044, Ch ina) Ab stract F loods across Ch ina have very d ifferent characteristics. T here are so m any types o f flood, depen ding on geog raphy, chmi ate /w eather characterist ics, and hum an populat ion, soc ia l and econom ic situat ions, etc. Thu s, assessm ent of flood dam age is a comp lex task. Base on the G IS techno logy and its database, the paper m anages to smi u late the flood depth and the d istribut ion o f econom y on each G IS gr id. A nd to so lve the uncer tainty o f the inf luences o f various com plex factors on the econom ic loss assessm ent precision, B ack P ropagation N eural N etworks are applied by the m at lab neura l ne wt ork too l- boxes. T hu s, a m ode l o f flood dam age assess m ent is constructed. In this w ay, w