Study Of Particle Swarm Optimization Method And Development Of Software For Overhead Travelling Crane Girders

At present, The optimal design methods of cranes structure mainly aretraditional optimization algorithms, but these algorithms to solve such highdimensions, many constrained, nonlinear, hybrid and discrete variable problems,will appear inevitably lots of problems, such as consuming too large amount ofcalculations, killing longer time and easy to falling into the local optimalsolution. This paper tries to use a new type of group intelligent optimizationalgorithm-Particle Swarm Optimization(PSO) to solve the problem.This paper, from the traditional structure design, taking the typical doublebox overhead travelling crane girders for example, according to the maintechnical parameters requirements, chooses the preliminary sectionsizes,calculates loads and internal forces,and checks the strength, stiffness andstability whether meetting the use performance requirements. In guaranteeingthe main technical parameters are same premise, the weight of girder structurerespecting as the objective function, section sizes as optimization variables,strength, stiffness and stability acting as constraint conditions, The paper uses atraditional optimization algorithm-Grid Method, and a new intelligentoptimization algorithm-PSO to conduct the optimize design.PSO, originating in the research of biological birds' group movementbehavior, getting through the principle of cooperative and competitive betweenindividuals in group to realize the solution of the optimization problems, is apopulation intelligent optimization algorithm. It is provided with simple theory,less parameters, rapider convergence speed. The algorithm has been widelyapplied in neural network training, function optimization and other fields, andget a good effect. Because of the complexity and particularity of the crane metalstructure optimization design, this paper analyzes deeply the principle andprocess of the basic particle swarm optimization algorithm, and applies theimproved particle swarm optimization algorithm to a project case which tooptimiz the structure of overhead travelling crane.The optimization results showthat the algorithm has good performance characteristics. The project by using object oriented software Visual C++6.0develops a setof particle swarm optimization CAD software of overhead travelling cranegirder. The software integrates the main technical parameters input, conventionaldesign, optimal design, and parameter drawing module, uses the database, twodimensional parametric drawing technology, and realizes these functions oftraditional calculation, optimal design, and2d drawing output.