Propagation of chaos for Belavkin equations beyond pure states

Abstract

We use probabilistic and stochastic-analysis methods to prove trace-norm propagation of chaos for finite-dimensional quantum mean-field systems governed by Belavkin equations. The particles are density matrices, interact through a mean-field Hamiltonian, and are continuously monitored through independent diffusive observation channels. The limiting dynamics is a nonlinear matrix-valued McKean-Vlasov diffusion, random through its local observation record and coupled through the deterministic averaged state. The main result treats arbitrary one-particle density matrices, including mixed states, and both perfect and inefficient measurement regimes. Under strong tensorization of the initial data, every fixed marginal converges uniformly on compact time intervals to the tensor product of the nonlinear limiting filters, with an explicit quantitative bound. The proof combines purification, fully observed dilation, conditional expectation, relative entropy, and uniform stability of the associated Zakai equations. In the skew-adjoint measurement case, exterior observation noises disappear from the marginal equations and a stochastic BBGKY hierarchy is recovered. Under only marginal chaoticity of permutation-invariant initial states, we prove convergence of fixed marginals by an iteration of this hierarchy.