Font Size: a A A

Study On Simulation And Economic Impact Estimation Of Land Subsidence

Posted on:2007-10-01Degree:MasterType:Thesis
Country:ChinaCandidate:J J ZhangFull Text:PDF
GTID:2120360212471113Subject:Environmental Engineering
Abstract/Summary:PDF Full Text Request
As a complicated environmental geological problem, the land subsidence which is the main geological disaster of Tianjin City has seriously impacted on the construction of city's basic establishment and restricted Tianjin's global economical development. At present, it has improved that the overexploitation of groundwater which unbalanced the soil dynamics is the uppermost reason of the land subsidence. So, it has positive and profound meaning to develop the suitable predict method for the specific area of Tianjin City ,on the reference of domestic and abroad newest prediction model, which can provide scientific basis for the effective control of land subsidence.The Tanggu district which has more complete historical monitoring datum was chosen as the demonstration zone. On the analysis of the datum's tendency, a BP neural network prediction model is established. Annual water table of different confined aquifer is introduced as model input, while land subsidence rate of optimized monitoring points is defined as model output. After compared and filtered, a network structure with double concealed layers, each layer has 8 nodes, has been affirmed finally.According to model estimation, the variation of land subsidence rate in Tanggu district has the same response to the water table. Lifting with water table of earlier stage, the alleviation of land subsidence is obvious, but function will be weakened gradually on later stage. The intensity of influence of different aquifer to land subsidence descends in order ofâ…£,â…¢,â…¤,â…¡confined aquifer.In order to overcome the shortcoming of BP neural network model such as being apt to fall into the local best and being slow to convergence, a combined model using genetic algorithm's strong overall search ability to optimize the primary connection weights of BP neural network has been defined. The simulation process of the coupling model is more stable, and the operation results of several times are unanimous basically which has testified the coupling model's validation.At last, on the basic of the prediction model, the evaluation index system of land subsidence economic loss is established, and the exterior economic theories have been introduced to analyze economic impact of land subsidence. It is proved that the government control and inspire system has great contribution to the alleviation and...
Keywords/Search Tags:Water Table, Land Subsidence, BP Neural Network Model, Genetic Algorithm, Economic Impact, Tanggu District
PDF Full Text Request
Related items