To enable comprehensive prediction of typical fuze cook-off processes and address the challenge of quantifying output pressure， an advanced strain-gage pressure bar sensor was utilized for dynamic pressure acquisition during experimental investigations. A comprehensive coupled numerical framework was developed， integrating heat transfer models， Arrhenius reaction kinetics， and ignition response mechanisms， to systematically analyze the cook-off behavior and generate detailed pressure profiles of booster explosives. The kinetic parameters， such as activation energy and pre-exponential factors， were inversely determined through the application of a Back Propagation （BP） neural network. Meanwhile， the state parameters that govern the ignition reaction equation were optimized using a multi-island genetic algorithm. Coupled simulations utilizing ANSYS Fluent and LS-DYNA within the Workbench platform were performed to numerically investigate the cook-off response under different heating rates. This approach enables comprehensive full-process characterization from thermal reaction to ignition. The results indicate that slower heating rates shift the ignition zone toward the central region of the charge， thereby intensifying the severity of the reaction.