Improved Ant Colony Algorithm And Its Application In Optimization Of Truss

Truss structure has the advantage of light quality, high material utilization andstrong bearing capacity, simple manufacture, convenient construction, good assemblyand reusability, so widely used in engineering structure, then the optimization design isparticularly important. Due to the increasingly complex problems in the practicalengineering, the traditional optimization methods are difficult to adapt to, so the studyof computing intelligence method with heuristic characteristics become active.Ant colony algorithm is a algorithm of new simulation optimization inevolutionary field. Combinating distributed computing, positive feedback mechanismand greed type search, the algorithm has a strong searching ability in searching theoptimal solution. However, the long searching time and tending to the phenomenon ofstagnation is the shortcomings. Citing particle swarm optimization to speed up theprevious search and jump out of local optimum, formed the improved ant colonyalgorithm. Applied successfully in optimum design of truss structure, and comparedwith other algorithms such as genetic algorithm, PSO algorithm.In this paper, the main work has the following several parts:(1) Presenting the improved ACO algorithm based on PSO algorithm, anddesigned the operation flow chart. Firstly, producing the optimal location used thestrong global search capability of PSO, then ACO algorithm do a further adjustment;(2) Solving the section optimization problem of truss. Compiled the structureoptimization program using Matlab7.0. Throughing the analysis of typical planar10bartruss, verified the improved ACO algorithm is feasible in optimization design of truss.And studied the optimization of72space truss with stress and displacement constraintsunder multi-loading.(3) Solving the shape optimization problem of truss. Mainly including two parts:firstly, Considering the two design variables of bar cross section and the node positionof truss at the same time, using the coupling method to overcome the weaknesses oftraditional hierarchical method difficult to solve and get the global optimal solution.Secondly, solving the problem of structural optimization with dynamic constraints,realized the shape optimization design of truss with local stability constraints. Throughthe example analysis and compared with the existent literature, results show that the improved ACO algorithm can well solve truss shape optimization problem.The successfull application in truss optimization with the improved ACO algorithm,obtained the best cross section and shape, increased the overall stiffness of the truss,thenlowered the construction cost. So the significance of this study is to provide a feasibleoptimization algorithm, which can be widely used in practical engineering, and producemore obvious economic benefits.