On preserving dissipativity properties of linear complementarity dynamical systems with the \(\theta\)-method (Q2435054)

From MaRDI portal
scientific article
Language Label Description Also known as
English
On preserving dissipativity properties of linear complementarity dynamical systems with the \(\theta\)-method
scientific article

    Statements

    On preserving dissipativity properties of linear complementarity dynamical systems with the \(\theta\)-method (English)
    0 references
    0 references
    0 references
    0 references
    3 February 2014
    0 references
    The authors consider initial value problems of linear dynamical systems including a linear complementarity problem, where the solution consists of state variables and Lagrange multipliers. This formulation can be seen as differential algebraic equations. For a numerical solution in the time domain, an extension of the \(\theta\)-method for ordinary differential equations is applied by using two different parameters \(\theta\) and \(\tau\) for the discretisation of the state variables and the multipliers, respectively. An energy storage function as well as a dissipation function is defined for the original time-continuous system and for the time-discrete system from the numerical method. Moreover, the authors specify a numerical dissipation property. Firstly, smooth solutions of the time-continuous system are considered. Assuming that both the continuous problem and the discrete problem are passive, the authors derive conditions on the parameters of the numerical method such that the two systems exhibit the same energy storage function (except for higher-order terms) and the same dissipation function independent of the choice of the time step size \(h>0\). Secondly, non-smooth solutions are investigated, i.e., the state variables are discontinuous in time. This class of problems was also studied in the previous work [\textit{V. Acary} et al., Nonsmooth modeling and simulation for switched circuits. Dordrecht: Springer (2011; Zbl 1208.94003)]. For the time-continuous problem, the authors identify a jump condition for the state variables such that the dissipative properties are preserved. For the time-discrete problem with \(\tau=1\) it is shown that a consistent approximation of the jump condition is guaranteed in the limit \(h \rightarrow 0\) of the time step size. The authors investigate two test examples with linear models of electric circuits, where the two different cases of smooth and discontinuous solutions are considered. The matrices of these linear problems are analysed with respect to the theoretical properties. Furthermore, the numerical simulations confirm the time-invariance of the sum of energy storage function and dissipation function for appropriate choices of the discretisation parameters.
    0 references
    0 references
    linear dynamical system
    0 references
    complementarity problem
    0 references
    differential algebraic equations
    0 references
    one-step method
    0 references
    passivity
    0 references
    dissipativity
    0 references
    consistency
    0 references
    jump condition
    0 references
    electric circuits
    0 references
    numerical examples
    0 references
    initial value problem
    0 references
    \(\theta\)-method
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references