Modeling Of Extreme Load Based On EM Algorithm And Application

The design of controllable fatigue life of key mechanical parts is an important means to ensure the service life and reliability of construction machinery.The preparation of the load spectrum is the prerequisite and basis for the design of fatigue life.In the process of designing the mechanical structure,only the full load information provided by the load spectrum can be used to ensure the reliability of the actual use process in the case of cost saving.In the fields of aerospace,water conservancy,construction and so on,the application of load spectrum has been more common and the compilation process of load spectrum has been more standardized.Some key technical problems have been studied in depth.In the field of engineering machinery,the preparation of load spectrum is not yet mature,and many problems still need to be studied.In particular,the measure of the extreme value of the load is very important because the extreme load has a significant effect on the mechanical structure.The study of the extreme load and the statistical method are very important for the study of the reliability design.Therefore,in order to study the statistical measurement of extreme load in the load spectrum of construction machinery,this paper presents a distribution model of extreme load,and designs the corresponding parameter estimation algorithm.The model and method are simulated,The validity of the algorithm and the parameter estimation algorithm are verified and further verified on the test load data.The main contents of this paper are as follows:(1)A statistical distribution model is proposed for the extreme value of the load.Based on the preliminary statistical analysis of the real load data,this paper presents the basic assumptions about the statistical model.On the basis of the hypothesis,a statistical model describing the distribution of extreme loads is proposed,and the mathematical expression of the model is given.(2)Based on the basic idea of EM algorithm,a parameter estimation algorithm for extreme value distribution model is designed.Through the derivation and analysis of the formula,this paper divides the more complex likelihood function in the algorithm into a simple likelihood function combination,which simplifies the calculation process.Then the algorithm is implemented by programming,and the computational efficiency of the program is optimized by parallel computing technology.This paper introduces the principle of another parameter estimation algorithm for comparative analysis.(3)Through the simulation data,the performance of the algorithm is analyzed.The mathematical model and program flow of the simulation data are designed.The model fitting and parameter estimation of the simulation data of single group are carried out to verify the fitting effect of the model and the unbiasedness of the parameter estimation.The validity of the parameter estimation algorithm is analyzed by analyzing the 1000 sets of simulation data.And compared the effect of another parameter estimation algorithm.The results show that the proposed algorithm has better effect,and the estimation error of the key parameters is less than 20%,and it is less than the true value.(4)The model and algorithm proposed in this paper are proved by the chi-square test,which is consistent with the experimental data.And through the use of self-help method to re-pumping thousands of sets of sampling,through repeated tests again on the effectiveness of the parameters of the algorithm is analyzed.The results are basically the same as the simulation data.In this paper,a statistical measurement model and a parameter estimation method for this model are proposed for the extreme load.The applicability of the model and the effectiveness of the algorithm are verified by means of simulation data analysis and experimental data analysis.Through the method of this paper,it is convenient and quick to measure the extreme load of engineering machinery,and establish an accurate statistical distribution model.The research content of this paper has some reference and reference significance for the study of the load spectrum of engineering machinery.