Research on grey modeling means and its application

Grey theory was applied in constructing the model of the estimating nuclear explosion yield, to compensate the practical value of the existing nuclear explosion yield detecting methods and improve the accuracy of the yield estimate value. The methods of state estimation, bad data detection and identification were applied in grey system modeling to get some more reasonable results compared with the conventional estimation methods of the nuclear explosion yield. The proposed methods can be used as data preparation techniques to detect bad data. The modeling accuracy of grey systems is greatly improved by such methods. Based on the grey theory and the processing of plenty nuclear explosion test data available from home and abroad, the result indicated that the new optimal estimation method using the grey theory and bad data detection and identification to estimate the nuclear explosion yield is feasible. Compared with the existing estimation model, the new estimation method can reduce error range, advance precision, offset the deficiencies of the existing means detecting the nuclear explosion yield, and enlarge the practical application value.