Research On Digital Twin Risk Prevention And Control Technology Of Beam Structure Based On Dynamic Load Identification

Accidents caused by structural safety problems of hoisting machinery often cause serious economic losses,even threaten personal safety and cause casualties.For the structure that often bears dynamic load,it is very necessary to accurately grasp its actual bearing situation and reduce the risk that abnormal dynamic load may cause to the structure.Taking the beam structure in hoisting machinery as the research object,this thesis proposes a digital twin risk prevention and control technology based on dynamic load identification.The dynamic load identification is carried out by real-time monitoring information,and a virtual model is established according to the dynamic load identification results,so that the virtual model and the physical model interaction to realize condition monitoring and early warning of beam structures.The specific research contents are described as follows:(1)Overall design of beam structure digital twin risk prevention and control technology based on dynamic load identification: The design goals and design principles of digital twin risk prevention and control technology are determined,and the overall structure is clarified.According to the functional requirements,the functional is divided into related modules.At the same time,the key problems of this thesis are expounded from the establishment of the dynamic load identification model and the establishment of the digital twin system,and the relevant solutions are proposed.(2)A dynamic load identification method based on Newmark method is proposed:Aiming at the limitations of the current dynamic load identification research methods in practical application,this thesis uses the time domain method based on direct integration to process the motion equation,and selects the Newmark method to discrete the time domain.By separating the position and size information of the load and constructing the objective function,a dynamic load identification method is proposed and verified.In the process of method verification,the effects of gradient descent method and genetic algorithm in solving the objective function are compared.According to the comparison results,the gradient descent method is selected to solve the objective function to improve the solution efficiency.At the same time,the dynamic load identification accuracy of the method under different influencing factors is analyzed.It further shows that the dynamic load identification method based on Newmark is feasible and has good robustness.(3)Algorithm optimization of dynamic load identification method based on Newmark method: Aiming at the problems of low search efficiency and slow convergence speed when using the traditional gradient descent method to solve the objective function of the dynamic load identification method based on Newmark method,the Lyapunov function is constructed to explore the stability of the objective function of the dynamic load identification method,so as to obtain the best step size of each iterative calculation.At the same time,the mechanical constraints are set on the search initial value in the gradient descent method through the deformation theory of beam in material mechanics to get a better initial search value.Finally,the accuracy of the optimization algorithm is verified.(4)System implementation: The main realization process of the digital twin risk prevention and control system is introduced,and the development of the Web system in the system realization process is discussed in detail.The research shows that under the premise of unknown beam structure bearing conditions,through the digital twin risk prevention and control system in this thesis,the full structure response data of beam structure can be obtained in real time,the state monitoring and early warning functions can be realized,and the safety of the structure can be effectively improved.At the same time,the digital twin risk prevention and control system has good practicability and scalability.