Studies On Damage Identification Of Framework Base On Residual Force Vector And Genetic Algorithm

During the normal operation civil engineering structures will be suffered byearthquakes, typhoons and other natural disasters inevitably,the integrity and security ofstucture will be hit seriously,physical properties(such as stiffness,etc.),mechanicalproperties will change.If the engineering structures can run normally will draw a bigquestion mark in the heart of people,thenational economy will also beaffected.Understanding of the structure’ defect early and making good perparedness isthe purpose of structural health monitoring and damage identification is one of the mostimportant work.However, with the development of science and technology, tectonicstructures become more complex,Many traditional identification methods have beenunable to diagnose the extent and location of structure’damage,In this context damageidentification method based on vibration testing have been recognized by the majority ofscientific research for its unique advantages(practical operation,simple calculation,easyto implement),with the strengthening of computer-aided capacity,many diagnostictechniques develop rapidlly,such as through improvement residual force can overcomethe detect which the location can be determined only,the emergence of computationalintelligence greatly enhance the GA optimization capability. Residual force vectormethod has a clear concept、simple calculation and it is easy to implement, etc,analysisthe modal of the damaged structure can be achieved on the diagnosticrequirements.Genetic algorithms have high intelligent recognition accuracy and simulatebiological evolution,it been widely applied with the greatest probability to obtain theoptimal solution.This article focuses on the following aspects of the research aftersummarizing on the basis of previous research.The first chapter is an introduction which Summarized the current status ofdamage identification technology research and difficult problems,Introducedidentification methods which based on the frequency response、modal、neural networksand modal strain energy,analysed the principces of each method.The second chapter describes some methods of the signal processing in damageIdentification,The results of signal processing can identify the response of structure, suchas modal、damping, the damage index can be obtained,analysing and identifyingstructural damage based on damage index.Chapter III on the basis of previous research introduces the theory of finite elementmethod and residual force vector theory,analysis the framework elements、 beamelements、truss element finite element theory, provide a theoretical basis for solving thecontact stiffness matrix.This paper uses residual force vector identify structural damagedirectly and ues contact stiffness matrix to obtain a new residual force vectorequation,solving this equation can identify the location and extent of the structure,thewhole process used the sensitivity analysis and expanding the stiffness matrix perturbation theory of knowledge.Finally, using the proposed method combined with thefinite element method to successfully identify the damage location and extent of framestructure、 beams and trusses.The fourth chapter describes the genetic algorithm theory,the measured frequencyresponse function obtained by Analysing the Assuming known damageframework,Calculated frequency response function obtained from the geneticidentification algorithm,based on this Get the necessary objective function, whichidentified the damaged frame structure.Next in order to simulate the node identificationof frame damage, the frame unit is divided into three sections,doing so leads to excessiveidentification parameters,to reduce the identification parameters, first calculate thedamage framework residual force vector directly, initially identified damage rods, usinggenetic algorithms to optimize identify the specific impact factor, using this methodidentified the frame structure node damage successfully.It can be seen as node injurywhen the cell size is short enoughThe fifth chapter summarizes the contents of this study, presented the next researchdirection and vision for the future research.