Modified TLBO Technique for Economic Dispatch Problem

摘要

Economic Dispatch plays a crucial role in the power system operation. The conventional problem of Economic Dispatch(ED) can be solved using classical optimization techniques. But the latest optimization techniques such as GA, PSO, TLBO etc assures better solution quality and convergence characteristics. TLBO is a simple optimization technique mimics the teaching and learning process in a classroom environment. The teacher tries to raise the mean knowledge level of the students using teacher phase. In the students phase the students interact among themselves to improve their knowledge levels. In the TLBO method the students are selected randomly during the student phase and during the teacher phase the teaching factor is considered as constant. In the present work tournament selection is incorporated for selecting the students for interaction, further during the teacher phase the teaching factor has been decreased iteration wise so as to improve the local search during the later iterations. These steps have improved the convergence characteristics of the solution. Algorithm has been developed for modified TLBO and programs are written in SCILAB software. The results obtained for three generator and six generator system considering losses are presented and discussed duly. The solutions show that TLBO and MTLBO gives almost same solution but MTLBO gives better convergence properties as can be seen from the plots.