Life prediction of LED lights by ADTs combined with luminance degradation and probability statistics

Abstract

For the purpose of improving the life accuracy calculated based on accelerated degradation data of light-emitting diodes (LED) lights, three-parameter Weibull right approximation method (TPWRAM) was applied to fit four groups of degradation data from accelerated degradation tests (ADTs) to obtain luminance degradation curves, and the pseudo-failure time of each sample was obtained combined with failure criteria. Then the life distribution under each stress was described by lognormal distribution, the maximum likelihood estimation (MLE) was employed to calculate the distribution parameters, and the accelerated average life was acquired. Finally, the LED light life under conventional stress was predicted by the inverse power law. The proposed life prediction method was also applied to predict the life of vacuum fluorescent display to verify the applicability of the proposed method to other data. The results show that the data obtained via the four groups of ADTs are highly reliable and the life distribution of LED is completely in line with the lognormal distribution. Besides, the determination coefficient R² of the LED life characteristic curve is 0.9841, which is close to 1, meaning that the conventional life prediction of LED is accurate. Then the result of applying the same method to the life prediction of vacuum fluorescent display (VFD) shows that the relative error between the predicted value and the actual conventional life of VFD is only 1.44 %, which proves that the proposed life prediction method also has good applicability to other photoelectric device. The proposed life prediction method can provide some guidance for the related standards establishment of the LED light life.