Gas Leakage-Based Fault Prediction for a Boron-Coated Proportional Counter

Large reaction cross-sections, low operating voltage, extended lifespan, good stability, and suitability for high-temperature and intense gamma radiation settings are among the features of boron-coated proportional counter tubes. Consequently, it has been extensively employed in the nuclear-measuring system outside of the reactor, where the source range detector is a boron-coated proportional counter tube. This paper studies boron-coated proportional counter tubes using a comprehensive approach that combines Monte Carlo simulation with lifespan prediction in health management. The wavelet packet decomposition approach was used to extract feature information from the pulse signals of the boron-coated proportional counter tube. A counter tube defect prediction model was established based on a multi-kernel correlation vector machine, where the impact of different wavelet packets and correlation vector machine kernel functions on the model's prediction performance was examined. The study's findings demonstrated that the particle swarm optimization algorithm-optimized multi-kernel correlation vector machine model exhibited superior prediction accuracy for the counter tubes' future operating conditions. The primary goal is to comprehend how gas loss affects gas pressure and to determine the lifespan of detectors within a detector system.