NIR Chemical-Image Analysis Based on Adaptive Local Optimization PSO

Principal component analysis (PCA) combined with cluster analysis has become an effective approach for Near-infrared (NIR) chemical image analysis. Traditional cluster algorithms are sensitive to initial starting conditions and can be trapped into local optimal solutions. To overcome the drawbacks, we develop a new algorithm in this paper which improves Particle swarm optimization with Adaptive local optimization (ALO-PSO). Simulation experiments performed on NIR image of tablet verify the feasibility and effectiveness of the proposed algorithm. Experimental results of the clustering performances indicate that ALO-PSO algorithm offers an alternative approach for solving data clustering problems in NIR chemical image analysis.