Probabilistic Reliability-based Optimization For Overhead Traveling Crane Metal Structure

Overhead traveling crane is used more and more frequently, and is developed towards large-scale. As the skeleton of the crane, metal structure makes up about 60% ~ 80% of the whole crane weight, so lightweight design is needed. In addition, failure usually occurs on metal structure, extremely easy to lead to a major accident, so reliability requirements are also needed.The allowable values of strength, stiffness and stability are often used to be the subject to limit the variable when to optimize the overhead traveling crane metal structure. But structure cannot always be secured with the optimization solution. In the design phase, economy and reliability requirements of the metal structure need to be considered all together. The combination of reliability analysis and optimization design can be realized by reliability-based optimization design. The reliability and economy requirements can be balanced by reliability-based optimization design. In this paper, probabilistic model is used to describe the uncertain factors of the structure, and the application of reliability-based optimization design is applied in the design of overhead traveling crane metal structure. The main research contents are as follows:(1) Based on improved PSO(Particle Swarm Optimization) for the overhead traveling crane metal structure optimization design, high accuracy,fast convergence, and the improvement of the optimization design efficiency are realized; Based on the Monte-Carlo method for the reliability analysis of traditional manual design solution and general optimization solution, the conclusions can be drawn that the former often has high reliability, while the latter can improve the design efficiency and realize the lightweight structure,but couldn't get a satisfactory reliability level.(2) By introducing the stress level to measure the design margin,optimization design of structure is performed with different stress level, and then the reliability analysis of optimization solution is performed at the same time. The influence principle of stress level on optimization and reliability isfound.The conclusions can be drawn that: with the decrease of stress level, the quality of overhead traveling crane metal structure is gradually increased;With the decrease of the stress level, the reliability of the structure is improved,but the effectiveness is smaller and smaller, and at last becomes a line tending to 1.In other words, the rate of return by improving the design margin in exchange for higher reliability is getting lower and lower.(3) Based on the nonlinear mapping and regression fitting techniques of the Artificial Neural Network, direct function relation between the random design variable and structure reliability is simulated. The mathematical model of probabilistic reliability-based optimization design was established by this relation to predict the structure reliability. The optimization solution with the requirement of economy and reliability is obtained.(4) Based on VC++ 6.0, overhead traveling crane metal structure design software was developed to realized visualization design, optimization design,reliability analysis, Artificial Neural Network predicting structure reliability,and probabilistic reliability-based optimization design.The innovation of this paper lies in: By introducing the improved Particle Swarm optimization combined with penalty function method, the constrained optimization problem can be solved effectively; By introducing the stress level to measure the design margin, the effect of stress level on optimum design and structure reliability is obtained; Based on the Artificial Neural Network, direct function relation between the random design variable and structure reliability is simulated to improve the calculation efficiency of probabilistic reliability-based optimization.