Relevancy between ground-motion signals caused by underground chemical explosion

Equations of motion were analyzed for governing underground explosion-induced stress waves spreading in an elastic domain to explain that this stress wave transmission processes could be described by a time-independent linear system model. According to the impulse response methods in the system identification field, the relationships were established to link the impulse series and source functions in the measured points with the response series. Based on the established relationships, a response model was developed to depict the relevancy between the stress series in the measured points. And a method was introduced to calculate the relevant parameters by adopting the adaptive optimization method of neural network and some important parameters were listed. The relevant response model was validated by the experimental data for a certain chemical explosion in soil. The predictions by the relevant response model are in agreement with the measured data and the relevancy between different measured points is determined only by the distributions of the explosion source, measured points and geological parameters.