Stationary Probability Research Articles

Probabilistic curvature limit states of corroded circular RC bridge columns: Data-driven models and application to lifetime seismic fragility analyses

Reinforcement corrosion has been recognized as an influential factor in the seismic fragility, both demand and capacity models, of aging reinforced concrete (RC) bridges. For capacity models, accurate and applied prediction tools accounting for aging effects are yet to be well established. Current practices usually perform numerical analyses to obtain time-variant capacity models, which are time-consuming particularly when multi-source structural and environmental uncertainties are considered and sometimes even suffer computational non-convergence. To address these issues, this study leverages a rigorously optimized artificial neural network architecture to develop data-driven models for rapid estimates of probabilistic curvature capacity of corroded circular RC bridge columns with flexural failure modes. An extensive database of multi-level curvature limit states (i.e. slight, moderate, severe, and complete) is created through experimentally validated moment–curvature analyses. A new threshold for the moderate limit state is defined based on the strain of core concrete, rather than cover concrete, to account for the potential full erosion of the cover with drastic corrosion. The data-driven probabilistic capacity models are applied to aid the lifetime seismic fragility assessment of a typical highway bridge, where the spectral acceleration at 1.0 s (Sa-10), peak ground velocity (PGV), and Housner intensity (HI) are found consistently, for the first time, as optimal intensity measures for probabilistic demand modeling of RC columns with different extents of aging effects. For the ease of application, the database and code for the data-driven probabilistic capacity models are accessible at https://bit.ly/3uAa8EY .

Probabilistic ductile deformation limit state prediction of monolithic exterior shear keys based on quantile regression machine learning techniques

Shear strength and ductile deformation capacity of monolithic exterior shear keys play important roles on the seismic behavior of highway bridges. Improving the ductility of the shear key through reasonable design could effectively restrict superstructure displacement and minimize damage to pier columns. However, most existing studies primarily focus on evaluating the shear strength and damage mechanisms of shear keys, with limited attention given to analytical models for assessing ductile deformation capacity. To this end, this study proposes a method that combines quantile regression and machine learning techniques to predict the ductile deformation limit states of shear keys. The proposed numerical modeling method for shear keys is validated using the experimental data on lateral load-displacement curves. On this basis, Latin hypercubic sampling and joint simulation are utilized to establish the ductile deformation limit state database. The ductile deformation limit states of shear keys are characterized by their lateral displacements in three distinct damage states. The main influencing factors of each limit states are investigated by correlation analysis. Three quantile regression machine learning models are trained and constructed to conduct point and interval prediction for the ductile deformation limit states. All three models, particularly the Gaussian process quantile regression model, are proven to demonstrate exceptional prediction accuracy and uncertainty quantification. Experimental data validation confirms that these models outperform the prediction accuracy of existing simplified models.

LN: A Flexible Algorithmic Framework for Layered Queueing Network Analysis

Layered queueing networks (LQNs) are an extension of ordinary queueing networks useful to model simultaneous resource possession and stochastic call graphs in distributed systems. Existing computational algorithms for LQNs have primarily focused on mean-value analysis. However, other solution paradigms, such as normalizing constant analysis and mean-field approximation, can improve the computation of LQN mean and transient performance metrics, state probabilities, and response time distributions. Motivated by this observation, we propose the first LQN meta-solver, called LN, that allows for the dynamic selection of the performance analysis paradigm to be iteratively applied to the submodels arising from layer decomposition. We report experiments where this added flexibility helps us to reduce the LQN solution errors. We also demonstrate that the meta-solver approach eases the integration of LQNs with other formalisms, such as caching models, enabling the analysis of more general classes of layered stochastic networks. Additionally, to support the accurate evaluation of the LQN submodels, we develop novel algorithms for homogeneous queueing networks consisting of an infinite server node and a set of identical queueing stations. In particular, we propose an exact method of moment algorithms, integration techniques for normalizing constants, and a fast non-iterative mean-value analysis technique.

Minimum magnitude boundaries in probabilistic seismic hazard analysis: an insight from structural engineering

In order to systematically advance our understanding of the minimum magnitude limit (Mmin) in the probabilistic seismic hazard analysis (PSHA) calculations, a novel and useful approach utilising a broad range of Single-Degree-of-Freedom oscillators and hazard conditions is being developed and tested. We have determined the most reasonable Mmin value for a variety of structures by examining the impact of Mmin on the mean annual frequency (MAF) of various limit states (LSs) (including the collapse capacity). The originality of the suggested methodology in the current work, known as the MAF saturation strategy, is the recommended Mmin, which is the cut-off value at which lesser magnitude events do add to the hazard but do not significantly change the MAF. The current work is the first to offer the MAF saturation strategy methodology, which searches for the cut-off magnitude at which the MAF value essentially remains constant even when smaller values of this cut-off are utilised as Mmin for hazard assessments. Therefore, given a series of carefully chosen ground motions in each oscillator instance, an incremental dynamic analysis is carried out (by applying the Hunt and Fill algorithm), and the appropriate LS (including the collapse capacity defined as global instability) points are calculated. Thus, the relationship between the distribution of LSs and the Engineering Demand Parameter and intensity measure is found. A simple point source hazard curve is convoluted with this distribution, yielding the structure-specific MAF. In order to find the cut-off lower magnitude (Mmin), this convolution is repeated for several Mmin values. This cut-off is defined as the point at which, when lower values are utilised as Mmin in the PSHA computation, the MAF’s values do not change considerably (with a five per cent threshold). The acquired data were thoroughly discussed in relation to various structural features and seismic input factors. The primary findings showed that each of the structures under consideration requires a Mmin value in the range of 4–4.3. Put otherwise, the suggestions seen in technical literature, which range from 4.5 to 5, are not cautious, at least not when it comes to probabilistic structural limit state frequency. The derived Mmin value is mostly controlled by the natural period of the structure and is largely unaffected by other structural characteristics like ductility, damping ratio and overstrength factor.

Twinning quality sensors in wastewater treatment process via optimized echo state network-based soft sensors

The presence of a large amount of quality-related but hard-to-measure variables usually makes effective monitoring of industrial processes difficult, and even impossible. Soft computing techniques and digital twins can revolutionize standard approaches to solve this issue, even though industrial data are frustrated by nonlinearity, disturbances, and dynamic behaviors. In this light, this paper proposes a data-driven soft sensor with the help of the Fourier amplitude sensitivity test (FAST) and improved probabilistic regularized echo state network (IPRESN), then to twin the quality hardware sensors. Within this framework, the initial step is to rely on the FAST method to measure the collaborated importance of multiple process variables with respect to each quality variable. This quantification facilitates the selection of auxiliary variables for the training of IPRESN. Furthermore, a probabilistic regularization method is learned to improve the robust performance of the echo state network (ESN) by re-designing a new objective function to minimize the mean and variance of the modeling error. Additionally, the whale optimization algorithm (WOA) is improved and then used to optimize the critical hyper-parameters of ESN, thus mitigating the risks associated with suboptimal hyperparameter selection. The effectiveness of the proposed method is verified through a Benchmark Simulation Model 2 (BSM2) and a full-scale wastewater treatment plant data set. The findings show the potential of the proposed method in facilitating digital twin implementations for hardware sensors.

Approximate averaged derivative moments and stochastic response of high-DOF quasi-non-integrable Hamiltonian systems under Poisson white noises

Stochastic response prediction is one of the principal topics in diverse branches of engineering, such as aerospace engineering, mechanical engineering, and vehicle engineering. However, the challenge of addressing high-dimensional partial differential equations or evaluating high-dimensional domain integrals in determining averaged derivative moments has limited the research. This study proposes a method to predict the stochastic response of high-degree-of-freedom (DOF) quasi-non-integrable Hamiltonian systems under Poisson white noises. The original finite DOF system is reduced to a one-dimensional one governing the Hamiltonian by applying the stochastic averaging method, and the associated reduced generalized Fokker–Planck–Kolmogorov (GFPK) equation is derived. Then, the two-step generalized elliptical coordinate transformation is extended to evaluate averaged derivative moments that are the key to obtaining the stationary probability density function (SPDF) for probabilistic information prediction. Moreover, the perturbation method is introduced to address the high-order GFPK equation. Finally, an 8-DOF strongly nonlinear system under purely additive Poisson white noises is given as an example to highlight the accuracy of the proposed method. Results from the proposed method show good agreement with those from Monte Carlo simulations, which indicates that the proposed method can be applied to assess the response of high-DOF quasi-non-integrable Hamiltonian systems under Poisson white noises. In addition, if the potential energy of an arbitrary finite n-DOF system is the same form as that in Appendix B, pertinent high-dimensional domain integrals are converted into explicit expressions when n is even.

Information flow and catalytic dyad in SARS-CoV2 main protease Mpro enzyme using embedded discrete Markov chains and centrality measures

Abstract In this work, we use a network representation of the globular crystalline structure of a given protein as a graph structure with N nodes and E edges in order to analyze quantitatively the flow information and to identify key sites within the globular structure. Each node nj represents a $ C_{\alpha}^{i}$ carbon of the main backbone whereas the node’s degree ki is a measure of its physical interactions. In order to identify biologically relevant and active nodes, we compute local per residue closeness $ C_{c}(i)$, betweenness $ C_{b}(i)$ and eigenvector centralities $ C_{e}(i)$. Further analysis is done by embedding a stochastic dynamic discrete Markov chain in order to evaluate the dynamics of a set of normal random walkers (NRW’s) within the network. From this, we compute the mean first passage time matrix M and the stationary occupation probability vector ψi for each node. These two measures provide very useful information on the dynamical process embed within the $ C_{\alpha}$ network. We apply this to SARS-CoV2 Mpro main protease which is a key enzyme in the virus replication cycle. In particular, we focus our attention to the properties of the catalytic dyad integrated by His41-Cys145 in Mpro main protease since this active site has been under intense scrutiny as a pharmaceutical target. In addition, our results show the existence of additional relevant aminoacids that might play a fundamental role on signal propagation and allosteric pathways in SARS-CoV main proteases.

Graph spatial sampling

We develop lagged Metropolis–Hastings walk for sampling from simple undirected graphs according to given stationary sampling probabilities. It is explained how the technique can be applied together with designed graphs for sampling of units‐in‐space. Compared with the existing spatial sampling methods, which chiefly focus on the sample spatial balance regardless of the associated outcomes of interest, the proposed graph spatial sampling method can considerably improve the efficiency because the graph can be designed to take into account the anticipated spatial distribution of the outcome of interest.

Stationary distribution and probability density function of a stochastic COVID-19 infections model with general incidence

With the sudden outbreak of COVID-19 all over the world, vaccination was one of the most effective treatments. In this paper, a stochastic COVID-19 model with vaccination is developed. We first prove that the stochastic model has a unique global solution by constructing the corresponding Lyapunov function. Then, sufficient condition for ergodic stationary distribution υ(⋅) is provided under the condition R0s>1. Moreover, by applying the six-dimensional Fokker–Planck equation, we obtain the explicit expression of the probability density function near quasi-local equilibrium Q∗, which has the form of lognormal density function. Finally, numerical simulations are presented to verify the above conclusions.

The coalescent in finite populations with arbitrary, fixed structure

The coalescent is a stochastic process representing ancestral lineages in a population undergoing neutral genetic drift. Originally defined for a well-mixed population, the coalescent has been adapted in various ways to accommodate spatial, age, and class structure, along with other features of real-world populations. To further extend the range of population structures to which coalescent theory applies, we formulate a coalescent process for a broad class of neutral drift models with arbitrary – but fixed – spatial, age, sex, and class structure, haploid or diploid genetics, and any fixed mating pattern. Here, the coalescent is represented as a random sequence of mappings C=Ctt=0∞ from a finite set G to itself. The set G represents the “sites” (in individuals, in particular locations and/or classes) at which these alleles can live. The state of the coalescent, Ct:G→G, maps each site g∈G to the site containing g’s ancestor, t time-steps into the past. Using this representation, we define and analyze coalescence time, coalescence branch length, mutations prior to coalescence, and stationary probabilities of identity-by-descent and identity-by-state. For low mutation, we provide a recipe for computing identity-by-descent and identity-by-state probabilities via the coalescent. Applying our results to a diploid population with arbitrary sex ratio r, we find that measures of genetic dissimilarity, among any set of sites, are scaled by 4r(1−r) relative to the even sex ratio case.

State estimation and structure identification of complex networks with time‐varying delays and model uncertainty based on probabilistic graph neural network

SummaryIn this paper, the state estimation problem and structure identification problem are investigated for an array of complex networks with time‐varying delays and model uncertainty. Model uncertainty is generated by uncertain connections between nodes in complex networks. Combining probabilistic graph model and graph neural network, a probabilistic graph neural network is proposed to deal with the uncertainty, and the multiple linear terms of the errors between nodes are estimated. The purpose of the addressed state estimation problem is to design a state estimator such that, in the presence of the model uncertainty, the estimation error to converge to zero can be guaranteed and the explicit expression of the estimator parameters is given. Based on a Lyapunov function, the parameters of the estimator and the updating laws of the weights of probabilistic graph neural network can be obtained by the solutions to linear matrix inequalities. The purpose of the addressed structure identification problem is to infer a definite graph structure model based on Bayes' theorem in the presence of a stable probabilistic graph neural network and state estimator. Finally, an illustrative example is provided to demonstrate the feasibility and effectiveness of the developed state estimation and structure identification scheme.

Modeling correlated uncertainties in stochastic compartmental models

We consider compartmental models of communicable disease with uncertain contact rates. Stochastic fluctuations are often added to the contact rate to account for uncertainties. White noise, which is the typical choice for the fluctuations, leads to significant underestimation of the disease severity. Here, starting from reasonable assumptions on the social behavior of individuals, we model the contacts as a Markov process which takes into account the temporal correlations present in human social activities. Consequently, we show that the mean-reverting Ornstein–Uhlenbeck (OU) process is the correct model for the stochastic contact rate. We demonstrate the implication of our model on two examples: a Susceptibles–Infected–Susceptibles (SIS) model and a Susceptibles–Exposed–Infected–Removed (SEIR) model of the COVID-19 pandemic and compare the results to the available US data from the Johns Hopkins University database. In particular, we observe that both compartmental models with white noise uncertainties undergo transitions that lead to the systematic underestimation of the spread of the disease. In contrast, modeling the contact rate with the OU process significantly hinders such unrealistic noise-induced transitions. For the SIS model, we derive its stationary probability density analytically, for both white and correlated noise. This allows us to give a complete description of the model’s asymptotic behavior as a function of its bifurcation parameters, i.e., the basic reproduction number, noise intensity, and correlation time. For the SEIR model, where the probability density is not available in closed form, we study the transitions using Monte Carlo simulations. Our modeling approach can be used to quantify uncertain parameters in a broad range of biological systems.

End-to-end wind turbine design under uncertainties: a practical example

This paper illustrates the process of design under uncertainty on a practical case study of an offshore wind farm. We document the entire process through selection and quantification of relevant uncertainties, definition of probabilistic limit states, reliability computation algorithms, as well as illustrating the impacts of the analysis through a design utilization study. The brief introduction in this study draws information and summarizes outcomes from the extensive works that took part within the EU H2020 HIPERWIND project. The results from the study show that significant material savings can be achieved by introducing probabilistic design methodologies, and particularly with the help of an integrated modelling approach where the entire structure (turbine, tower & foundation) is considered as a whole.

Response of Gaussian white noise excited oscillators with inertia nonlinearity based on the RBFNN method

Although stochastic averaging methods have proven effective in solving the responses of nonlinear oscillators with a strong stiffness term under broadband noise excitations, these methods appear to be ineffective when dealing with oscillators that have a strong inertial nonlinearity term (also known as coordinate-dependent mass) or multiple potential wells. To address this limitation, a radial basis function neural network (RBFNN) algorithm is applied to predict the responses of oscillators with both a strong inertia nonlinearity term and multiple potential wells. The well-known Gaussian functions are chosen as radial basis functions in the model. Then, the approximate stationary probability density function (PDF) is expressed as the sum of Gaussian basis functions (GBFs) with weights. The squared error of the approximate solution for the Fokker-Plank-Kolmogorov (FPK) function is minimized using the Lagrange multiplier method to determine optimal weight coefficients. Three examples are presented to demonstrate how inertia nonlinearity terms and potential wells affect the responses. The mean square errors between Monte Carlo simulations (MCS) and RBFNN predictions are provided. The results indicate that RBFNN predictions align perfectly with those obtained from MCS.

State Identification Via Symbolic Time Series Analysis for Reinforcement Learning Control

Abstract This technical brief makes use of the concept of symbolic time-series analysis (STSA) for identifying discrete states from the nonlinear time response of a chaotic dynamical system for model-free reinforcement learning (RL) control. Along this line, a projection-based method is adopted to construct probabilistic finite state automata (PFSA) for identification of the current state (i.e., operational regime) of the Lorenz system; and a simple Q-map-based (and model-free) RL control strategy is formulated to reach the target state from the (identified) current state. A synergistic combination of PFSA-based state identification and RL control is demonstrated by the simulation of a numeric model of the Lorenz system, which yields very satisfactory performance to reach the target states from the current states in real-time.

Group consensus of nonlinear semi-Markov multi-agent systems by a quantity limited controller

This paper solves the group consensus problem of incremental quadratic constraints nonlinear multi-agent systems with semi-Markov switching parameters via a quantity limited controller. The quantity limited controller is designed according to the stationary probability distribution of the system modes, and the number of controllers can be selected from 1 to the number of system modes. Therefore, the controller can be applied in cases where there is a mismatch between the number of system modes and the number of controllers. Next, the more general incremental quadratic constraint is used to handle the nonlinear terms. Finally, in the simulation section, the method of adding topology links is used to implement external equitable partitions of arbitrary topologies, and the validity of the proposed method is verified by a spring damping system.

Cost optimization for system stability and orbital search under working vacation and starting failure by using the ANFIS soft computing

This article deals with the concepts of a single server retrial queueing model with balking, orbital search under working vacation and starting failure. If the server is busy, the prospective consumer may either enter the pool or leave the system; if the server is idle, the arriving client begins providing its services instantaneously. We assume the server is unstable since it did not start during the usual busy period. Instead of being completely unavailable at some point during the working vacation, the server will operate at a slow rate. Following the completion of the vacation period, the server either searches the orbit for clients or returns to idle mode, during which it resumes normal operations and continues to serve customers. The essential and adequate conditions for system stability have been discovered. The supplementary variable approach has been utilised to generate the stationary probability distribution and a few performances metrices. To justify the suggested paradigm, several numerical examples and special cases are presented. We have used the ANFIS soft computing strategy to assess the outcomes of the analytical and neural fuzzy methods.

Stochastic analysis and vibration suppression of a time-delayed system with nonlinear energy sink

Nonlinear energy sink (NES) is of great significance in attenuating structural vibration. Time-delay phenomenon in a mechanical system is inevitable. Different from existing works, the influences of time-delayed factors in a nonlinear oscillator attached with NES are considered in this paper under Gaussian white noise (GWN). Firstly, the time-delayed term is transformed into a linear combination of stiffness and damping terms, and an equivalent system without the time-delayed term is obtained. Stationary probability density function (SPDF) is deduced by leveraging stochastic averaging method. Secondly, good agreement is observed between analytical and numerical results, which verifies the effectiveness of the method. Furthermore, bifurcation conditions for the amplitude of SPDF of the primary system are obtained, indicating that delayed time, feedback gain and damping coefficients can cause stochastic P-bifurcation of the primary system. Finally, the vibration suppression performance of NES is analyzed via energy variation and the time–history diagram of displacement.

Feedback stabilization of probabilistic finite state machines based on deep Q-network.

As an important mathematical model, the finite state machine (FSM) has been used in many fields, such as manufacturing system, health care, and so on. This paper analyzes the current development status of FSMs. It is pointed out that the traditional methods are often inconvenient for analysis and design, or encounter high computational complexity problems when studying FSMs. The deep Q-network (DQN) technique, which is a model-free optimization method, is introduced to solve the stabilization problem of probabilistic finite state machines (PFSMs). In order to better understand the technique, some preliminaries, including Markov decision process, ϵ-greedy strategy, DQN, and so on, are recalled. First, a necessary and sufficient stabilizability condition for PFSMs is derived. Next, the feedback stabilization problem of PFSMs is transformed into an optimization problem. Finally, by using the stabilizability condition and deep Q-network, an algorithm for solving the optimization problem (equivalently, computing a state feedback stabilizer) is provided. Compared with the traditional Q learning, DQN avoids the limited capacity problem. So our method can deal with high-dimensional complex systems efficiently. The effectiveness of our method is further demonstrated through an illustrative example.

Complexity-calibrated benchmarks for machine learning reveal when prediction algorithms succeed and mislead

Recurrent neural networks are used to forecast time series in finance, climate, language, and from many other domains. Reservoir computers are a particularly easily trainable form of recurrent neural network. Recently, a “next-generation” reservoir computer was introduced in which the memory trace involves only a finite number of previous symbols. We explore the inherent limitations of finite-past memory traces in this intriguing proposal. A lower bound from Fano’s inequality shows that, on highly non-Markovian processes generated by large probabilistic state machines, next-generation reservoir computers with reasonably long memory traces have an error probability that is at least ∼60%\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\sim 60\\%$$\\end{document} higher than the minimal attainable error probability in predicting the next observation. More generally, it appears that popular recurrent neural networks fall far short of optimally predicting such complex processes. These results highlight the need for a new generation of optimized recurrent neural network architectures. Alongside this finding, we present concentration-of-measure results for randomly-generated but complex processes. One conclusion is that large probabilistic state machines—specifically, large ϵ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\epsilon$$\\end{document}-machines—are key to generating challenging and structurally-unbiased stimuli for ground-truthing recurrent neural network architectures.