Balancing Convergence and Diversity in Meta-Heuristics for Calibration of Instrument Transformers

Instrument transformers (ITs) are integral components of power systems characterized by a certain level of failure probability. Monitoring the errors of ITs and promptly replacing faulty ones is important for ensuring the secure operation of power systems. Existing online calibration methods exhibit specific problem formulation and optimization algorithm selection issues, rendering them inadequate to meet practical engineering requirements. To address the above deficiencies, we first formulate an optimization problem that aligns with the practical considerations of engineering. Notably, parameter normalization and objective integration are incorporated for engineering considerations. Next, we propose a hybrid meta-heuristic algorithm that balances convergence and diversity for handling variables in the complex number field and the nonlin-earity of the formulated optimization problem. This algorithm is expected to fill the gap between mathematical local search and the meta-heuristic global search for real-world applications, considering the accuracy and reliability of the optimization results simultaneously. Numerical validation results indicate the superiority of the proposed method in terms of convergence rate and assessment precision over existing algorithms.