Thesis (M. Eng.) - Asian Institute of Technology, 2011

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Energy.

This thesis proposes an improved binary PSO method for determining distribution substation location, sizing, and installation period. The proposed method uses nonlinear functions to evaluate the total investment cost of substation. The maximum sizes of substations and transformers, power balance, radial line placement and voltage drop constraints are included into the problem formulation. The problem is solved by improved binary PSO method. To validate the performance, the proposed method is tested on 9 sector area systems. Each sector area system represents different load forecast for ten-year intervals. An improved binary PSO can directly solve nonlinear objective function and constraints without the need to linearize the objective function and constraints by MILP. The substation location, sizing and installation period through improved binary PSO has an advantage over mixed integer linear programming as it leads to optimal solution in substation investment planning. The improved BPSO can overcome local optimization and has better performance of global searching. Test result indicates that the proposed method reaches a lower entire investment and annual operational cost than mixed integer linear programming in GAMS, leading to substantial cost savings.