Elitist teaching-learning-based optimization (ETLBO) algorithm is inspired by practical teaching-learning process. A novel group search optimizer

modified elitist teaching-learning-based optimization (mETLBO)

was proposed to improve low precision and poor stability of the ETLBO. First

an autonomous learning process was introduced to strengthen local search of high quality solution so as to improve algorithm's elite-searching speed. Second

differentiated support and self-adaptive strategy providing appropriate and flexible learning approach to students at various levels

were applied to offer desirable assistance and balance searching rate and accuracy of the algorithm. Third

global searching ability of the algorithm was enhanced by increasing communication frequency between students. Optimization results on standardized functions show that the proposed algorithm is obviously superior to the original one in performance and efficiency. Finally

satisfactory results were achieved by applying the improved algorithm to process optimization with mechanism model of methanol synthesis.