LaplaceDiag Uses Actuator Effect Changes for Fault Diagnosis

LaplaceDiag identifies faulty components in complex plants by decoupling nonlinear history prediction from actuator-to-sensor effect estimation in Laplace space. By keeping the LaplaceNet predictor fixed, the system uses linear regression to isolate changes in actuator influence, enabling root-cause analysis without labeled fault data.