Mitigating wind energy uncertainties and operational constraints in solving the unit commitment problem in power systems through enhanced arithmetic optimization techniques

Abstract

The Unit Commitment Problem (UCP) is a critical aspect of managing operational processes in power grids, aiming to meet energy demand reliably and cost-effectively consistently. The evolution of UCP with advanced stochastic features heightens the complexity of the problem, mainly attributed to the inherent uncertainties linked to wind energy generation. This study introduces two innovative meta-heuristic approaches, namely the Arithmetic Optimization Technique (AOT) and its enhanced version, referred to as (IAOT), to address the daily UCP while considering operational and spinning reserve constraints. AOT leverages the distributional properties of fundamental mathematical operators, while IAOT enhances performance by utilizing natural logarithms and high-density values of exponential operators. Key parameters of both algorithms are thoroughly adjusted to optimize their speed and accuracy. Moreover, these selected techniques are examined for their effectiveness in solving the UCP while accommodating the considerable uncertainties associated with wind energy. The performance indicators of the chosen algorithms are rigorously validated against those of established evolutionary techniques, assessing the minimal overall generated cost, execution time, and convergence capabilities. Both techniques consistently demonstrate reductions in overall generation costs by providing ideal costs ranging from 0.02% to 6.26% across numerous situations and reference systems indicating the efficiency of AOT and IAOT in optimizing generating costs. AOT and IAOT regularly indicate percentage reductions in overall generating costs ranging from 0.10% to 6.26% for the 20-unit, 40-unit, and 60-unit reference systems. In addition, the percentage reductions reached 0.56% in the 80-unit and 100-unit reference systems, highlighting their usefulness across a wide range of system sizes. The study demonstrates the applicability and superiority of employing AOT and IAOT for determining optimal solutions to real-time UCP challenges through statistical analysis and comparative results.