Developing a Quadratic Interpolation Northern Goshawk Optimiser and Its Application for Solving Electrical Engineering Problems

This study presents an advanced metaheuristic approach, termed quadratic interpolation northern goshawk optimisation (QINGO), which extends the Quadratic Interpolation Optimisation to enhance the search capabilities of the northern goshawk optimisation (NGO). Like many metaheuristic algorithms, NGO faces challenges in maintaining convergence accuracy and stability, particularly when addressing complex and high‐dimensional optimisation problems. To overcome these limitations, QINGO integrates two search strategies based on the generalised quadratic interpolation method, enabling efficient searches across all dimensions of the search space and enhancing the algorithm's exploration capabilities. QINGO effectively balances exploration and exploitation, achieving superior performance. Its evaluation on a range of benchmark functions, including 23 test functions from CEC2017 and 12 complex functions from CEC2022, demonstrates consistent superiority over established optimisation techniques, such as NGO, artificial hummingbird algorithm, spider wasp optimiser, weighted mean of vectors optimisation, whale optimisation algorithm, and particle swarm optimisation. Moreover, QINGO is successfully applied to the parameter estimation of fuel cell (FC) and photovoltaic (PV) models, including two proton exchange membrane fuel cell stacks (250 W FC stack and BCS 500 W) and two PV models (single diode model and double diode model). Results highlight QINGO's robust convergence, ability to locate local and global optima, and overall supremacy over alternative methods.