MaOTLBO: Many-objective teaching-learning-based optimizer for control and monitoring the optimal power flow of modern power systems
Pradeep Jangir, Premkumar Manoharan, Sundaram B. Pandya, R. Sowmya
Abstract
This paper recommends a new Many-Objective Teaching-Learning-Based Optimizer (MaOTLBO) to handle the Many-Objective Optimal Power Flow (MaO-OPF) problem of modern complex power systems while meeting different operating constraints. A reference point-based mechanism is utilized in the basic version of Teacher Learning-Based Optimizer (TLBO) to formulate the MaOTLBO algorithm and directly applied to DTLZ test benchmark functions with 5, 7, 10-objectives and IEEE-30 bus power system with six different objective functions, namely the minimization of the voltage magnitude deviation, total fuel cost, voltage stability indicator, total emission, active power loss, and reactive power loss. The results obtained from the MaOTLBO optimizer are compared with the well-known standard many-objective algorithms, such as the Multi-Objective Evolutionary Algorithm based on Decomposition with Dynamical Resource Allocation (MOEA/D-DRA) and Non-Dominated Sorting Genetic Algorithm-version-III (NSGA-III) presented in the literature. The results show the ability of the proposed MaOTLBO to solve the MaO-OPF problem in terms of convergence, coverage, and well-Spread Pareto optimal solutions. The experimental outcomes indicate that the suggested MaOTLBO gives improved individual output and compromised solutions than MOEA/D-DRA and NSGA-III algorithms.