Stress point prediction method for large beam deformation based on neural network

Abstract

The invention discloses a stress point prediction method for large beam deformation based on a neural network. The stress point prediction method includes the steps: firstly, processing data calculated by a finite element model and coefficient data of a fitting function expression; using the two types of data as learning samples; training neural networks, carrying out sample expansion by using a neural network which is trained to be qualified; using the expanded sample data as a prediction neural network training sample; verifying the trained neural network; and finally, inputting a polynomialcoefficient of any large deformation curve of the beam in the prediction neural network, and obtaining a stress point predicted value of the large deformation of the beam; according to the invention,forward training of the neural network is learned. According to the stress point prediction method, the capacity of prediction neural network samples is increased, and the number of training samplesof the prediction neural network is increased, and good precision of training of the prediction neural network can be achieved, and the stress position and size of the beam can be rapidly obtained, and the problem that only deformation is solved and beam deformation stress is not solved under general conditions is solved.

Description

technical field [0001] The invention relates to the research on the prediction of the stress point of the large deformation of the rigid beam, in particular to the prediction method of the stress point of the large deformation of the beam based on the neural network. Background technique [0002] Beams can be divided into statically indeterminate beams and super-statically indeterminate beams according to structural mechanics properties. Statically indeterminate beams include simply supported beams, outrigger beams, cantilever beams, and multi-span statically indeterminate beams (rarely used in housing construction projects, but used in road and bridge projects); Statically indeterminate beams include single-span fixed-end beams and multi-span continuous beams. Beams are widely used in various engineering applications, such as longitudinal beams in automobiles, track beams in gantry cranes, girders and wing beams in aircraft, etc. Many researchers have also done a lot of res...

AI Technical Summary

Problems solved by technology

Secondly, in view of the problem that the damage and deformation of beams in previous projects cannot be judged, the damage load and the position of the stress point, the method provided by the present invention solves this problem, and can effectively judge the damage load on beam deformation and the position of the stress point

Images