Particle-Based H-infinity Filter Scheme and Numerical Experiments

The H_∞ filter technique, which was developed in signal processing and control field in recent years, is introduced into data fusion method. Based on the principle of particles of forward model, a new sequential data fusion method called "particle H_∞ filter" is introduced into this paper by integrating H_∞ filter technique and Monte Carlo method. This method can be used for nonlinear systems lacking the statistical properties of observational errors. To find the smallest level factor γ that could bring optimal estimations, bisection search method is embedded in the arithmetic of particle H_∞ filter and their calculation steps are also given. The numerical results also show that particle H_∞ filter data fusion method is effective and suitable to nonlinear systems in that it does not rely on the statistical properties of observational errors and has better robustness.