Numerical simulation of stochastic evolution equations associated to quantum Markov semigroups

From MaRDI portal





Consider the problem of computation of expectations \(E \langle X_t, A X_t\rangle\), where \(A\) is a linear operator on a separable complex Hilbert space \((H, \langle \cdot, \cdot \rangle)\) and \((X_t)_{t \in [0,T]}\) is an \(H\)-valued diffusion process satisfying the Itô-type stochastic evolution equation \[ dX_t = G X_t dt + \sum_{k=1}^m L_k X_t dW^k_t, \qquad X_0 = x_0 \in H, \] driven by independent standard Brownian motions \(W^k\), \(L_1, \dots, L_m\) operators on \(H\) and \(G = - i B - \frac{1}{2} \sum^m_{j=1} L^*_j L_j\) with self-adjoint operator \(B\) on \(H\). The author introduces approximations \(X_{t,n}\) of \(X_t\) via orthogonal projection solving an approximating (finite-dimensional) equation of the above one with truncated operators \(L_{1,n}, \dots, L_{m,n}\) and \( G_{n}\). These approximating equations preserve the mean square conservative relationship \(E | | X_t| | ^2 = | | x_0| | ^2\) in a wide class of applications. As the main result, he verifies (naturally very complex) worst-case estimates on the rate of convergence of \(E \langle \tilde{X}_{t,n}, A \tilde{X}_{t,n} \rangle\) to \(E \langle X_t, A X_t \rangle\) where \(\tilde{X}_{t,n}\) is obtained by the explicit Euler approximation. He also considers an extrapolation method based on the explicit Euler scheme. Eventually, a quantum master equation for a boson system (quantum harmonic oscillator) is studied as a first application.











This page was built for publication: Numerical simulation of stochastic evolution equations associated to quantum Markov semigroups

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4813589)