Structural Reliability Analysis Based On Improved Ant Colony Algorithm And Particle Swarm Optimization Algorithm

Engineering structures have a lot of uncertainty，it is directly related to ourproduction practice and the safety of people’s life and property security．Therefore，considering the effect of uncertainty on security has important practical significance inthe field of civil engineering．Structural reliability theory is a powerful tool to solvestructure uncertainties and evaluate structural performance．This dissertation concludes domestic and foreign research present situation of thestructure reliability，on the basis of existing reliability analysis introduces the intelligentalgorithm for reliability calculation．The main task of this dissertation is to find a newmethod for structural reliability analysis and try to apply it to engineering practice．Themain work of this dissertation is as follows：(1)Respectively，the ant colony algorithm and particle swarm algorithm wasimproved．Through the introduction of information entropy，achieve the adaptiveregulation of ant colony algorithm．Set up the inertia weight，introduce the smoothfunction，construct the threshold function, realize particle swarm algorithm the fastconvergence of solving the constrained optimization problem．(2) Using the characteristics of the ACO and the PSO algorithm，propose a hybridalgorithm based on fusion of ant colony and particle swarm optimization，which isapplied to structural reliability analysis．Using ant colony algorithm of heuristicinformation， find out the key structural failure paths， then use particle swarmoptimization algorithm to seek the optimal structural failure paths and establish themathematical model，the structural reliability index is calculated．The hybrid algorithmrealize complementary advantages，so as to achieve the purpose of accurate and rapidanalysis of structure reliability．(3) Carry on analysis for example，the calculated results are made comparativeanalysis with the traditional calculation method and other intelligent algorithm．Thisdissertation proves that this method is effective in structural reliability analysis andcalculation．Due to multiple components，complex structure for structure system，the traditionalstructural reliability difficult to find failure paths and handling big data in the process ofcalculation．In view of the traditional calculation method is difficult to solve reliable indicator and the poor convergence efficiency of other intelligent algorithm，the hybridalgorithm of improved ant colony and particle swarm algorithm is applied to structuralreliability analysis．It achieves very good effect on the precision and efficiency and hasvery important significance for improving the level of our country structure reliability.