Simulating CERN's Quark-Gluon Plasma Depletion Wakes: A Unified AMDCF-CSD Predictive Control Kernel

Recent direct observations by the CMS Collaboration at CERN have revealed hydrodynamic "depletion wakes" left by partons traversing the Quark-Gluon Plasma (QGP). Simulating this supersonic particle-fluid interaction is numerically challenging due to the singularity at the parton's core, which often causes computational blow-up. In this study, we present a unified AMDCF-CSD (Adaptive Memory-Driven Coupling Framework with Critical Slowing Down) kernel that combines wave propagation, artificial viscosity, a non-Markovian memory field, and a Predictive Safety Valve. The CSD algorithm forecasts impending instabilities by analyzing the variance and autocorrelation of local gradients. When the parton approaches the core, the CSD risk spikes, dynamically lowering the safety valve threshold. This proactive control bounds the core singularity, reproducing the theoretical V-shaped Mach cone without numerical blow-up. The simulation successfully transitions from reactive to proactive control, demonstrating the framework's capability in computational high-energy physics. Keywords: Quark-Gluon Plasma, Mach Cone, CERN CMS, AMDCF, CSD, Predictive Control, Numerical Stability, Singularity Control.