Optimal wavelength selection method of absorption spectroscopy for simultaneous detection of nitrates and nitrites

Simultaneous detection of nitrate and nitrite using absorption spectroscopy still faces many challenges due to the severe overlap of their spectra. Therefore, it is necessary to select optimal wavelengths to enhance the detection capability. The existing wavelength selection methods, including observational method, traversal method, derivation method, and optimization algorithm, have limitations in extracting global features from the entire spectral range. In this paper, an optimal wavelength selection method is proposed based on meta heuristic optimization learning, named the Modified Crayfish Optimization Algorithm (MCOA), which integrates global spectral information not only focusing on discriminative spectral regions. Specifically, an innovative evading predator stage was added to the basis of the traditional crayfish optimization algorithm (COA). During the iteration process, the overlapping wavelength regions that were initially searched by traditional COA were marked as predator snake. By simulating the behavior of crayfish escaping from predators, the selected wavelengths were guided to migrate to unexplored regions, thereby enhancing the global search capability for optimal wavelengths. Finally, the backpropagation (BP) neural network was used to predict the concentrations of nitrate and nitrite in the mixed samples, and the proposed method achieved the lowest Root Mean Square Error (RMSE) of nitrite and nitrate respectively, compared with observational method, traversal method, interval splitting method, Simulated Annealing (SA) Algorithm, and traditional COA. The results show that the proposed method achieves optimal wavelengths selection across the entire spectral range, and the selected wavelengths with richer information improves the accuracy of simultaneous detection of nitrite and nitrate.