Optimal design of dissipation for the array power LED by the genetic algorithm

This study applies the ANSYS finite element analysis software to analyze the thermal behaviors of the array LED. The array LED is modeled by 4 same-size LEDs with MCPCB (Metal Core Print Circuit Board) on a heat-sink. Prior to the process of the optimal design on the chip junction temperature, the most significant parameters are chosen by the fractional factorial design method. The regressive models are set up by the double response surface method and the mixed response surface method respectively. Furthermore the genetic algorithm combined with the response surface is applied to obtain the optimal design parameters, and then the results obtained by both two methods are discussed. Afterwards, a mixed response surface method is applied to analyze the interactive effects among various parameters on the chip junction temperature. Finally it is found that the conductivity of MCPCB and the heat-sink height are the most significant factors. Besides, the interactive effects between the size of chip and the thermal conductivity of the chip adhesion layer are the most obvious effects on the chip junction temperature.