【摘要】 地质灾害危险性预测评价的准确性,主要取决于基础资料的可靠性和数学模型的合理性。论文结合工程实例,尝试用人工神经网络方法(改进的神经网络BP模型)对区域地质灾害危险性预测进行评价研究。然后与目前常用的方法(如层次分析法、信息量法和模糊综合评判法等)所得出的结果相比较。结果表明,运用人工神经网络方法对区域地质灾害危险性预测评价相对常用方法更准确、可靠,具有一定的实用意义及推广价值。更多还原

【Abstract】 The veracity of risk and assessment of risk in geologic hazards lies on the reliability of basic datum and the rationality of math ematical model.Combining with the examples of project,the paper tries to use the method of artificial NN(improved BP NN model) to forcast and estimate regional geologic hazards.Then,compared with the result of routine procedure at present(such as AHP,amount of information method and fuzzy synthesis judge method),the results of the method of artificial NN are more accurate and more reliable than the results of routine procedure in regional geologic hazards’ forecast and assessment of risk,and have some practical significance and popularization value.更多还原