Variable Selection Based on Hybrid Ant Colony Optimization-Simulated Annealing for Spectroscopy Quantitative Analysis of Underground Cable Tunnel

摘要: Cable tunnel is widely used in urban electric power construction but environment of cable may contain toxic and harmful gas. To analysis the densities different components mixture gases, near-infrared(NIR) spectroscopy adopted. However, full spectrum prediction model be inaccurate redundant. This paper proposes a hybrid ant colony optimization-simulated annealing variable selection method for quantitative analysis. The proposed has two parts: SA part calculate fitness new candidate solutions determine whether to accept solution, ACO generates based on pheromone deposited each uses solution update after finished. demonstrate effectiveness method, PLS with spectrum, uninformative elimination by PLS, elastic net, genetic algorithm optimization are build models real gas data results compared discussed. experimental verify precision higher.