Real-time warning method and application of microseismic multi-feature parameters rockburst based on PSO-GRNN model

Rockburst disasters in the construction of deep-buried tunnels seriously threaten the safety of underground projects. Traditional monitoring methods have limitations in the analysis of nonlinear and small-sample microseismic data. To effectively reduce the risk of rockburst disasters, a microseismic multi-parameter monitoring method based on the PSO-GRNN model is proposed. Meanwhile, a field sound and light alarm system is independently developed to provide real-time feedback on the prediction results. This method collects the characteristic parameters of microseismic signals in real time and combines the dynamic comprehensive hazard index W _Z ( t ) of the grey correlation degree method to construct a multi-parameter early warning criterion standard, effectively solving the scenarios of nonlinearity and small samples of microseismic data in deep-buried tunnels. The developed method and systems are applied on-site in the DJ Tunnel in western China, with good results. The early warning accuracy rate is 92.8%. A complete closed loop of data collection-intelligent analysis-multi-level early warning- emergency response is constructed, providing valuable references for on-site rockburst early warning.