Reliability constrained generation expansion planning by a modified shuffled frog leaping algorithm

This paper introduces a modified shuffled frog leaping algorithm (MSFLA) to solve reliability constrained generation expansion planning (GEP) problem. GEP, as a crucial issue in power systems, is a highly constrained non-linear discrete dynamic optimization problem. To solve this complicated problem by MSFLA, a new frog leaping rule, associated with a new strategy for frog distribution into memeplexes, is proposed to improve the local exploration and performance of the SFLA. Furthermore, integer encoding, mapping procedure and penalty factor approach are implemented to improve the efficiency of the proposed methodology. To show the effectiveness of the method, it is applied to a test system for two planning horizon of 12 and 24 years. For the sake of methodology validation, an ordinary SFLA as well as a Genetic Algorithm (GA) are both applied to solve the same problem. Simulation results show the advantages of the proposed MSFLA over the SFLA and GA.