Machine Learning to Predict Junction Temperature Based on Optical Characteristics in Solid-State Lighting Devices: A Test on WLEDs

Abstract

While junction temperature control is an indispensable part of having reliable solid-state lighting, there is no direct method to measure its quantity. Among various methods, temperature-sensitive optical parameter-based junction temperature measurement techniques have been used in practice. Researchers calibrate different spectral power distribution behaviors to a specific temperature and then use that to predict the junction temperature. White light in white LEDs is composed of blue chip emission and down-converted emission from photoluminescent particles, each with its own behavior at different temperatures. These two emissions can be combined in an unlimited number of ways to produce diverse white colors at different brightness levels. The shape of the spectral power distribution can, in essence, be compressed into a correlated color temperature (CCT). The intensity level of the spectral power distribution can be inferred from the luminous flux as it is the special weighted integration of the spectral power distribution. This paper demonstrates that knowing the color characteristics and power level provide enough information for possible regressor trainings to predict any white LED junction temperature. A database from manufacturer datasheets is utilized to develop four machine learning-based models, viz., k-Nearest Neighbor (KNN), Radius Near Neighbors (RNN), Random Forest (RF), and Extreme Gradient Booster (XGB). The models were used to predict the junction temperatures from a set of dynamic opto-thermal measurements. This study shows that machine learning algorithms can be employed as reliable novel prediction tools for junction temperature estimation, particularly where measuring equipment limitations exist, as in wafer-level probing or phosphor-coated chips.

An analytical and machine learning model for SPD estimation and its prediction for lumen and chromaticity shift based LED lifetime performance analysis

“…According to the study, there are various ML algorithms that can help with LED reliability assessment and lifetime prediction. The study [35] findings indicate that knowing the power level and colour attributes is sufficient information for prospective regressor trainings to forecast any white LED junction temperature. ML methods such as k-Nearest Neighbour (KNN), Radius Near Neighbours (RNN), Extreme Gradient Booster (XGB) and Random Forest (RF) are employed.…”

Holographic thermal mapping in volumes using acoustic lenses

“…To generate precisely targeted acousto-thermal patterns, discrepancies between the imposed and forward-propagated pressure fields necessitate iterative simulations and processing efforts. In a variety of applications, deep learning techniques have been offered as a solution to inverse heat and design problems [42][43][44][45][46][47][48]. For the first time, we use a machine learning (ML) system to compress the process into a single inverse issue for speedy and efficient holographic plate design.…”

A Critical Review on the Junction Temperature Measurement of Light Emitting Diodes

“…The authors claimed that with a ratio of 0.005, the temperature prediction accuracy of 1 K can be achieved in commercial white LEDs with this relation. In constant forward current density, Azarifar et al [ 87 ] performed four machine learning regressions including k-nearest neighbor (KNN), radius near neighbors (RNN), random forest (RF), and extreme gradient booster (XGB) on temperature sensitive optical data from over 500 commercial white LED packages and tested the accuracy of prediction with experimental measurements. With near unity in R 2 scores and small root mean square deviation values, the XGB regressor showed close-to-perfect correlation capability to assess T j based on SPD behavior.…”