System Optimization Design Under Time Variant Reliability Constraints

2011 ◽  
Author(s):  
Xiao-Ling Zhang ◽  
Hong-Zhong Huang ◽  
Zhong-Lai Wang ◽  
Pei-Nan Ge
Keyword(s):  
Probability Theory ◽  
Random Variables ◽  
System Optimization ◽  
Probability Of Failure ◽  
Time Dependent ◽  
Design Parameters ◽  
Total Probability ◽  
Engineering Design Problems ◽  
Reliability Method ◽  
Time Invariant

Due to the degradation, input loading and uncertainty in the design parameters usually involve random variables and random processes, reliability analysis for engineering design problems are usually time dependent. Many problems related to degradation have been treated as monotonic or statistically independent, therefore, the probability of failure only at the end of the lifetime of the structure are considered. To the issues of parameters with stochastic process, the outcrossing rate methods have been extensively developed to calculate the upper bound of time-dependent reliability. In these methods, the issue of proper choice of time interval is crucial and difficult. In this paper, a new method for time dependent reliability optimization based on the total probability theory and universal generating function is proposed. In the proposed method, firstly, Parameters with stochastic processes are discretized into some discrete random variables. Secondly, the discrete parameters are reformed into a new random process by the operation of the universal generating functions. Finally, based on the total probability theory, the probability of failure for each limit state function is analyzed using sequential optimization and time invariant reliability assessment method. Only the time invariant reliability method is needed in the proposed method, by conditioning the continuous random variables on the discrete random parameters. Numerical example is presented to demonstrate the performance of the proposed method.

Reliability-Based Optimization With Discrete and Continuous Decision and Random Variables

2008 ◽  
Vol 130 (6) ◽  
Author(s):  
Mark McDonald ◽  
Sankaran Mahadevan
Keyword(s):  
Discrete Optimization ◽  
System Reliability ◽  
Greedy Algorithms ◽  
Optimization Technique ◽  
Random Variables ◽  
System Level ◽  
Mixed Integer ◽  
Total Probability ◽  
Engineering Design Problems ◽  
Reliability Method

Engineering design problems frequently involve both discrete and continuous random and design variables, and system reliability may depend on the union or intersection of multiple limit states. Solving reliability-based design optimization (RBDO) problems, where some or all of the decision variables must be integer valued, can be expensive since the computational effort increases exponentially with the number of discrete variables in discrete optimization problems, and the presence of both system and component level reliability makes RBDO more expensive. The presence of discrete random variables in a RBDO problem has usually necessitated the use of Monte Carlo simulation or some other type of enumeration procedure, both of which are computationally expensive. In this paper, the theorem of total probability is used to allow for the use of the first-order reliability method in solving mixed-integer RBDO problems. Single-loop RBDO formulations are developed for three classes of mixed-integer RBDO with both discrete and continuous random variables and component and system-level reliability constraints. These problem formulations can be solved with any appropriate discrete optimization technique. This paper develops, for each of the three problem classes, greedy algorithms to find an approximate solution to the mixed-integer RBDO problem with both component and system reliability constraints and/or objectives. These greedy algorithms are based on the solution of a relaxed formulation and require hardly additional computational expense than that required for the solution of the continuous RBDO problem. The greedy algorithms are verified by branch and bound and genetic algorithms. Also, this paper develops three algorithms, which can allow for calibration of reliability estimation with a more accurate reliability analysis technique. These algorithms are illustrated in the context of a truss optimization problem.

Reliability of Buried Pipeline Using a Theory of Probability of Failure

2006 ◽  
Vol 110 ◽  
pp. 221-230 ◽  
Author(s):  
Ouk Sub Lee ◽  
Dong Hyeok Kim ◽  
Seon Soon Choi
Keyword(s):  
Failure Probability ◽  
Probability Of Failure ◽  
Operating Conditions ◽  
Reliability Estimation ◽  
Design Parameters ◽  
Buried Pipeline ◽  
Safety Level ◽  
First Order ◽  
Reliability Method ◽  
Corrosion Defects

The reliability estimation of buried pipeline with corrosion defects is presented. The reliability of corroded pipeline has been estimated by using a theory of probability of failure. And the reliability has been analyzed in accordance with a target safety level. The probability of failure is calculated using the FORM (first order reliability method). The changes in probability of failure corresponding to three corrosion models and eight failure pressure models are systematically investigated in detail. It is highly suggested that the plant designer should select appropriate operating conditions and design parameters and analyze the reliability of buried pipeline with corrosion defects according to the probability of failure and a required target safety level. The normalized margin is defined and estimated accordingly. Furthermore, the normalized margin is used to predict the failure probability using the fitting lines between failure probability and normalized margin.

Time-Dependent Reliability Analysis of Vibratory Systems With Random Parameters

2016 ◽  
Vol 138 (3) ◽  
Author(s):  
Zissimos P. Mourelatos ◽  
Monica Majcher ◽  
Vasileios Geroulas
Keyword(s):  
Large Scale ◽  
Probability Of Failure ◽  
Time Dependent ◽  
Total Probability ◽  
Conditional Probabilities ◽  
Random Parameters ◽  
Random Vibrations ◽  
Input Parameters ◽  
Total Probability Theorem ◽  
Dependent Probability

The field of random vibrations of large-scale systems with millions of degrees-of-freedom (DOF) is of significant importance in many engineering disciplines. In this paper, we propose a method to calculate the time-dependent reliability of linear vibratory systems with random parameters excited by nonstationary Gaussian processes. The approach combines principles of random vibrations, the total probability theorem, and recent advances in time-dependent reliability using an integral equation involving the upcrossing and joint upcrossing rates. A space-filling design, such as optimal symmetric Latin hypercube (OSLH) sampling, is first used to sample the input parameter space. For each design point, the corresponding conditional time-dependent probability of failure is calculated efficiently using random vibrations principles to obtain the statistics of the output process and an efficient numerical estimation of the upcrossing and joint upcrossing rates. A time-dependent metamodel is then created between the input parameters and the output conditional probabilities allowing us to estimate the conditional probabilities for any set of input parameters. The total probability theorem is finally applied to calculate the time-dependent probability of failure. The proposed method is demonstrated using a vibratory beam example.

Surrogate-model-based reliability method for structural systems with dependent truncated random variables

2017 ◽  
Vol 231 (3) ◽  
pp. 265-274 ◽  
Author(s):  
Ning-Cong Xiao ◽  
Libin Duan ◽  
Zhangchun Tang
Keyword(s):  
Surrogate Model ◽  
Failure Modes ◽  
Random Variables ◽  
Probability Of Failure ◽  
Structural Systems ◽  
Multiple Failure Modes ◽  
Model Based ◽  
Reliability Sensitivity ◽  
Reliability Method ◽  
Uncertainty Space

Calculating probability of failure and reliability sensitivity for a structural system with dependent truncated random variables and multiple failure modes efficiently is a challenge mainly due to the complicated features and intersections for the multiple failure modes, as well as the correlated performance functions. In this article, a new surrogate-model-based reliability method is proposed for structural systems with dependent truncated random variables and multiple failure modes. Copula functions are used to model the correlation for truncated random variables. A small size of uniformly distribution samples in the supported intervals is generated to cover the entire uncertainty space fully and properly. An accurate surrogate model is constructed based on the proposed training points and support vector machines to approximate the relationships between the inputs and system responses accurately for almost the entire uncertainty space. The approaches to calculate probability of failure and reliability sensitivity for structural systems with truncated random variables and multiple failure modes based on the constructed surrogate model are derived. The accuracy and efficiency of the proposed method are demonstrated using two numerical examples.

Extreme Value Metamodeling for System Reliability With Time-Dependent Functions

2015 ◽  
Author(s):  
Zhifu Zhu ◽  
Xiaoping Du
Keyword(s):  
Computational Efficiency ◽  
System Reliability ◽  
Extreme Values ◽  
Random Variables ◽  
Probability Of Failure ◽  
High Accuracy ◽  
Extreme Value ◽  
Time Dependent ◽  
System Failure ◽  
Kriging Method

The reliability of a system is usually measured by the probability that the system performs its intended function in a given period of time. Estimating such reliability is a challenging task when the probability of failure is rare and the responses are nonlinear and time variant. The evaluation of the system reliability defined in a period of time requires the extreme values of the responses in the predefined period of time during which the system is supposed to function. This work builds surrogate models for the extreme values of responses with the Kriging method. For the sake of computational efficiency, the method creates Kriging models with high accuracy only in the region that has high contributions to the system failure; training points of random variables and time are sampled simultaneously so that their interactions could be considered automatically. The example of a mechanism system shows the effectiveness of the proposed method.

Reliability-Based Topology Optimization Using Mean-Value Second-Order Saddlepoint Approximation

2018 ◽  
Vol 140 (3) ◽  
Author(s):  
Dimitrios I. Papadimitriou ◽  
Zissimos P. Mourelatos
Keyword(s):  
Topology Optimization ◽  
Limit State ◽  
Random Variables ◽  
Saddlepoint Approximation ◽  
Probability Of Failure ◽  
Mean Value ◽  
Second Order ◽  
State Function ◽  
Sensitivity Derivatives ◽  
Reliability Method

A reliability-based topology optimization (RBTO) approach is presented using a new mean-value second-order saddlepoint approximation (MVSOSA) method to calculate the probability of failure. The topology optimizer uses a discrete adjoint formulation. MVSOSA is based on a second-order Taylor expansion of the limit state function at the mean values of the random variables. The first- and second-order sensitivity derivatives of the limit state cumulant generating function (CGF), with respect to the random variables in MVSOSA, are computed using direct-differentiation of the structural equations. Third-order sensitivity derivatives, including the sensitivities of the saddlepoint, are calculated using the adjoint approach. The accuracy of the proposed MVSOSA reliability method is demonstrated using a nonlinear mathematical example. Comparison with Monte Carlo simulation (MCS) shows that MVSOSA is more accurate than mean-value first-order saddlepoint approximation (MVFOSA) and more accurate than mean-value second-order second-moment (MVSOSM) method. Finally, the proposed RBTO-MVSOSA method for minimizing a compliance-based probability of failure is demonstrated using two two-dimensional beam structures under random loading. The density-based topology optimization based on the solid isotropic material with penalization (SIMP) method is utilized.

Time-Dependent System Reliability Analysis Using Random Field Discretization

2015 ◽  
Vol 137 (10) ◽  
Author(s):  
Zhen Hu ◽  
Sankaran Mahadevan
Keyword(s):  
Random Field ◽  
Reliability Analysis ◽  
System Reliability ◽  
Probability Of Failure ◽  
Two Dimensions ◽  
Time Dependent ◽  
Time Interval ◽  
System Reliability Analysis ◽  
Reliability Method ◽  
Time Dependent System

This paper proposes a novel and efficient methodology for time-dependent system reliability analysis of systems with multiple limit-state functions of random variables, stochastic processes, and time. Since there are correlations and variations between components and over time, the overall system is formulated as a random field with two dimensions: component index and time. To overcome the difficulties in modeling the two-dimensional random field, an equivalent Gaussian random field is constructed based on the probability equivalency between the two random fields. The first-order reliability method (FORM) is employed to obtain important features of the equivalent random field. By generating samples from the equivalent random field, the time-dependent system reliability is estimated from Boolean functions defined according to the system topology. Using one system reliability analysis, the proposed method can get not only the entire time-dependent system probability of failure curve up to a time interval of interest but also two other important outputs, namely, the time-dependent probability of failure of individual components and dominant failure sequences. Three examples featuring series, parallel, and combined systems are used to demonstrate the effectiveness of the proposed method.

Time-Dependent Reliability Analysis by a Sampling Approach to Extreme Values of Stochastic Processes

2012 ◽  
Author(s):  
Zhen Hu ◽  
Xiaoping Du
Keyword(s):  
Stochastic Processes ◽  
Reliability Analysis ◽  
Extreme Values ◽  
Limit State ◽  
Time Dependent ◽  
State Function ◽  
Limit State Function ◽  
Reliability Method ◽  
Time Invariant ◽  
State Functions

Maintaining high accuracy and efficiency is a challenging issue in time-dependent reliability analysis. In this work, an accurate and efficient method is proposed for limit-state functions with the following features: The limit-state function is implicit with respect to time, and its input contains stochastic processes; the stochastic processes include only general strength and stress variables, or the limit-state function is monotonic to these stochastic processes. The new method employs random sampling approaches to estimate the distributions of the extreme values of the stochastic processes. The extreme values are then used to replace the corresponding stochastic processes, and consequently the time-dependent reliability analysis is converted into its time-invariant counterpart. The commonly used time-invariant reliability method, the First Order Reliability Method, is then applied for the time-variant reliability analysis. The results show that the proposed method significantly improves the accuracy and efficiency of time-dependent reliability analysis.

Time-Dependent Reliability Analysis Using the Total Probability Theorem

2015 ◽  
Vol 137 (3) ◽  
Author(s):  
Zissimos P. Mourelatos ◽  
Monica Majcher ◽  
Vijitashwa Pandey ◽  
Igor Baseski
Keyword(s):  
Reliability Analysis ◽  
Limit State ◽  
Random Processes ◽  
Time Dependent ◽  
Normal Space ◽  
Total Probability ◽  
Conditional Probabilities ◽  
Reliability Method ◽  
Standard Normal ◽  
Total Probability Theorem

A new reliability analysis method is proposed for time-dependent problems with explicit in time limit-state functions of input random variables and input random processes using the total probability theorem and the concept of composite limit state. The input random processes are assumed Gaussian. They are expressed in terms of standard normal variables using a spectral decomposition method. The total probability theorem is employed to calculate the time-dependent probability of failure using time-dependent conditional probabilities which are computed accurately and efficiently in the standard normal space using the first-order reliability method (FORM) and a composite limit state of linear instantaneous limit states. If the dimensionality of the total probability theorem integral is small, we can easily calculate it using Gauss quadrature numerical integration. Otherwise, simple Monte Carlo simulation (MCS) or adaptive importance sampling are used based on a Kriging metamodel of the conditional probabilities. An example from the literature on the design of a hydrokinetic turbine blade under time-dependent river flow load demonstrates all developments.

Time-Dependent Reliability Analysis of Vibratory Systems With Random Parameters

2015 ◽  
Author(s):  
Zissimos P. Mourelatos ◽  
Monica Majcher ◽  
Vasileios Geroulas
Keyword(s):  
Large Scale ◽  
Probability Of Failure ◽  
Time Dependent ◽  
Total Probability ◽  
Conditional Probabilities ◽  
Random Parameters ◽  
Random Vibrations ◽  
Input Parameters ◽  
Total Probability Theorem ◽  
Dependent Probability

The field of random vibrations of large-scale systems with millions of degrees of freedom is of significant importance in many engineering disciplines. In this paper, we propose a method to calculate the time-dependent reliability of linear vibratory systems with random parameters excited by non-stationary Gaussian processes. The approach combines principles of random vibrations, the total probability theorem and recent advances in time-dependent reliability using an integral equation involving the up-crossing and joint up-crossing rates. A space-filling design, such as optimal symmetric Latin hypercube sampling, is first used to sample the input parameter space. For each design point, the corresponding conditional time-dependent probability of failure is calculated efficiently using random vibrations principles to obtain the statistics of the output process and an efficient numerical estimation of the up-crossing and joint up-crossing rates. A time-dependent metamodel is then created between the input parameters and the output conditional probabilities allowing us to estimate the conditional probabilities for any set of input parameters. The total probability theorem is finally applied to calculate the time-dependent probability of failure. The proposed method is demonstrated using a vibratory beam example.

Sign in / Sign up

Export Citation Format

Share Document