Research On Response Analysis Of Structure Random Vibration And Random Loading Identification Based On Orthogonal Decomposition

In the aerospace field, random vibration phenomenon is widespread. For example, in the process of propelling the rocket, the satellite receives random excitation from jet noise, pulsating thrust, turbulent boundary layer noise, and the vibration transfers to satellite by the adapter between the rocket through the satellite, affecting the equipments' performance and reliability. Determine the random loading excitation on the satellite, to get the response of the satellite, according to the results of response to do structure design and optimization, and to estimate the structural reliability and satellite life are the main tasks of satellite structure and strength design. Therefore, the random vibration response analysis has an important role for reducing the spacecraft development costs and r&d cycle, and ensure that the structure and equipment safe and reliable.Random vibration analysis of spacecraft structure is the background of this paper, and random vibration analysis for linear structures, nonlinear structure, and random structure are in-depth studied in this paper, the research contents are as follows:Stochastic process orthogonal expansion expressions is introduced, precise integration format is derived to accurately obtain the K_L orthogonal decomposition vector of the non-stationary filtered white noise random excitation, the energy concentration characteristics of K_L vector is found by simulation analysis; conventional methods of calculating linear structure random vibration response are introduced in the paper, and direct method of response variance analysis is used for the white noise and non-white noise random vibration response analysis; Orthogonal decomposition of non-stationary filtered white noise signal is used to solve non-stationary random vibration response of linear structure, and by using the K_L energy concentration characteristics, it increases the computational efficiency.For Random vibration of nonlinear system response analysis, equivalent linearization method based on two equivalent criteria of equation difference and energy difference is introduced. For the problem of iteration process of equivalent linear method's efficiency is not high, direct method of response variance analysis is proposed to solve the nonlinear system stationary random response, and orthogonal decomposition method is proposed to solve the nonlinear system non-stationary random response, comparison and simulation analysis is done to verify the correctness and feasibility.For double random vibration response analysis of random structure, orthogonal decomposition method of random structure is introduced to obtain the equivalent deterministic system expanding equation, the orthogonal decomposition method of non-stationary random excitation and Ritz dynamic condensation method are proposed to conquer the order increasing problem caused by the expanding equation and to improve the computation efficiency of response analysis, and simulation and analysis work is done.Stationary random loading identification is study in the paper. For random load identification error near the natural frequency, combined with inverse pseudo excitation method of load identification, the causes of the errors is deeply studied, and method of conditions of frequency response function is proposed to select the measurement points. Based on inverse pseudo excitation method, the condition number weighted algorithm is proposed to reduce the errors caused by the ill-conditioning during the process of random load identification, and simulation analysis is done to verify the correctness and feasibility.