The degradation behavior and lifetime of LEDs for field operation is currently predominantly estimated using LM-80 data in combination with TM-21 extrapolations. If the environmental conditions differ significantly from the laboratory conditions of the LM-80 tests, differences in in-situ degradation behavior and TM-21 predictions could occur. Independently of the previous described methods, the following publication evaluates the use of an electro-optical model in order to allow an estimation of the in-situ degradation dynamics. The correlation between optical degradation and forward voltage decrease is determined on the basis of degradation tests performed on green high-power LEDs.

Depending on the forward current selected, the correlation can be described by a linear or logarithmic function and can be used to build an electro-optical model that takes the degradation behavior for different operating conditions into account.

A final validation of the model shows that the degradation behavior of individual LEDs can be determined, while a comparison of the measured mean values of four test LEDs with the model predictions shows a maximum root-mean-square error (RMSE) of 0.012. Accordingly, the method offers the possibility of condition monitoring that can be transferred into a practical measurement circuit.

当前，主要使用LM-80数据结合TM-21外推法估算用于现场操作的LED的退化行为和寿命。如果环境条件与LM-80测试的实验室条件明显不同，则可能会发生原位降解行为和TM-21预测的差异。独立于先前描述的方法，以下出版物评估了电光模型的使用，以便可以估计原位降解动力学。光衰减与正向电压下降之间的相关性是根据对绿色大功率LED进行的衰减测试确定的。

根据所选择的正向电流，相关性可以通过线性或对数函数来描述，并且可以用于建立考虑到不同工作条件下的退化行为的电光模型。

该模型的最终验证表明，可以确定单个LED的退化行为，而将四个测试LED的测量平均值与模型预测值进行比较，则可以得出最大均方根误差（RMSE）为0.012。因此，该方法提供了状态监视的可能性，该状态监视可以被转移到实际的测量电路中。