Papers by Keyword: Random Variable

Abstract: With all parameters being seemed as random variables with Gaussian distribution,and Latin Hypercube Sampling method being used, the strength of ribbed cylindrical shell is computed based finite element and Monte Carlo Method. By probability sensitivity diagram, scatter diagram and related coefficient, the sensitivity of random variable of ribbed cylindrical shell is analyzed based on the method of random sensitivity analysis. The results show that shell thickness, external pressure, frame spacing and internal radius are more sensitivity than the other variables, so these variables can be treated as sensitivity factors.

Abstract: The probabilistic methods are used in engineering tasks where a computational model contains random variables. The new method the Direct Optimised Probabilistic Calculation (DOProC) - which is being developed now seems to be highly efficient in terms of calculation time and accuracy of solution. The computation is purely numerical and does not use any simulation techniques. The algorithm has been implemented in several software applications which have been used in probabilistic tasks and probabilistic reliability assessments.

Abstract: In this paper, the reliability design method was used during the process of mechanical parts designing, the design parameters was regarded as random variables, the reliability index of products was involved in the design of the entire process, making the design more in line with the results of the actual situation. The general process of the reliability design was explained by the design of typical part. A good method to effectively improve product quality, reduce part size, save materials and reduce the cost was founded, and the same time the method can make the design of the products more competitive in the market.

Abstract: The paper is focuses on one of probabilistic methods which can be used for failure analysis and reliability assessment of steel structures which are subject to cyclic loads and exposed to fatigue. A particular attention is paid to creation and propagation of fatigue cracks from edges and surface. On the basis of the reliability assessment, a system of inspections is proposed for structural details which tend to be sensitive to fatigue damage. A new probabilistic method which is still under development - Direct Optimized Probabilistic Calculation (DOProC) was used for this probabilistic task. This method is the basis of the FCProbCalc code.

Abstract: This paper takes each weight as a normal distribution random variable. The values of weight are given by several experts, which are used to determine the distribution of this weight random variable. Taking the weight random variables as parameters of the decision models, we can calculate the distribution of result. This method evaluates the alternatives by their expectations and variances of the calculation results. The evaluation result coincides with people's understanding about the problem of evaluation. The method of this paper is the improvement of power exponent method and weight sum method.

Abstract: There is a great deal of uncertainty in spare parts inventory support process of aircraft hydraulic system, which cause its support decision difficult. Built integrated support decision model of the spare parts inventory based on sequence operation theory. Regard the quantity of the spare parts demand, maintenance and store as one-dimensional discrete random variable, analyzing the mutual effect among the random events utilizing addition type convolution, subtraction type convolution, and type product, thus give a dynamic description of the spare parts inventory integrated support process. Provide optimal decision basis for the spare parts inventory integrated support of aircraft hydraulic system. Results of the experimental indicated that the result of the support decision modeling consistent with the actual demand and the support decision model proved to be validated.

Abstract: The Stochastic Finite Element Method (SFEM) represents a new approach to solve mechanical systems with stochastic characteristics. The SFEM is based on the deterministic Finite Element Method in which some variables related to the structural state (variables involved in the stiffness matrix) and to the applied actions (involved in the load vector) are uncertain. In other words, the SFEM tries to look for the stochastic properties of the mechanical response. The solution of a stochastic mechanical system is completely defined through the evaluation of the probability density function of the response process. This cannot be analytically achieved through most of the available methods and techniques such as Fokker-Planck equation, Wiener-Hermite expansion, perturbation methods, stochastic linearization, WHEP technique, decomposition method and stochastic finite element methods [1; 2]. Some exact solutions are available for the mean and standard deviation, not for the Probability Density Function (pdf), of the solution process. In this paper, we develop the theory of method named the Stochastic Transformation Method (STM) in order to determine the pdf of the response of a stochastic mechanical system with random excitation and/or stiffness. The Stochastic Transformation Method is based on one-to-one mapping between the random output(s) and input(s) where the transformation Jacobean J can be computed. The pdf of the output(s) is then computed through the known joint pdf of the inputs multiplied by the determinant of transformation Jacobean matrix. The one-to-one mapping condition can be relaxed through some mathematical tricks. This STM allows us to express the “exact” pdf of the mechanical response [3], provided that the transformation Jacobean can be defined. For many cases, the pdf of the response can be obtained in a closed-form in terms of the joint distribution of the input random variables.

Abstract: In this paper variance of preformance function is calculated according to law of covariance propagation and matrix algorithm is used to calculate reliability index of structures in generalized random space, which is applicable to independent and correlative random variable, linear or nonlinear limit state equation. The given example has shown that the proposed method is very simple and convenient.

Abstract: Considering the up-down deflection of vehicle frame as a random variable when a car runs unsteadily, the alternating stress cased by deflection is random. The probability density function is determined by date of the alternating stress acting on the frame. Using the concepts of upper and lower approximation in the rough set theory and confidence degree, random parameters of vehicle frame can be transferred into certain interval values. When the confidence degree took a value from 0 to 1, a series of intervals between the upper and lower approximation are determined respectively. With the algorithm of the interval operation, a method of fatigue analysis is put forward.

Abstract: The mixed distribution model and the maximum entropy model are used to represent service load of the vehicle clutch. The parameters of those models are estimated with different methods. The findings indicate that maximum entropy distribution can accurately describe different statistical features of random variables as minimally prejudiced probability distribution if order of the distribution function is properly selected, and that the mixed Weibull distribution shows super performance of the complicated statistical model expression. The parameters of those models are estimated by optimization based on non-linear least squares.