A hybrid teaching–learning-based optimization technique for optimal DG sizing and placement in radial distribution systems

Abstract

Distributed generation (DG) technology has proved to be an efficient and economical way of generation of power. DGs are intended to generate power near the load centers. Optimal allocation of DG resources enhances the overall performance of distribution systems. This paper presents a hybrid teaching–learning-based optimization (HTLBO) technique for the optimal allocation of DGs in distribution systems. The proposed technique is proficient in handling continuous as well as discrete variables and has the capability to escape strong local minima/maxima trappings. The validity and effectiveness of HTLBO are tested on well-defined standard mathematical benchmark functions. The proposed method is further implemented for optimal allocation of DGs in the IEEE 33-bus, 69-bus and 118-bus radial distribution test systems for minimization of power losses, voltage deviation and maximization of voltage stability index. The multi-objective function for DG allocation uses the ɛ-constraints approach. The obtained results reveal improved convergence characteristics over both teaching–learning-based optimization and quasi-oppositional teaching–learning-based optimization.

Appendix

A MOF optimizes all the SOFs at the same time, subject to the equality and the inequality constraints. In this proposed work, a MOF (Sultana and Roy 2014) simultaneously minimizes F₁ and F₂ along with maximization of F₃.

$$ {\text{MOF}} = {\text{Minimize}}\left[ {b_{1} *F_{1} + b_{2} *F_{2} + b_{3} *(1/F_{3} } \right)] $$

(A.1)

If ‘m’ is total number of objective functions and $b_{i}$ are penalty coefficients,

$$ b_{i} \in \left( {\left[ {0,1} \right]} \right)\quad {\text{and}}\quad \mathop \sum \limits_{i = 1}^{m} b_{i} = 1 $$

If DGs are allocated in RDS with the purpose of meeting a particular objective, the corresponding penalty coefficient value is increased. The priority of objective function in the MOF decides the value of the penalty coefficient. However, the sum of the penalty coefficients must be unity for a normalized objective function.