D-S理论在多传感器短临震预报中的应用

Application of D-S theory in short-term and impending earthquake prediction of multi-sensor

在D-S理论的基础上,结合模糊数学理论,依据地震前兆,给出了短临震预报的信息融合算法。首先,对在现场工作的地电、地应力、地下水位、短水准、水氡等传感器的数值进行特征提取,把模糊数学中隶属度函数的概念运用到D-S理论中,得出每种前兆出现时地震发生的信度函数分配,然后,采用多传感器分布式融合算法进行信息融合,利用判决规则来预报地震。通过对1976年唐山7.8级地震预报的仿真,结果表明:D-S理论在地震预报中具有一定的有效性和优越性。

An information fusion algorithm for short-term and impending earthquake prediction by seismic precursor is presented, The algorithm is based on D-S theory and fuzzy mathematics. Firstly, the characteristic values of Geoelectricity, ground stress, groundwater level, short levelling and groundwater radon sensors in the locale are extracted. Then, the concept of membership function in fuzzy mathematics is generalized to D-S theory to produce basic probability assignment（BPA） of earthquake when each precursor occurs. The BPAs fuse according as D-S rule, and through the judge rule of target mode, the situation of earthquake is predicted at last. Through the prediction simulation to 7.8 earthquake of Tangshan in 1976, it is shown that the D-S theory is more efficient and superiority in earthquake prediction.