An efficient on-line thermal input estimation method using Kalman filter and recursive least square algorithm

Conference · Tue Jul 01 04:00:00 EDT 1997

OSTI ID:20014414

Tuan, P C; Lee, S C; Hou, W T

The efficient on-line thermal unknowns estimation using the Kalman filter and recursive least square with forgetting weighting algorithm is presented. The efficiency is dominated by the best choice of the forgetting factor under different scales of covariance of process and measurement noise. In this paper the roots mean square error is mainly used as the performance index to discuss the role and effect of forgetting factor. Simultaneously, the performances of the proposed algorithm in time domain and in frequency domain of the estimation are also discussed. In summary, an rigorously efficient robust forgetting factor zone, which provides a excellent tracking time-lag and noise filtered estimation result, is introduced. This zone is applicable to any type of time varied thermal unknown functions in the Inverse Heat Conduction Problem (IHCP) and suitable for the hardware loop realization under the global uncertainties. In addition, the thermal diffusion lag, is also discussed and compensated in this paper. The superior results are verified through one dimensional IHCP simulation.

• Research Organization: Chung Cheng Inst. of Tech., Tahsi, Taoyuan (TW)
• Sponsoring Organization: National Science Council of Republic of China
• OSTI ID: 20014414
• Country of Publication: United States
• Language: English

Related Subjects

42 ENGINEERING
99 GENERAL AND MISCELLANEOUS
ALGORITHMS
DIGITAL FILTERS
HEAT FLUX
LEAST SQUARE FIT
PROBABILISTIC ESTIMATION
THEORETICAL DATA
THERMAL CONDUCTION
THERMAL DIFFUSION
WEIGHTING FUNCTIONS