Detection of chemical oxygen demand in water based on UV absorption spectroscopy and PSO-LSSVM algorithm

Abstract

A method of detecting chemical oxygen demand (COD) of water based on ultraviolet (UV) absorption spectra is proposed. The modeling and analysis of the standard samples and the actual water samples are carried out respectively. For the standard solution samples, the univariate linear models based on single wavelengths and the partial least square (PLS) model based on synergy interval partial least square (SiPLS) and moving window partial least square (MWPLS) are established. For the actual water samples, different pre-processing methods are used. SiPLS and MWPLS are used to select the characteristic bands. The least squares support vector machine algorithm optimized by particle swarm optimization (PSO-LSSVM) algorithm is used to establish the prediction model, and the prediction results of various models are compared. The results show that the optimal model is PSO-LSSVM which uses SiPLS to select the characteristic bands of the first derivative spectra (preprocessing method). The determination coefficient of the prediction set is 0.963 1, and the root mean square error of prediction (RMSEP) is 2.225 4 mg/L. PSO-LSSVM algorithm has good prediction performance for the analysis of COD in actual water samples by UV spectra. This paper provides a new design idea for the research and development of water quality detection optical sensor.