NEURAL-NETWORK METHODS IN HYDRODYNAMIC YIELD ESTIMATION

Hydrodynamic theory allows us to use the speed of a shock wave front t o determine the yield of an explosion. On the basis of this theory we developed a neural network to estimate a yield of underground explosio ns from the shock wave radius versus time (RVT) data, as measured by c ontinuous reflectometry for radius versus time experiments (CORRTEX). The proposed method not only replaces the subjective elements of conve ntional algorithms, but produces significantly improved yield estimate s. The network was trained with rear hydrodynamic data and its perform ance on unseen test events was studied. A backpropagation network was employed; the architecture consisted of ten input units, a hidden laye r with eleven hidden units, and one output unit. The network was train ed with thousands of input-output measurement vectors, the feasible in put set, derived from the RVT data from only four other training or st andard events (selected on the basis of the given RVT data from the un known event). The feasible input vectors were propagated through the t rained network and the network output was used to select the optimum y ield estimate. Elements of the input vector were: center of energy (CO E) offsets, shock front radii, and time onset and interval of analysis for both the standard and unknown events. We studied the performance of the proposed system using 24 Nevada Test Site (NTS) events that. we re located in the geologic medium tuff. Sensitivity analysis of the pr oposed method to the assumed nominal COE offset is discussed. Variatio ns of the proposed system that might lead to further improvements in p erformance are suggested.