Performance evaluation of approximated artificial neural network (AANN) algorithm for reliability improvement

摘要: Approximated artificial neural network (AANN) is a meta-heuristics optimization algorithm mixing the features of approximating computed combinatorial spectrum and ability to approximate input from problem domain desired output. This paper proposes use an approximated approach for case reliability when complex design considered. A mesh 256 nodes hyper-tree 496 are considered evaluating performance AANN improving minimizing cost network. Since, using formal requires substantial computational effort time equivalent NP-Hard. The work presented in this compares with that Monte Carlo simulation (MCS) particle swarm (PSO) problems. results show comparable those mentioned algorithms can be used improve reduce problems amount complexity relatively higher.