机构地区: 中国冶金科工集团有限公司中冶建筑研究总院有限公司
出 处: 《工程抗震与加固改造》 2007年第3期109-113,85,共6页
摘 要: 本文基于可辨识矩阵和信息熵理论,提出了震害预测影响因子的启发式属性约简算法,并通过实例验证了方法的有效性和可行性。算例表明,该法可在不影响震害预测结果的前提下,使得由建筑物及其属性构成的二维决策系统的信息量大为简化,从而大大降低了震害预测前期数据采集和后期数据处理的工作量。本文所做的工作,对进行抗震防灾并制定相应的规划具有较好的参考和借鉴价值。同时,在理论上为从事震害预测方面的研究提供了一条新的思路。 Based on cognizable matrix and information entropy theory, the influenced factors reduction algorithm in seismic prediction are put forwand, and its feasibility is proved by an example. The illustration indicates the method can predigest information of the planar decision- making that buildings and their properties compose, assuring result of the seismic prediction invariability. So the workload of prophase data collection and anaphase data processing in seismic prediction are reduced. The work of the paper has good referenced value for anti-seismic, as well as provides a new insight for developing seismic prediction.