Occurrence assessment of earth fissure based on genetic algorithms and artificial neural networks in Su-Xi-Chang land subsidence area, China

摘要: Earth fissures in Su-Xi-Chang land subsidence area have induced massive damages to the area. The non-linear characteristic associated with process of earth fissure formation requires method for evaluating occurrence hazard. Based on quantification influence factors breeding hazard, GA-ANN method, which integrates artificial neural networks (ANN) genetic algorithms (GA), is developed Six indicators, that include depth bedrock burial (DBB), degree relief (DBR), water level (WL) (the II confined aquifer), gradient (GLS), transmissivity (T) aquifer) and thickness clay soil (TCS), are selected as input patterns integrated approach, danger index (DI) output pattern. A multilayer back-propagation network trained 30 sets data samples including 15 safety defining architecture ANN. Subsequently, GA employed by optimizing initial weights ANN minimizing deviation output. efficacy approach demonstrated comparing from 5 testing result shows more accurate than identifying fissure. applied assessment hazard typical regions According classification DI, divided into four zones — zone, sub-danger sub-safe zone safe zone.