State estimation for uncertain discrete-time nonlinear systems via the coding-decoding scheme

Abstract

The state estimation problem is investigated for a class of discrete-time networked systems with uncertainty and nonlinearity via a coding-decoding communication approach. The signal is transmitted through digital channels, which means that only the sequence of finite coded data is sent to the estimator. To be specific, the observation must be encoded to certain codewords by designing a coder, and then the received codewords are decoded by the relevant decoder. The purpose of the addressed problem is to design a proper coding-decoding procedure such that the estimator error is asymptotically stable across the limited capacity of the communication channel. By applying a modified uniform quantization approach, sufficient detection conditions involving the size of the coding alphabet and the coding period are obtained. Subsequently, the estimator parameter is explicitly determined in terms of the solutions to a set of matrix inequalities. Finally, a simulation example is provided to demonstrate the effectiveness of the obtained results.