Nonlinear optimal on-line heat-dissipation control methodology in electronic devices

The essential element in obtaining semiconductor electronic device enhanced reliability involves solving the heat-dissipating issue. Certain electronic components possess varying thermal properties and the strength of the heat generated by the semiconductor chip is unknown. Therefore, the heat dissipation control problem is both complicated and perplexing. This paper proposes a methodology, the LQG/IE tracking algorithm, to solve the nonlinear heat dissipation control problem. The IE method, which is a combination of an Extended Kalman filter and Recursive Least Squares Estimator, estimates (in real time) the unknown time-varying heat source generated by the semiconductor chip using temperature measurements of the package seal surface. The LQG tracking algorithm is adopted to analyze the feedback gain to control the heat dissipation. The simulation results reveal that an effective and optimal heat dissipation controller can be implemented for the cooling system using the LQG/IE tracking algorithm.