A Graphic Processing Unit (GPU) Algorithm for Improved Variable Selection in Multivariate Process Monitoring

Process monitoring is extremely important for producing high quality product and at the same time ensuring safe working environment in chemical process industry. Recently, it has been shown that selection of an appropriate subset of variables can improve the monitoring performance. The main contribution of this work is the development of a parallel version of the Genetic Algorithm-Principal Component Analysis algorithm which was proposed by Ghosh et al. [2] for variable selection. The developed algorithm has been implemented using NVIDIA's Compute Unified Device Architecture, CUDA parallel computing platform. Experimental results show that the proposed parallel approach is 12 times faster than the original serial code when applied to the Tennessee Eastman challenge problem.