Integro-differential Chapman-Kolmogorov equation for continuous-jump Markov processes and its use in problems of multi-component renewal impulse process excitations

In the present paper the applications of the integro-differential Chapman-Kolmogorov equation to the problems of pure-jump stochastic processes and continuous-jump response processes are discussed. The pure-jump processes considered herein are the counting Poisson process, a two-state jump process, and a multi-state jump process. The differential equations governing the Markov state probabilities are obtained from the degenerate, pure differential form, of the general, integro-differential Chapman-Kolmogorov equation, with the aid of the jump probability intensity functions. The continuous-jump response process is the response of a dynamic system to a multi-component renewal impulse process excitation. The excitation consists of a number of n statistically independent random trains of impulses, each of which is driven by an Erlang renewal process with parameters ν_j,k_j. Each of the impulse processes is characterized by an auxiliary zero-one jump stochastic process, which consists of k_j negative exponential distributed phases. The Markov states for the whole problem are determined by the coincidences of the phases of the individual jump processes. Thus the response probability distribution may be characterized by a joint probability density-discrete distribution of the state variables of the dynamic system and of the states of the pertinent Markov chain. The explicit integro-differential equations governing the joint probability density-discrete distribution of the response are obtained from the general forward integro-differential Chapman-Kolmogorov equation, after the determination of the jump probability intensity functions for the continuous-jump and pure-jump processes.     

Response of dynamic systems to renewal impulse processes: Generating equation for moments based on the integro-differential Chapman–Kolmogorov equations

In the present paper the method is developed for the derivation of differential equations for statistical moments of the state vector (response) of a non-linear dynamic system subjected to a random train of impulses. The arrival times of the impulses are assumed to be driven by a non-Poisson counting process. The state vector of the dynamic system is then a non-Markov process and no method is directly available for the derivation of the equations for response moments. The original non-Markov problem is converted into a Markov one by recasting the excitation process with the aid of an auxiliary, pure-jump stochastic process characterized by a Markov chain. Hence the conversion is carried out at the expense of augmentation of the state space of the dynamic system by auxiliary Markov states. For the augmented problem the sets of forward and backward integro-differential Chapman–Kolmogorov equations are formulated. The general, generating equation for moments is obtained with the aid of the forward and backward integro-differential Chapman–Kolmogorov operators. The developed method is illustrated by the examples of several renewal impulse processes.     

Response of linear dynamic systems to non-Erlang renewal impulses: Stochastic equations approach

Linear dynamical systems under random trains of impulses driven by a class of non-Erlang renewal processes are considered. The class considered is the one where the renewal events are selected from an Erlang renewal process. The original train of impulses is recast, with the aid of an auxiliary stochastic variable, in terms of two independent Poisson processes. Thus, by augmenting the state vector of the dynamic system with the auxiliary stochastic variables, the original non-Markov problem is converted to a Markov one.

The differential equations for the response statistical moments can then be derived from the generalised Ito's differential rule.

Numerical results obtained for a few different models and various sets of parameters, show that the present approach allows to account for a variety of inter arrival time's probability distributions. Transient mean value and variance of the response of a linear oscillator have been obtained from the equations for moments.     

A method for the evaluation of the response probability density function of some linear dynamic systems subjected to non-Gaussian random load

This paper deals with the characterization of the random response of linear systems subjected to stochastic load. It proposes a new method based on the new version of the Probabilistic Transformation Method (PTM) that allows obtaining, with a very low computational effort, the probability density function of the response. An important aspect of the proposed approach is the ability to join directly the pdfs of the input load with those of the response. Based on the step-by-step integration method, explicit solutions will be proposed for the random response of systems loaded by seismic and windy sampled inputs.     

Stochastic direct integration schemes for dynamic systems subjected to random excitations

In this work, a stochastic version of direct integration schemes is constructed, based on a general recursive state space formulation. The technique is applicable to evaluating the second order response statistics of systems subjected to non-stationary random excitations, and is potentially able to handle non-proportional damping. A specific group of these algorithms obtained using one of the most frequently used families of time stepping schemes is introduced. It is shown that statistics obtained for two multi-degree of freedom (MDOF) dynamic systems using the proposed method are in close agreement with those obtained from Monte Carlo simulation when a proper time step is used. Stability of the proposed scheme is systematically investigated and it is concluded that, for the response statistics to be bounded, the time step used in the algorithm has to lie within certain limits.     

Probability density evolution method for dynamic response analysis of structures with uncertain parameters

Probability density evolution method is proposed for dynamic response analysis of structures with random parameters. In the present paper, a probability density evolution equation (PDEE) is derived according to the principle of preservation of probability. With the state equation expression, the PDEE is further reduced to a one-dimensional partial differential equation. The numerical algorithm is studied through combining the precise time integration method and the finite difference method with TVD schemes. The proposed method can provide the probability density function (PDF) and its evolution, rather than the second-order statistical quantities, of the stochastic responses. Numerical examples, including a SDOF system and an 8-story frame, are investigated. The results demonstrate that the proposed method is of high accuracy and efficiency. Some characteristics of the PDF and its evolution of the stochastic responses are observed. The PDFs evidence heavy variance against time. Usually, they are much irregular and far from well-known regular distribution types. Additionally, the coefficients of variation of the random parameters have significant influence on PDF and second-order statistical quantities of responses of the stochastic structure.The support of the Natural Science Funds for Distinguished Young Scholars of China (Grant No.59825105) and the Natural Science Funds for Innovative Research Groups of China (Grant No.50321803) are gratefully appreciated.     

Reliability of dynamic systems under limited information

A method is developed for reliability analysis of dynamic systems under limited information. The available information includes one or more samples of the system output; any known information on features of the output can be used if available. The method is based on the theory of non-Gaussian translation processes and is shown to be particularly suitable for problems of practical interest. For illustration, we apply the proposed method to a series of relevant examples and compare with results given by traditional statistical estimators. It is demonstrated that the method delivers accurate results for the case of linear and nonlinear dynamic systems, and can be applied to analyze experimental data and/or mathematical model outputs.     

Analysis of multi-degree of freedom strongly non-linear mechanical systems with random input: Part II: equivalent linear system with random matrices and power spectral density matrix

We present a method for estimating the (power spectral density) PSD matrix of the stationary response of lightly damped randomly excited multi-degree of fredom mechanical systems with strong non-linear asymmetrical restoring forces. The PSD matrix is defined as the mean value of the PSD matrix response of an equivalent linear system (ELS) whose damping and stiffness matrices depend on non-linear vibration modes of the associated conservative system, the frequencies and modes shapes being amplitude dependent. Based on a generalized van der Pol transformation and using a stochastic averaging principle, as developed in a companion paper, a stationary probability density function for the amplitude process is derived to characterize the ELS fully. Some possible simplifications of the method, such as modal reduction and/or local linearization, are also discussed. The results obtained are in good agreement with those of direct numerical simulations taking two typical examples.     

Equivalent nonlinearization of nonlinear systems to random excitations

The broadest class of solvable reduced Fokker-Planck equation is given and a new equivalent nonlinear method is presented to obtain an approximate probability density function for the response of a nonlinear oscillator to Gaussian white noise excitations. The method is based on the least mean-square criterion and Euler equation. It is shown that this method, which is simpler and more reasonable, generalizes Caughey's method and gives the same results as Cai and Lin under purely additive excitations. Examples are given to show the applications of the method. In one of the examples, this method leads to a better approximation than that obtained from the energy dissipation criterion.     

Almost-sure stability of a class of distributed parameter systems subjected to random excitations

The stability of a class of distributed parameter systems subjected to stationary and non-stationary randomx excitations is studied. Several criteria regarding the almost-sure stabihty of the equilibrium state of the system are developed. A variational scheme for obtaining sharp stability criteria is formulated-Examples of flutter of a panel in a random wind and stability of a tall structure under a seismic load are solved to illustrate the applicability of the method.     

On deriving exact probability density functions for functionally graded exponential profile orthotropic plates driven by laterally random excitation

In this article, the exact probability density function for a nonlinear exponential functionally graded orthotropic plate under lateral random excitation is investigated. After that, a study on the probability density function (PDF) is done to examine the dynamic instability and bifurcation. Using nondimensional parameters, the results are justified for a wide range of plates. The outcomes are studied with respect to the variation of both mean value of lateral loads and in-plane forces. The famous Monte Carlo simulation is employed to validate the obtained analytical PDFs. An analogical study is done between the behaviors of homogenous plates with their corresponding functionally graded material ones. Finally, it is shown how the exponential profile functionally graded orthotropic material can affect the quality of instability and bifurcation.     

State-space independent component analysis for nonlinear dynamic process monitoring

The cost effective benefits of process monitoring will never be over emphasised. Amongst monitoring techniques, the Independent Component Analysis (ICA) is an efficient tool to reveal hidden factors from process measurements, which follow non-Gaussian distributions. Conventionally, most ICA algorithms adopt the Principal Component Analysis (PCA) as a pre-processing tool for dimension reduction and de-correlation before extracting the independent components (ICs). However, due to the static nature of the PCA, such algorithms are not suitable for dynamic process monitoring. The dynamic extension of the ICA (DICA), similar to the dynamic PCA, is able to deal with dynamic processes, however unsatisfactorily. On the other hand, the Canonical Variate Analysis(CVA) is an ideal tool for dynamic process monitoring, however is not sufficient for nonlinear systems where most measurements follow non-Gaussian distributions. To improve the performance of nonlinear dynamic process monitoring, a state space based ICA (SSICA) approach is proposed in this work. Unlike the conventional ICA, the proposed algorithm employs the CVA as a dimension reduction tool to construct a state space, from where statistically independent components are extracted for process monitoring. The proposed SSICA is applied to the Tennessee Eastman Process Plant as a case study. It shows that the new SSICA provides better monitoring performance and detect some faults earlier than other approaches, such as the DICA and the CVA.     

Model selection for a class of stochastic processes or random fields with bounded range

Methods are developed for finding an optimal model for a non-Gaussian stationary stochastic process or homogeneous random field under limited information. The available information consists of: (i) one or more finite length samples of the process or field; and (ii) knowledge that the process or field takes values in a bounded interval of the real line whose ends may or may not be known. The methods are developed and applied to the special case of non-Gaussian processes or fields belonging to the class of beta translation processes. Beta translation processes provide a flexible model for representing physical phenomena taking values in a bounded range, and are therefore useful for many applications. Numerical examples are presented to illustrate the utility of beta translation processes and the proposed methods for model selection.     

Numerical solutions of Fokker-Planck equation of nonlinear systems subjected to random and harmonic excitations

A new approach for an efficient numerical implementation of the path integral (PI) method based on non-Gaussian transition probability density function (PDF) and the Gauss-Legendre integration scheme is developed. This modified PI method is used to solve the Fokker-Planck (FP) equation and to study the nature of the stochastic and chaotic response of the nonlinear systems. The steady state PDF, periodicity, jump phenomenon, noise induced changes in joint PDF of the states are studied by the modified PI method. A computationally efficient higher order, finite difference (FD) technique is derived for the solution of higher-dimensional FP equation. A two degree of freedom nonlinear system having Coulomb damping with a variable friction coefficient subjected to Gaussian white noise excitation is considered as an example which can represent a bladed disk assembly of turbo-machinery blades. Effects of normal force and viscous damping on the mean square response are investigated.     

Stationary probability density of a class of non-linear dynamical systems to stochastic excitation

The response of a dynamical system to Gaussian white-noise excitations may be represented by the Markov process whose probability density is governed by the well-known Fokker-Plank equation. In this paper a general procedure is developed to obtain the solution for Fokker-Plank equation in the state of statistical stationary. The dynamical systems considered are generally non-linear. This paper also demonstrates that the stationary joint probability density of the coordinates and velocity may possess the form of a separate product for the systems both non-linear damping and non-linear restoring forces.     

A study of an n-unit system operating in a random environment

An n-unit system operating in a random environment is considered. The environment determines the number of units required for the satisfactory performance of the system. Assuming that the environment is described by a Markov process, we obtain expressions for: (1) the distribution and the moments of the time to the first disappointment; and (2) the mean number of disappointments over an arbitrary interval.     

Linear dynamical systems under random trains of non-overlapping, arbitrary-shape pulses: generalized approach

The excitation considered is a train of non-overlapping pulses, defined as a train of alternate random pulse durations and time gaps between the consecutive pulses. The time gaps and durations are assumed to be independent, continuous random variables. In this model the pulses are not only non-overlapping, but they are also non-adjoining, i.e. they can only adjoin with probability zero. All durations are assumed to be characterized by one common, Erlang (or integer parameter gamma) type, probability distribution. Likewise, the time gaps are all identically, Erlang distributed, but the probability distributions of the durations and of the time gaps have different parameters. Pulse shapes are given by an arbitrary, deterministic function and pulse heights are given by independent, identically distributed random variables. The mean value and autocorrelation function of the pulse train are evaluated analytically and are shown to be expressed in terms of the renewal densities of the renewal process governing the pulse arrivals. It is shown that for rectangular pulses with equal, deterministic heights these expressions reduce to the ones obtained independently with the aid of another approach, where the mean value and autocorrelation function are given in terms of Markov ‘on’ state probabilities. The mean value and variance of the response of a linear oscillator are obtained via a time domain analysis. Illustrative numerical analysis is carried out for example trains of pulses with parabolic and rectangular pulse shapes.     

An original approximate method for estimating the invariant probability distribution of a large class of multi-dimensional nonlinear stochastic oscillators

This paper proposes an original method for obtaining analytical approximations of the invariant probability density function of multi-dimensional Hamiltonian dissipative dynamic systems under Gaussian white noise excitations, with linear non-conservative parts and nonlinear conservative parts. The method is based on an exact result and a heuristic argument. Its pertinence is attested by numerical tests.     

A posteriori estimate of the random response of a dynamic system with autocorrelated additive noises

The signal recorded during the measurement of a mechanical system response is a mixture of a useful signal and random noises. On the one hand, these noises enter the investigated system directly, and on the other hand they are added to the signal during its transmission to the measuring device and its further processing. Thus, the measured signals are a stochastic function of the response and, consequently, provide only indirect, but the only information of the actual response of the investigated system. The problem is to determine most accurately the actual response of the investigated system by virtue of known outputs from the measuring system. As such a problem is not sufficiently determined in a general case, but it can be solved as an optimum estimate of actual response. It is coming to light, however, that in the meaning of the minimum mean square deviation of actual response and of this estimate the optimum a posteriori first stochastic moment of the response is conditioned by the measured signals. This article describes the principal characteristics of the mutually influenced investigated system and the measuring system. After a short analysis of the principal variant of the problem, assuming that all noises entering the system are Gaussian white noises, principal attention is afforded to the cases when the noises are described by more complicated autocorrelation functions. Cross correlation of noises, which are considered as additive only, is neglected.