Residual-Based Inverse Modeling for Fragment Mass Estimation

This project presents a novel residual-based multidimensional inverse modeling system designed to estimate fragment masses and detect structural interference following an explosion event. By reformulating the momentum conservation law as an overdetermined linear system, the model treats residual vectors not as error terms, but as informative indicators of system imbalance and external forces. The proposed framework allows: Mass estimation even under partial information conditions, Directional inference of external perturbations, Detection of sensor inconsistency or spoofing, Scalable modeling in high-dimensional multi-body systems. A full LaTeX source file, compiled PDF, and supporting datasets (if any) are provided here for public access. The project is intended as a bridge between physics-based modeling and intelligent inference systems, and is applicable to fields such as aerospace dynamics, sensor fusion, AI-based fault detection, and quantitative finance.