When the coupling of column height,load,structural size and other variables affects the structural performance of stacker columns,the numerical simulation calculation will become very complicated and the workload will increase sharply,which makes it difficult to meet the actual needs of large-scale personalized agile design of enterprise stackers.Through numerical simulation and orthogonal test,the mechanical properties of variable cross-section columns of 25-40 m high-rise stackers were analyzed.Based on this,an intelligent prediction model of column performance was established by convolution neural network algorithm.Compared with the finite element calculation results,the average relative error is less than 10%and the calculation efficiency is greatly improved,which provides a new enabling tool for rapid evaluation of personalized design of high-rise variable cross-section columns.
stackernumerical simulationconvolutional neural networkprediction modelorthogonal test methodvariable cross-section column
万方数据
维普
