This paper presents a reliability modeling and analysis framework for load-sharing systems with identical components subject to continuous degradation. It is assumed that the in system suffer from degradation through an additive impact under increased workload caused by consecutive failures. A log-linear link function used describe relationship between rate load stress levels. By assuming component well modeled step-wise drifted Wiener process, we construct maximum likelihood estimates...

The transient stabilization of virtual synchronous generator-voltage source converter (VSG-VSC) system considering the current limitation is analyzed, which widely used in grid integration renewable energy. mechanism and adverse effect are revealed through qualitative quantitative analysis. Further, low-pass filter (LPF)-based power angle feedback control proposed. improvement stability elimination verified by phase trajectory method based on a nonlinear state space model during fault...

Abstract Anaerobic sludge digestion is the main technology for reduction and stabilization prior to disposal. Nevertheless, methane production from anaerobic of waste activated (WAS) often restricted by poor biochemical potential slow hydrolysis rate WAS. This work systematically investigated effect PHA levels WAS on production, using both experimental mathematical modeling approaches. Biochemical tests showed that increased with in Model-based analysis suggested PHA-based method enhanced...

In this paper, we develop a maintenance model for systems subjected to multiple correlated degradation processes, where multivariate stochastic process is used the and covariance matrix employed describe interactions among processes. The system considered failed when any of its features hits pre-specified threshold. Due dormancy degradation-based failures, inspection implemented detect hidden failures. are replaced upon inspection. We assume an imperfect inspection, in such way that failure...

Summary This paper presents an online data‐driven composite adaptive backstepping control for a class of parametric strict‐feedback nonlinear systems with mismatched uncertainties, where both tracking errors and prediction are utilized to update estimates. Hybrid exact differentiators applied obtain the derivatives virtual inputs such that complexity problem integrator can be avoided. Closed‐loop error equations integrated in moving‐time window generate recorded data improve parameter...

In Federated Learning (FL), two-way model exchanges are required between the server and workers every training round. Due to large size of machine learning models, communications them lead high delay economic cost. At present, communication-efficient FL methods, for examples, top-k sparsification quantization, taking advantages sparseness gradients fact that gradient-based updating can tolerance small deviations, effectively reduce communication cost single However, these schemes cannot be...

Given the multivariable coupling, strong nonlinearity and time-varying features in wastewater treatment processes, adaptive strategies, including just-in-time learning (JITL), time difference (TD), moving window (MW) methods have been chosen this paper to enhance multi-output soft-sensor models ensure online prediction for a variety of hard-to-measure variables simultaneously. In proposed soft-sensors, partial least squares (MPLS), relevant vector machine (MRVM) Gaussian process regression...

Process monitoring of wastewater treatment plant (WWTP) is a challenging industrial problem, due to its exposure the hostile working environment and significant disturbances. This paper proposed novel fault diagnosis method, termed as optimization forecast components-support vector machine (OFC-SVM). The method firstly improved forecastable component analysis (ForeCA) for feature extraction. Secondly, in order further enhance quadratic Grid Search (GS) algorithm utilized optimize parameters...

This paper investigates a condition-based maintenance policy for systems subject to non-homogeneous degradation process. A nonhomogeneous process occurs as result of deterioration nature and the environmental effect. In first step, it develops two models, which consider constant inspection interval non-periodic interval. The then optimizes with monotone preventive replacement thresholds. optimal decision is shown control-limit policy, where threshold monotonically decreasing system age. An...

Stereo matching is a challenging problem in the field of computer vision and has recently received extensive attention. However, traditional methods are labor intensive premised on specific conditions. In this paper, we propose robust stereo cost algorithm that relies refined features extracted by stack auto-encoder. These for different types image pairs, which significantly improve generality proposed algorithm. addition, smoothed belief volume with guided filter to performance propagation...

We propose that ultrasonic pretreatment could significantly improve the degradation of anaerobically digested sludge with economic favorability in post aerobic digestion.

Multivariate statistical methods have gained significant popularity in past decades. However, process dynamics and insufficient training data usually result degradation or even failure of a trained model. To deal with these problems, this paper proposes novel monitoring method, called cross-spatiotemporal adaptive boosting transfer learning (CS-AdBoostTrLM). Different from the standard methods, CS-AdBoostTrLM has following advantages: first, source domain (SD) data, which are discarded by...

Due to the closed-loop control strategy, current sensor faults in inverter systems are very destructive. Aiming at fault estimation and tolerance of sensors, this paper proposes a reduced-order observer-based method for grid-connected three-level neutral point clamped(NPC) inverter. Firstly, model is established, augmented system inverte proposed, state vector composed original system, signal disturbance signal. Then, matrix transformation used decouple from state. observer designed...

Abstract This article investigates robust multiple fault detection (FD) and estimation (FE) schemes, based on adaptive finite‐time observer (AFTO), applied to nonlinear systems with disturbances, actuator faults, sensor faults. The AFTO is designed, a parameter law constructed continuously optimize the residual performance function improve robustness. schemes are able simultaneously detect estimate unknown states, AFTO's design, convergence stability conditions guaranteed in terms of linear...