Lakhbab, Halima; El Bernoussi, Souad A hybrid method based on particle swarm optimization and nonmonotone spectral gradient method for unconstrained optimization problem. (English) Zbl 1280.90117 Int. J. Math. Anal., Ruse 6, No. 57-60, 2963-2976 (2012). Summary: The best characteristics of the particle swarm optimization (PSO) are combined with the good local search characteristics of the nonmonotone spectral gradient (NSG) to develop a novel hybrid algorithm based on PSO algorithm; the proposed algorithm is called PSO-NSG. The basic idea of PSO-NSG is to use the NSG algorithm after each \(l\) iterations of the PSO, where \(l\) is a prefixed integer. The novel algorithm can be widely applied to a class of global optimization problems for continuously differentiable functions. Simulations for benchmark test functions and also for sensor network localization problems (SNLP) illustrate the robustness and efficiency of the method presented. Cited in 1 Document MSC: 90C30 Nonlinear programming 90C59 Approximation methods and heuristics in mathematical programming Keywords:nonmonotone spectral gradient method; particle swarm optimization; sensor network localization problem Software:SFSDP PDFBibTeX XMLCite \textit{H. Lakhbab} and \textit{S. El Bernoussi}, Int. J. Math. Anal., Ruse 6, No. 57--60, 2963--2976 (2012; Zbl 1280.90117) Full Text: Link