Distributed quantized multi-modal H_∞ fusion filtering for two-time-scale systems

In this paper, we deal with the distributed H_∞ fusion filtering problem for a class of discrete-time system with significant modal difference in the time-scale. The nonstandard singularly perturbed model (SPM) is employed to describe the addressed two-time-scale system, where the small perturbation parameter ε reflects the discrepancies in the time-scale. Some sufficient conditions are provided to guarantee the existence of the parameters of the local filters for the SPM, where O(ε²) (the second-order correction terms in ε) is eliminated and an upper bound of ε is calculated. To enhance the filtering accuracy and scalability, the estimates provided by the local filter are transmitted to the fusion centre via the communication network. In the presence of the bandwidth constraint, a quantized fusion scheme is developed and the design of the fusion parameters is cast into a convex optimization problem. Finally, an example on the F-8 airplane is provided to verify the validity and applicability of the proposed methodology.