Monitoring And Evaluation Of Damage Process Of Sparse Array Large-scale Structures

For the damage monitoring and safety assessment of large-scale structures,it is particularly important to determine the potential safety hazards in time and establish a set of safety assessment system.Large scale structures have a wide range of applications and large material area.Damage monitoring on large-scale structures is not only cumbersome in monitoring equipment,but also more difficult in signal processing.In this paper,the damage monitoring mechanism of Lamb wave on large-scale composite structures is analyzed,the layout of triangular sparse array sensors is designed,and the multi characteristic parameters of Lamb wave on typical damage are extracted.The damage orientation and degree evaluation model is established by machine learning to realize the comprehensive evaluation of damage in different directions and degrees.The content of this paper mainly includes the following specific aspects:(1)The structural health monitoring technology and the theory of Lamb wave propagation in composites is discussed,including the dispersion characteristics of Lamb wave,excitation and sensing mode and piezoelectric effect,analyzes the basic method of damage assessment based on active Lamb wave,and expounds the basic principle of machine learning.(2)The signal processing methods before and after structural damage in the traditional structural health monitoring technology(SHM)are analyzed.The mechanism of the collected signals is analyzed by combining time domain,frequency domain and time-frequency domain.The multi feature parameters of damage in different dimensions are extracted respectively to supplement additional damage feature parameters and improve the accuracy of damage assessment.(3)A triangular sparse sensor signal acquisition array is designed and established.The traditional methods of damage assessment are analyzed and compared.Support vector machine(SVM)is selected as the assessment module of damage orientation and degree,and the typical voting mechanism is used as the structure of SVM.On this basis,a damage diagnosis model based on SVM is constructed.(4)A set of Lamb wave structural damage diagnosis system based on SVM is designed.It is completed from two aspects: hardware construction and software design.Through this system,we can realize from damage location to damage diagnosis.Finally,experiments are carried out on composite plates,triangular sparse sensor arrays are arranged,and the system is verified by typical damage.SVM and other typical machine learning classification algorithms are used to compare the damage evaluation.The experimental results show that the arrangement of triangular sparse array can have a certain identification effect on the orientation and tracking evaluation of delamination damage of large-scale plate structure,and the trained model can realize the location and quantitative evaluation of damage,so as to improve the research on damage prediction and evaluation and provide valuable information for reducing potential safety hazards,It has a good application prospect.