An improved evaluation method of industrial noise based on level of noise pollution and kurtosis-weighted cumulative noise exposure

While industry development has improved people's lives, industrial noise has also caused negative impacts. Industrial noise cannot be avoided in industrial environments. It is essential to accurately evaluate industrial noise. Most existing industrial noise evaluation methods consider the factors of industrial noise but ignore the factors of human perception and cannot accurately reflect the subjective auditory perception of humans. The paper proposes an industrial noise evaluation method, LCINE, based on level of noise pollution (LNP) and kurtosis-weighted cumulative noise exposure. First, statistical regression is used to determine the LNP constant based on improved Robinson's method. The consideration of both industrial noise and individual worker in improved method makes the constant suitable for evaluating industrial noise. Second, the improved LNP is combined with kurtosis to determine the intensity evaluation item. Kurtosis is used to capture the extremity of outliers and amplitude variations effectively in statistics, so kurtosis accurately reflects sound peaks and intensity here. Then, the exposure time evaluation item is determined by combining kurtosis with exposure time. Finally, the normalization method is used to allocate theweights of the intensity evaluation item and the exposure time evaluation item, which can unify the measurement units and reduce the adverse effects caused by extreme data. The superiority of LCINE is verified through methods including correlation test, paired-samples t test, partial regression sum of squares test, and subjective evaluation. The results indicate that the method LCINE can more accurately evaluate industrial noise and is more consistent with the subjective auditory perception of humans.