Application of enhanced sand cat swarm optimization algorithm in weapon target assignment problem

The efficient allocation of available ammunition to threat elements in the defence industry is of critical importance for resource management and operational success. Known in the literature as the Weapon Target Assignment (WTA) problem, this issue is a multi-dimensional complex optimization problem where the solution space grows exponentially with the increase in the number of targets and weapons. In this field, where classical deterministic methods fail to produce solutions within reasonable timeframes, meta-heuristic algorithms have gained prominence. In this thesis study, a novel and hybrid algorithm named Adaptive Golden Sine Orthogonal Sand Cat Swarm Optimization (AGO-SCSO) is proposed based on the nature-inspired Sand Cat Swarm Optimization (SCSO) algorithm for the effective solution of the WTA problem. To overcome the structural limitations of the original SCSO algorithm, such as premature convergence and entrapment in local optima, particularly in high-dimensional problems, three fundamental mechanisms responsive to problem dimension (adaptive) have been integrated into the proposed method. In low-dimensional problems, population diversity is maintained via Gaussian Perturbation and momentum strategy, while in high-dimensional problems, solution precision is improved through the Golden Sine mechanism, and escape from local traps is ensured using the Orthogonal Learning strategy. The performance of the developed algorithm was validated in two stages. In the first stage, tests conducted on 20 standard benchmark functions revealed that AGO-SCSO achieves high convergence precision in unimodal functions and outperforms the original algorithm in multimodal functions. In the second stage, WTA problems of varying scales, ranging from 5 to 200 targets, were simulated. Experimental results and Wilcoxon statistical analyses demonstrated that AGO-SCSO exhibits statistically significant and stable success in minimizing threat cost compared to Spotted Hyena Optimization (SHO), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), and the original SCSO. Even with a high-dimensional problem like SHA200, AGO-SCSO has maintained its competitive edge over other methods with a runtime of 0.59 seconds and a best solution value of 2036.37. Consequently, the developed AGO-SCSO algorithm is presented to the literature as a robust and scalable method for solving complex military optimization problems.