Research On Damage Identification Of Grid Structure Based On Genetic Algorithm

With the rapid development of the national economy and the strong support of the state,the large-span space grid structure has been rapidly developed in recent years,it is widely used in buildings with high requirements for span and clearance,especially in some clinker banks,dry coal sheds,large equipment repair workshops.Although the production and fabrication of space truss structural components are becoming more and more standardized,random defects inevitably occur,which are mainly manifested in the defects of bar materials,the negative deviation of section area and the initial deflection caused by collision during transportation or installation.The grid structure tends to withstand some large loads during use.Under the excitation of equipment and the external loads,bolt loosening,bar bending and even fracture may occur.These damages are usually difficult to observe by visual method.With the long-term operation of the structure,local damage may accumulate and cause local or even overall collapse of the structure,which may lead to catastrophic consequences.Therefore,it is of great practical significance for eliminating potential safety hazards to carry out periodic damage detection,timely detection and early warning for existing grid structures.In this paper,the Monte-Carlo method in the finite element software ANSYS is used to analysis,the Latin Hypercube Sampling method is selected to reduce the number of samples and improve the computational efficiency.Four random input variables of material density,elastic modulus,Poisson's ratio and cross-sectional area of the member were set in the simulation process.The structural vibration frequency was defined as the output variable,and the reliability of the grid structure is analyzed.The sensitivity of the vibration frequency of the grid structure to the random variables and the correlation coefficient of the random variables are obtained.The simulation results show that among the four random input variables,the sensitivity of structural vibration frequency to elastic modulus and density is higher.When there are defects in elastic modulus and density respectively,the ratio of correlation coefficient of structural vibration frequency between them is about 3.1:1.0,and the correlation between them is opposite.Finally,the elastic modulus,which has a great influence on the frequency of the structure,is applied to the grid structure in the form of normal distribution of defects to simulate the damage identification of the structure with random defects.In this paper,the damage identification method based on genetic algorithm is used to study the damage identification of space truss structures.The principle of the damage identification method is model modification,that is,modify the non-destructive structure and make the modal information of the modified structure basically consistent with the damaged structure,and the damage location and degree was identified.In order to obtain the location and extent of the damage,the target value function is constructed in Matlab using the genetic algorithm toolbox as the corresponding damage identification index.Based on the measured vibration frequency and mode shape of the structure,the stiffness reduction coefficient of each element of the structure is taken as the parameter to be identified.The problem of structural parameter identification is transformed into one that only needs vibration analysis and genetic algorithm calculation.Finally,the genetic algorithm is used to find out the problem.The optimum individual that accords with the measured data can determine the location and extent of the damage.In the process of numerical simulation,firstly,three damage conditions of upper chord,lower chord and web member are set according to the type of members of grid structure;secondly,different damage conditions of single member are set to improve the accuracy of damage identification;finally,the defect of elastic modulus is added to the structure,and two members are damaged at the same time to simulate the damage condition.In the experimental analysis stage,the hammering method is used to excite the grid structure.The response data of the structure are obtained by the acceleration sensor arranged at the bolt sphere node.The first five modal information of the structure is output by DASP software,and the damage condition is analyzed by the structural damage identification method based on genetic algorithm.Through numerical simulation and experimental analysis,it is known that the damage detection method based on genetic algorithm can effectively identify the location and extent of damage in the grid structure.By comparing three different objective functions,it can be seen that the target function constructed by synthesizing frequency and mode shape has better recognition effect and higher efficiency.Ideal results can also be obtained by using this method for damage identification of grid structures with random defects.