Intelligent Identification of Multicomponent Mixture Components Using an Artificial Intelligence
Abstract:
This article presents an advanced method for the qualitative and quantitative identification of components in mul ticomponent mixtures using two densitometers and artificial intelligence (AI). The proposed approach eliminates dependence on chromatographic operating conditions by incorporating a standardized dual-detector system and a mathematical model for determining molecular mass. Regression analysis and neural networks enhance the model’s performance, enabling automated, high-precision identification without manual calibration. This article presents an advanced method for the qualitative and quantitative identification of components in multicomponent mixtures using two densitometers and artificial intelligence (AI). The proposed approach eliminates dependence on chro matographic operating conditions by incorporating a standardized dual-detector system and a mathematical model for determining molecular mass. The core equations are enhanced through regression analysis and neural network techniques, enabling automated and high-precision identification without manual calibration. The article introduc es Chrom AI ID Pro, a dedicated software platform that integrates signal acquisition, real-time analysis, AI-based predictions, visualizations, and report generation. The system utilizes machine learning algorithms such as Random Forest and MLP Regressor to predict molecular weights based on detector signals and known standard param eters. Comparative simulations demonstrate a significant reduction in prediction error and improved reliability under variable conditions. Experimental data and illustrative examples, including model accuracy comparisons and graphical outputs, are provided to validate the efficiency of the method. The proposed solution is applicable to industrial and laboratory settings where fast, accurate component identification is essential.
