Input Estimation Method in the Use of Electronic Device Temperature Prediction and Heat Flux Inverse Estimation

Abstract

The input estimation method (IEM) is proposed in this research to estimate the real-time time-variant core heat flux of electronic components. The IEM is an on-line real-time estimation method, which combines the Kalman filter mechanism and recursive least-square estimation (RLSE). On the premise that there exists both system processing error and measurement error, using temperature measurements measured from the exterior of the device package, the core time-variant heat flux of the device can be precisely estimated inversely.