Fault detection and diagnosis algorithms for HV-LED lamp driven by segmented linear solution using first-order difference calculation and peak-detection methods

Abstract Fault detection and diagnosis (FDD) algorithms based on first-order difference calculation and peak-detection methods for HV-LED lamp driven by segmented linear solution is proposed. The proposed FDD algorithms analyze the illuminance fluctuation characteristics of HV-LED lamp, and then identify whether the HV-LED lamp occurs fault or not and confirm the fault-type. The proposed FDD algorithms can be divided into five steps: illuminance signal smoothing, illuminance signal process using first-order difference calculation, difference-value sequence smoothing, peak detection of difference-value sequence, and fault-type recognition. Taking the HV-LED lamp driven by four-segment linear driver for example, the illuminance data sets of the fault-free and three different faults are investigated for algorithm verification experiments. The experimental results demonstrate the proposed FDD algorithms can effectively detect and diagnose the fault-type of HV-LED lamp driven by segmented linear driver, and locate the faulty HV-LED sub-strings.