Optimization Of Crane Main Beams Based On Dynamic RBF Agent Models And Evolutionary Algorithms

As an important equipment in the field of construction equipment,cranes have played an essential and vital function in the development of domestic market.Concomitant with the advancement of modern design and the evolution of the building process,in line with the better performance of cranes,crane design requirements are more stringent,in the field of crane design structure lightweight research is always a matter of concern.The main beam as the main load supporting components of the crane,its heavy weight of approximately 60% of the entire machine,it is also natural to become the crane of the lightweight research focus object.The emergence of intelligent optimization algorithms to structural optimization problems to the climax,but indispensable also brings new problems,intelligent optimization algorithms are huge computational volume and for the need for finite element simulation analysis of optimization problems,especially complex model time is immeasurable.Therefore,it is important to study how to improve the efficiency and reduce the cost of structural optimization of bridge machines.The proxy model technique is widely used in complex structural optimization problems because it can proxy the simulation analysis of the model to cut down the number of calls to the simulation model and decrease the cost,but it is less used in the lightweighting of crane structures.To address the problem that the computational effort of crane structure optimization based on finite element simulation model is unaffordable in engineering,a global optimization strategy based on dynamic radial basis agent model is proposed by combining differential evolution algorithm and radial basis agent model.A dynamic radial basis agent model is constructed during the optimization process through an adding point strategy of locally exploiting the optimal solution and globally exploring the region with the largest error.The strategy constructs a dynamic radial basis agent model in the optimization process by using an additive strategy of locally developing the optimal solution and globally exploring the region of maximum error,and constructs an optimization termination condition with the estimation error of the bound function model and the drop level of the goal function to guarantee the accuracy of the global convergence of the optimization and the exactness of the model at the optimal solution.The approach is verified by numerical cases and I-beam optimization cases,which not only can obtain the global best resolution,as well as a distinct reduction in the recurrence of the original function,but also a significant improvement for the improvement of the simulation rate.In the end,together with the finite component theory of the bridge crane bridge,this method is used to solve the main beam optimization problem.The results show that under the condition that the constraints are met,the main beam cross-sectional area is reduced by about 22.36%,saving a large amount of computational cost,improving the optimization efficiency and solving the expensive calculation cost problem resulted from the direct combination of the intelligent swarm algorithm and the finite element model of the crane structure for optimization.