The quantitative and quantitative analysis of trace polycyclic aromatic hydrocarbons (PAHs) was carried out based on three-dimensional fluorescence spectroscopy combined with alternating penalty four linear decomposition (APQLD). The experiment was carried out with acenaphthene (ANP) and naphthalene (ANA). First, in order to solve the redundant information of the three-dimensional fluorescence spectral data, the experimental spectral data was compressed by wavelet transform. The four-dimensional data were constructed by using the combination of excitation and emission spectra of PAHs in ethanol solvent and methanol solvent and ultrapure water respectively. The four-dimensional spectral data were analyzed by APQLD and compared with PAHs under three solvent conditions The respective recovery rate. The experimental results showed that the higher order data can be used to determine the concentration of PAHs more accurately under the three solvent conditions, and the recovery rate was higher. Compared with the second-order correction and other four-dimensional correction algorithms, APQLD can better reflect the superiority of the four-dimensional algorithm The recovery rate of ANA was 96.5%~103.3% and the predicted root mean square error was 0.04 μg·L-1. The recovery rate of NAP was 93.3%~110.0%, and the predicted root mean square error was 0.08 μg·L-1.