Wanqing Song

The reliability analysis of integrated circuit (IC) isolation structures (ISs) faces two challenges: how to accurately describe the distribution of time-to-failure (TTF) data, and how to obtain the TTF distribution under actual working conditions (AWCs). To address these issues, this paper introduces the flexible five-parameter Beta Generalized Weibull (BGW) distribution, which can represent the five characteristic failure-rate shapes (IFR, DFR, CFR, BFR and UBT). The sign of the hazard derivative is analyzed through Glaser’s eta-function theorem together with the asymptotic behavior of the hazard rate at the origin and at infinity; this yields exact, necessary-and-sufficient parameter conditions for each shape, with the origin controlled by the product aγω and the tail controlled by γ. A TTF model based on the BGW distribution is then proposed to predict the TTF distribution under AWCs, and it is proved analytically that a change of stress affects only the scale parameter δ, leaving all shape parameters unchanged. Because this invariance was previously stated as an assumption, it is verified empirically on the time-dependent dielectric breakdown (TDDB) data using per-level Weibull shape estimates with bootstrap confidence intervals, a Kruskal–Wallis test on standardized failure times, and a parametric-bootstrap likelihood-ratio test within the BGW family; all three tests support stress-invariant shapes. The reliability function is recast as a convergent infinite series, with the correct beta-integral kernel and an explicit ratio-test bound. Finally, the effectiveness and superiority of the method are verified using TDDB TTF data, and the untested 3800 V condition is predicted from the fitted linear voltage-acceleration model.

Since the degradation process of IGBTs is a slowly varying nonstationary stochastic process with random switching among three modes (active region, saturation region, and cut-off region), remaining useful life (RUL) prediction of IGBTs remains an urgent research issue. This paper investigates IGBT RUL prediction under DC voltage-controlled operating conditions. First, a health feature parameter (HFP) is constructed via the weighted linear combination of gate voltage, collector voltage, and temperature for fault feature extraction of degradation processes. Second, the Markov state transition matrix describing the three operating modes of IGBTs is established. Furthermore, a long-range dependent fractional Weibull process (fWp) prognostic model with three-mode characteristics is developed. Finally, validation is carried out on the public NASA IGBT thermal overload aging dataset. Error analysis demonstrates that the proposed method achieves superior prediction accuracy compared with four prognostic approaches: APE, OSGP, TDNN-GP, and SA-NN.

The reliability of the power supply is intrinsically linked to the robustness of the distribution network. Constructing a distribution network model is highly challenging due to its complex structure, the large number of equipment and users, and the randomness of load variations and equipment failures. Knowledge graphs, a structured data representation describing entities and their relationships, can address challenges by modelling various devices, users and real-time operation data as entities in a graphical manner. As a result, the constructed large network graphs are not only easy to interpret, but also efficient in data management and reliability analysis. The results are promising and validate our hypothesis.

Aiming at the non-stationary and slowly varying stochastic nature of bearing degradation from normal operation to failure, this paper proposes a fratcional Generalized Cauchy (fGC) prediction model with 1/f process and dual parameters: fractal dimension and Hurst exponent. First, 1/f process sequences exhibit long-range dependence and power-law characteristics. Next the fGC degradation model is established, and the Hurst exponent and fractal dimension are calculated using the R/S method and box-counting dimension method, respectively. Then a dimensionless jump descriptor is employed as a Health Indicator to detect incipient faults and estimate degradation parameters. The maximum likelihood algorithm method is applied to parameter estimation. Finally, a experiment verifies the satisfactory prediction performance through compared with CNN and LSTM predicting model.

IGBTs often operate under stochastic external conditions, and their degradation process is slow, progressive, non-stationary, and characterized by long-range dependence. In particular, an IGBT may switch randomly among three operating modes: active mode, saturation mode, and cut-off mode. Consequently, predicting the remaining useful life becomes highly challenging. In this study, a complex health indicator is first proposed as a failure parameter to identify fault features, constructed through a weighted linear combination of the gate voltage, collector voltage and temperature. Next, a fractional Weibull process model incorporating long-range dependence is developed, which accounts for the three operating modes corresponding to the cut-off, saturation, and active regions of the IGBT. Finally, the experimental dataset is obtained from the publicly available thermal overload aging data of power IGBTs provided by the NASA Research Center. The results show that the proposed approach achieves higher prognosis accuracy compared to four other prognostic optimization strategies, as confirmed through error analysis.

In this study we develop a novel approach for improvement of the precision of numerical integration by using equal tangent slopes in order to partition the integration interval. As a result, two new composite numerical integration formulas are derived. Using continuous function properties, it is possible to analyze the error and convergence of these formulas. A practical example confirms the effectiveness of this method in improving the accuracy of numerical integration. Additionally, the paper delves into the convergence properties and the error reduction capabilities of the new formulas compared to the traditional trapezoidal and Simpson’s rules. The limitations of the developed approach are briefly discussed.

IGBTs often operate under stochastic external conditions, and their degradation process is slow, progressive, non-stationary, and characterized by long-range dependence. In particular, an IGBT may switch randomly among three operating modes: active mode, saturation mode, and cut-off mode. Consequently, predicting the remaining useful life becomes highly challenging. In this study, a complex health indicator is first proposed as a failure parameter to identify fault features, constructed through a weighted linear combination of the gate voltage, collector voltage and temperature. Next, a fractional Weibull process model incorporating long-range dependence is developed, which accounts for the three operating modes corresponding to the cut-off, saturation, and active regions of the IGBT. Finally, the experimental dataset is obtained from the publicly available thermal overload aging data of power IGBTs provided by the NASA Research Center. The results show that the proposed approach achieves higher prognosis accuracy compared to four other prognostic optimization strategies, as confirmed through error analysis.

Estimation of remaining useful life (RUL) of roller bearings is a critical task for predictive maintenance in manufacturing. Bearings are subject to a variety of complex factors during operation, making their degradation process nonlinear with non-Gaussian characteristics. This issue represents a challenge for estimation of the RUL. However, slow degradation process takes on long-range dependence (LRD) and self-similarity. In terms of Pareto’s probability density function (PDF), we proposed the Generalized Pareto Motion (GPm) and its fractional-order extension (fGPm) as degradation predictive models. Estimation of the Hurst exponent shows that this model has long correlation and self-similarity. Through the analysis of the uncertainty of the end point of the bearing’s RUL and the prediction process, not only verified the high adaptability of fGPm in simulating complex degradation process, but also the criteria for judging self-similarity and LRD characteristics were established. One case study demonstrated the advantages of this method, providing an effective analytical tool for a deeper understanding of the degradation mechanism of bearings.

Estimation of remaining useful life (RUL) of roller bearings is a prevalent problem for predictive maintenance in manufacturing. However, roller bearings are subject to a variety of factors during their operation. As a result, we deal with a slow nonlinear degradation process, which is long-range dependent, self-similar and has non-Gaussian characteristics. Proper data pre-processing enables us to use Pareto’s probability density function (PDF), Generalized Pareto motion (GPm) and its fractional-order extension (fGPm) as the degradation predictive model. Estimation of the Hurst exponent shows that this model has a long-range correlation and self-similarity. Through the analysis of the uncertainty of the end point of the bearing’s RUL and the prediction process, not only did it verify the high adaptability of fGPm in simulating complex degradation processes but also the criteria for judging self-similarity, and LRD characteristics were established. The case study mainly proves the validity of the theory, providing an effective analytical tool for a deeper understanding of the degradation mechanism.