A Modified Teaching-Learning Optimization Algorithm for Economic Load Dispatch Problem

Abstract

For the original Teaching-learning algorithm, it is weak in global search and prone to local search when solving complex optimization problems of high dimension. A modified algorithm based on space reverse-solution is proposed in this paper. Improvement of teacher phrase is based on the chaotic mapping and that of student phrase is based on the multi learning strategy. Then Self-learning phrase is added. The modified algorithm is applied to the complex high-dimensional benchmark functions for simulation experiments. Finally, the modified algorithm is applied to two typical power load distribution problems including 13 units and 40 units. The validity of the algorithm is verified from the aspects of convergence speed, convergence accuracy and stability.