Multi-objective driving training-based optimization algorithm for optimizing power losses and voltage profile in power systems

Managing active power losses and maintaining an acceptable voltage profile are two conflicting yet critical objectives in modern power system operations. This paper presents a novel Multi-Objective Driving Training Based Optimization (MODTBO) algorithm to solve the multi-objective optimal power flow problem. Unlike the original mono-objective DTBO algorithm which operates on three separates phases, the proposed multi-objective variant streamlines the search process into two main phases: instructor training, and combined patterning and personal practice. To effectively handle the multi-objective space, MODTBO integrates non-dominated sorting and crowding distance mechanisms to rank the population and adaptively select driving instructors. The performance of the MODTBO algorithm is evaluated on standard IEEE 30-bus system by simultaneously minimizing total active power transmission losses and bus voltage deviations while satisfying strict equality and inequality constraints. Furthermore, a Fuzzy Logic-based approach is employed to extract the best compromise solution from the generated Pareto-optimal front, aiding operators in decision-making. Simulation results demonstrate that MODTBO provides a well-distributed Pareto front with high intensification and diversification. It outperforms the single-objective Modified Driving Training Based Optimization (MDTBO) frameworks, specifically those focusing individually on either voltage profile enhancement or active loss minimization, by successfully compromising both conflicting goals in a single optimization run with superior convergence. Consequently, the proposed algorithm offers an efficient and practical tool for power system operators to balance system efficiency and voltage stability.