软件可靠性预测的核函数方法

通过合理的假设和抽象,把软件可靠性预测问题转化成从高维空间向低维空间的非线性映射,采用核函数回归估计理论对软件失效时间数据之间的对应关系进行建模,在新建立的模型中,随着软件失效不断发生,模型参数将不断自动调整以适应失效过程的动态变化,从而实现软件可靠性的自适应预测,并对模型中核函数以及回归估计算法的选择进行了研究。最后,对14个软件失效数据集进行了实验分析,并对新建模型与部分其它模型的预测能力和适用能力进行了比较。     

Gompertz software reliability model: Estimation algorithm and empirical validation

Gompertz curve has been used to estimate the number of residual faults in testing phases of software development, especially by Japanese software development companies. Since the Gompertz curve is a deterministic function, the curve cannot be applied to estimating software reliability which is the probability that software system does not fail in a prefixed time period. In this article, we propose a stochastic model called the Gompertz software reliability model based on non-homogeneous Poisson processes. The proposed model can be derived from the statistical theory of extreme-value, and has a similar asymptotic property to the deterministic Gompertz curve. Also, we develop an EM algorithm to determine the model parameters effectively. In numerical examples with software failure data observed in real software development projects, we evaluate performance of the Gompertz software reliability model in terms of reliability assessment and failure prediction.     

Optimal test distributions for software failure cost estimation

We generalize the input domain based software reliability measures by E.C. Nelson (1973) and by S.N. Weiss and E.J. Weyuker (1988), introducing expected failure costs under the operational distribution as a measure for software unreliability. This approach incorporates in the reliability concept a distinction between different degrees of failure severity. It is shown how to estimate the proposed quantity by means of random testing, using the Importance Sampling technique from Rare Event Simulation. A test input distribution that yields an unbiased estimator with minimum variance is determined. The practical application of the presented method is outlined, and a detailed numerical example is given     

Reliability evaluation and component importance measure for manufacturing systems based on failure losses

Little work has been done to assess the reliability of a vital system like the manufacturing system. In this article, a novel and effective system reliability evaluation method in terms of failure losses has been proposed for manufacturing systems of job shop type, and then the failure losses based component importance measure (CIM) is used for importance analysis of equipment. The former indicates the present system reliability situation and the latter points the way to reliability improvement efforts. In this scheme, the problem is described and modeled by a dynamic directed network. Consider that the actual processing time of machines is to contribute to failure occurrence, it is used to calculate the failure times and failure losses. The obtained total failure times and failure losses of the system are applied to evaluate its reliability. Techniques to estimate two kinds of failure losses based CIMs are presented. They offer guidelines to realize system reliability growth cost-effectively. A case study of a real job shop is provided as an example to demonstrate the validity of the proposed methods. Comparison to other commonly used methods shows the efficiency of the proposed methods.     

Doubly weighted moving least squares and its application to structural reliability analysis

In this paper, we proposed a two-stage hybrid reliability analysis framework based on the surrogate model, which combines the first-order reliability method and Monte Carlo simulation with a doubly-weighted moving least squares (DWMLS) method. The first stage consists of constructing a surrogate model based on DWMLS. The weight system of DWMLS considers not only the normal weight factor of moving least squares, but also the distance from the most probable failure point (MPFP), which accounts for reliability problems. An adaptive experimental design scheme is proposed, during which the MPFP is progressively updated. The approximate values and sensitivity information of DWMLS are chosen to determine the number and location of the experimental design points in the next iteration, until a convergence criterion is satisfied. In the second stage, MCS on the surrogate model is then used to calculate the probability of failure. The proposed method is applied to five benchmark examples to validate its accuracy and efficiency. Results show that the proposed surrogate model with DWMLS can estimate the failure probability accurately, while requiring fewer original model simulations.     

Time-variant reliability-based design optimization using sequential kriging modeling

This paper presents a sequential Kriging modeling approach (SKM) for time-variant reliability-based design optimization (tRBDO) involving stochastic processes. To handle the temporal uncertainty, time-variant limit state functions are transformed into time-independent domain by converting the stochastic processes and time parameter to random variables. Kriging surrogate models are then built and enhanced by a design-driven adaptive sampling scheme to accurately identify potential instantaneous failure events. By generating random realizations of stochastic processes, the time-variant probability of failure is evaluated by the surrogate models in Monte Carlo simulation (MCS). In tRBDO, the first-order score function is employed to estimate the sensitivity of time-variant reliability with respect to design variables. Three case studies are introduced to demonstrate the efficiency and accuracy of the proposed approach.     

Software architecture reliability analysis using failure scenarios

With the increasing size and complexity of software in embedded systems, software has now become a primary threat for the reliability. Several mature conventional reliability engineering techniques exist in literature but traditionally these have primarily addressed failures in hardware components and usually assume the availability of a running system. Software architecture analysis methods aim to analyze the quality of software-intensive system early at the software architecture design level and before a system is implemented. We propose a Software Architecture Reliability Analysis Approach (SARAH) that benefits from mature reliability engineering techniques and scenario-based software architecture analysis to provide an early software reliability analysis at the architecture design level. SARAH defines the notion of failure scenario model that is based on the Failure Modes and Effects Analysis method (FMEA) in the reliability engineering domain. The failure scenario model is applied to represent so-called failure scenarios that are utilized to derive fault tree sets (FTS). Fault tree sets are utilized to provide a severity analysis for the overall software architecture and the individual architectural elements. Despite conventional reliability analysis techniques which prioritize failures based on criteria such as safety concerns, in SARAH failure scenarios are prioritized based on severity from the end-user perspective. SARAH results in a failure analysis report that can be utilized to identify architectural tactics for improving the reliability of the software architecture. The approach is illustrated using an industrial case for analyzing reliability of the software architecture of the next release of a Digital TV.     

Integrating subset simulation with probabilistic re-analysis to estimate reliability of dynamic systems

It is often expensive to estimate the failure probability of highly reliable systems by Monte Carlo simulation. Subset Simulation breaks the original problem of estimating a small probability into the estimation of a sequence of large conditional probabilities, which is more efficient. The conditional probabilities are estimated by Markov Chain simulation. Uncertainty in the power spectral density of the excitation makes it necessary to re-evaluate the reliability for many power spectral densities that are consistent with the evidence about the system excitation. Subset Simulation is more efficient than Monte Carlo simulation, but still requires a new simulation for each admissible power spectral density. This paper presents an efficient method to re-evaluate the reliability of a dynamic system under stationary Gaussian stochastic excitation for different load spectra. We accomplish that by re-weighting the results of a single Subset Simulation. This method is applicable to both linear and nonlinear systems provided that all of the spectra contain the same amount of energy. The authors are currently working on an extension of the method to nonlinear systems, even when the sampling and true power spectral density functions contain different amounts of energy.     

An efficient evidence-based reliability analysis method via piecewise hyperplane approximation of limit state function

For evidence-based reliability analysis, whether a focal element belongs to the failure domain is commonly judged by the corresponding extreme values of a performance function in its response domain. In contrast, in this paper, an efficient method by which the ownership relationship between a focal element and the failure domain is directly determined in uncertain variable domain, is proposed via the piecewise hyperplane approximation of limit state function (LSF). The whole uncertainty domain is divided into several sub uncertainty domains on the defined reference direction. The approximate LSF is constructed by the piecewise hyperplane in each sub uncertainty domain, the belief measure and the plausibility measure of reliability analysis can be directly calculated in uncertainty domain through the approximate piecewise hyperplanes of LSF. The proposed evidence-based reliability analysis method is demonstrated by two numerical examples and two engineering applications.