Improved Newton-Raphson-based optimizer with gap evolution strategy for multiple engineering problems

Abstract

Newton-Raphson-based optimizer is a new meta-heuristic algorithm based on Newton-Raphson search rules and trap avoidance operator. To further boost its convergence performance and population diversity, an improved Newton-Raphson-based optimizer is proposed in this paper. First, the dynamic fitness-distance balance selection method is used to establish a balance between fitness values and distance, balancing the global exploration and local exploitation capabilities. Second, a novel gap evolution strategy is proposed to improve the search performance through the dual guidance of global and local gaps. Finally, the last optimization strategy is adopted to enhance population diversity, enabling a faster focus on the more promising solution space and reducing unnecessary computations and searches. We tested the performance of the improved Newton-Raphson optimizer through CEC2022 and then applied it to four challenging engineering problems involving characteristics such as multivariables, nonlinearity, and strong coupling, including parameter identification of photovoltaic model, coverage for wireless sensor network, safe path planning for unmanned aerial vehicle, and synchronous optimal pulsewidth modulation for 5-level inverter. Compared with 4 advanced algorithms, the proposed optimizer performs excellently in terms of convergence, population diversity and balance, and the experimental results demonstrate that it has the strongest competitiveness in multiple engineering application problems.