Study On State Evaluation Of Spatial Grid Structure Model Based On Neural Network

Spatial grid structure is the most commonly used structure form in large-span structure.In recent years,safety accidents of this type of structure are frequent,so effective evaluation of its health state is the key point to prevent accidents.First of all,this paper made a 14m-long 2 20 positive quadrangle full scale grid test model,established 13 damage and non-damage working conditions,repeated sampling was conducted for each working condition through dynamic test,data extraction program was written by Matlab to screen and establish a state evaluation data set.Secondly,based on the theory of fuzzy hierarchy analysis established the underlying index(performance)for location,stress ratio and the damage area of health assessment system,through the choice of weight coefficient and membership function,draw a space truss model table and membership function of evaluation standard,hierarchy figure,etc.,and preliminary estimates of the working condition of different damage level of health status.The evaluation system established by fuzzy analytic hierarchy process(fahp)can fully consider the representativeness and operability of the factors affecting damage,and can provide an effective method of state evaluation grade for the test grid model.Finally,based on the state evaluation results of the test grid structure model by the analytic hierarchy process(ahp),a state evaluation method of the grid structure based on BP neural network is established.This method takes the AR model parameters of acceleration signals in different working conditions as the characteristic quantity,and the results of fuzzy analytic hierarchy process as the label.The data set is divided into training set and test set,and the BP neural network learning method and training set are used to train the supervised BP neural network learning evaluation model.The test set is used to test the effectiveness and accuracy of the BP neural network learning method.The study shows that the evaluation system based on BP neural network has the advantages of fast evaluation speed,high accuracy and largereduction of manual detection cost.In a word,the fuzzy analytic hierarchy process(fahp)and the evaluation method based on BP neural network studied in this paper can effectively evaluate the state of grid structure,providing reference for researchers and engineers.